Configurable DB Instance Parameters

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

cn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

cn:behavior_compat_options

Specifies database compatibility configuration items. After the value of proc_outparam_override is changed, the database must be connected again or the instance must be rebooted. Otherwise, stored procedures and functions cannot be correctly called.

dn:recyclebin_retention_time

Specifies how long files will be kept in the recycle bin, in seconds. Files in the recycle bin will be automatically deleted after this length of time.

dn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a session. If the number of full or slow SQL statements exceeds this value, new statements will not be traced until the flush thread flushes the cached statements to the disk to reserve free space. The default value is recommended.

timezone

Specifies the time zone for displaying and interpreting time stamps.

cn:auto_increment_offset

Specifies the initial value of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is greater than that of auto_increment_increment, there will be an error when the values in the auto-increment column automatically increase.

dn:enable_xid_abort_check

Specifies whether to check the status of transaction ID rollback when a transaction is committed.

cn:audit_internal_event

Specifies whether to audit the connections and operations of internal tools cm_agent, gs_clean, and WDRXdb, and whether to audit the logins and logouts from CNs on DNs.

cn:codegen_compile_thread_num

Specifies the number of Codegen compilation threads.

dn:static_thread_pool_num

Specifies the number of threads used to create a static thread pool (static pool). This parameter takes effect only on CNs of a distributed instance after enable_thread_pool is enabled.

cn:auto_increment_increment

Specifies the auto-increment step of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is smaller than that of auto_increment_offset, there will be an error when the values in the auto-increment column automatically increase.

cn:sql_mode

Specifies the SQL behavior control configuration item in M-compatible mode.

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

cn:llvm_max_memory

Specifies the upper limit of the memory used by IRs (including cached and in-use IRs) generated during compilation in Codegen execution mode. The memory used by Codegen is not applied for by preoccupation. It is a part of max_dynamic_memory and is restricted by the llvm_max_memory parameter. Unit: KB

cn:hll_default_log2explicit

Specifies the threshold for switching from the explicit mode to the sparse mode.

dn:fix_func_selection

Specifies whether to optimize the function matching policy. The value catlist indicates the catlist sequence is optimized. (The non-B-compatible mode has been optimized. In non-B-compatible mode, system functions are always preferentially selected and executed. The policy in B-compatible mode is the same as that in versions earlier than 505.1.0. An error message indicating that the function is not unique may be displayed, or a system function may be selected for execution.)

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

global_syscache_threshold

Specifies the maximum memory usage of the global system cache. Recommended value range: 2,048 KB to 16,384 KB on average per database. If this parameter is set to a large value, the cache link may be too long and performance will deteriorate. If this parameter is set to a small value, the parameter is not applied, and after the memory usage can exceed the threshold, performance deteriorates. Unit: KB

dn:sql_mode

Specifies the SQL behavior control configuration item in M-compatible mode.

dn:verify_log_buffers

Specifies the size or pages of verifyLog buffers in memory mode. The unit is 8 KB. For example, if the value of this parameter is 4, the requested memory is 4 x 8 KB = 32 KB. This parameter is valid only when page_version_check is set to persistence. If page_version_check is set to another value, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to persistence.

dn:resilience_ctrlstmt_control_iopslimit

Specifies the maximum IOPS that can be used by slow SQL statements after normal SQL statements are marked as slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. 0(None): The IOPS is not limited. 10(LOW): The limit level for IOPS is LOW. 20(MEDIUM): The limit level for IOPS is MEDIUM. 50(HIGH): The limit level for IOPS is HIGH.

cn:convert_illegal_char_mode

Specifies the placeholders of invalid characters that can be displayed on the client when enable_convert_illegal_char is enabled. Value range: 95 characters whose decimal codes range from 32 to 126 in the ASCII coding table.

cn:page_version_check

Specifies the type of page version verification. off indicates that page version verification is disabled. memory indicates that page version verification in pure memory mode is enabled. The page version information will be lost after a restart. persistence indicates that persistent page version verification is enabled. The page version information will not be lost after a restart.

cn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

recovery_time_target

Specifies whether the standby DB instance completes log writing and replay in streaming DR mode. If this parameter is set to a small value, the performance of the primary node is affected. If it is set to a large value, the log flow is not effectively controlled. The value 0 indicates that log flow control is disabled. Unit: second

dn:enable_convert_illegal_char

Specifies whether the database supports characters not included the character sets.

dn:system_view_version

Determines the version of the system view. All versions are backward compatible. For example, when system_view_version is set to 3, all features of version 2 and version 1 are also supported. For details, see the product documentation.

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:codegen_compile_thread_num

Specifies the number of Codegen compilation threads.

dn:resilience_ctrlstmt_detect_timelimit

Specifies the execution time of a normal SQL statement that will be marked as a slow SQL statement. The value 0 indicates that slow SQL statements are not identified. A value greater than 0 indicates that a normal SQL statement whose execution time exceeds the value of this parameter is marked as a slow SQL statement. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. Unit: millisecond

cn:a_format_date_timestamp

Specifies whether to enable current_date, current_timestamp, and localtimestamp to return the system time, instead of the transaction start time, when a transaction starts.

cn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:max_compile_functions

Specifies the maximum number of function compilation results stored in the server. Excessive functions and compilation results of stored procedures may occupy large memory space. Setting this parameter to an appropriate value can reduce the memory usage and improve system performance. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

cn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions). The value 0 means to disable the function, and 1 means to enable it.

dn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: millisecond

cn:num_internal_lock_partitions

Specifies the number of internal lightweight lock partitions. Changing the value of this parameter affects performance and memory usage. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

dn:recovery_parse_workers

Specifies the number of ParseRedoRecord threads in the ultimate RTO feature. This parameter must be used together with recovery_redo_workers. If both recovery_parse_workers and recovery_redo_workers are greater than 1, ultimate RTO is enabled. If you do not want to enable ultimate RTO, retain the default value 1 for recovery_parse_workers. When enabling ultimate RTO, ensure that replication_type is set to 1. If both ultimate RTO and parallel replay are enabled, only ultimate RTO takes effect, and parallel replay is ineffective. Ultimate RTO does not support column-store tables. Therefore, disable ultimate RTO in a system where column-store tables are used or are to be used. Ultimate RTO also does not support flow control. Flow control is determined by the recovery_time_target parameter.

dn:enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

cn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

dn:page_version_check

Specifies the type of page version verification. off indicates that page version verification is disabled. memory indicates that page version verification in pure memory mode is enabled. The page version information will be lost after a restart. persistence indicates that persistent page version verification is enabled. The page version information will not be lost after a restart.

cn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

cn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: millisecond

dn:num_internal_lock_partitions

Specifies the number of internal lightweight lock partitions. Changing the value of this parameter affects performance and memory usage. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

cn:page_version_max_num

Specifies the maximum number of page versions that can be cached in memory. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

dn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions). The value 0 means to disable the function, and 1 means to enable it.

cn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:auto_increment_cache

Specifies the number of reserved auto-increment cache values when auto-increment is triggered by batch insertion or import of auto-increment columns. When auto-increment values are reserved, the auto-increment counter value is updated to the maximum auto-increment cache value. Before the cache values are used up, the auto-increment counter value remains unchanged, and the triggered auto-increment uses the cache values. If this parameter is set to 0, the auto-increment cache values are automatically reserved. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, 2^n values are reserved each time. For example, one value is reserved in the first auto-increment, two values are reserved in the second auto-increment, four values are reserved in the third auto-increment, and eight values are reserved for in fourth auto-increment. However, if the number of reserved values exceeds 65,535, 65,535 values are reserved. If this parameter is not set to 0, the number of reserved cache values is the value of this parameter. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, the value of auto_increment_cache is the number of auto-increment values reserved each time. The reserved cache values are valid only in the statement. If the reserved auto-increment cache values are used up and subsequent INSERT statements trigger auto-increment based on the auto-increment counter, the values in the auto-increment column in the table are discontinuous. This parameter does not affect the auto-increment column in the local temporary table.

dn:enable_codegen

Specifies whether code optimization can be enabled. Currently, code optimization uses the LLVM optimization.

dn:instr_unique_sql_combination_options

Specifies the configuration items of combining unique SQL statements of the same type. If this feature is enabled, the IDs of unique SQL statements of the same type are normalized, and the generated unique SQL strings are normalized.

dn:behavior_compat_options

Specifies database compatibility configuration items. After the value of proc_outparam_override is changed, the database must be connected again or the instance must be rebooted. Otherwise, stored procedures and functions cannot be correctly called.

dn:enable_early_free

Specifies whether the operator memory can be released in advance.

dn:page_missing_dirty_check

Specifies whether to enable the verification for pages not marked as dirty. The verification checks whether the modified pages are not marked as dirty. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

cn:undo_retention_time

Specifies how long undo logs are kept. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained. Unit: second

dn:wdr_snapshot_full_backup_interval

Specifies the interval at which a full WDR snapshot is created. The interval specified by this parameter is about a number instead of time. For example, if the parameter is set to 12, a full snapshot and then 11 incremental snapshots are generated for each group. If the parameter is set to 1, all snapshots generated are full snapshots.

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:wdr_snapshot_space_threshold

Specifies the threshold for controlling the space used by snapshots. When the space used by snapshots reaches 80% of the value of this parameter, the control logic of the database is enabled to stabilize the space usage. Unit: KB

cn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

dn:hll_duplicate_check

Specifies whether duplicate check is enabled by default.

cn:gs_perf_interval

Specifies the automatic perf data collection interval. The value 0 indicates that the collection is stopped. If the value is greater than 0 and less than 5, the value 5 is used. Unit: minute

cn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:gs_perf_interval

Specifies the automatic perf data collection interval. The value 0 indicates that the collection is stopped. If the value is greater than 0 and less than 5, the value 5 is used. Unit: minute

dn:extra_float_digits

Adjusts the number of digits displayed for floating-point values, including float4, float8, and geometric data types. The parameter value is added to the standard number of digits (FLT_DIG or DBL_DIG as appropriate). This parameter can also be set to a negative value to suppress unwanted digits.

cn:disable_keyword_options

Specifies the configuration items for database compatibility. Multiple items are separated by commas (,). An identifier with this parameter set will not be used as a keyword.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

cn:max_standby_archive_delay

Specifies the wait period before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

track_stmt_stat_level

Controls the level of statement execution tracking.

dn:max_standby_archive_delay

Specifies the wait period before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

cn:b_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

dn:copy_special_character_version

Specifies whether to report an error when there are invalid characters during data import and export using COPY FROM.

cn:page_version_recycler_thread_num

Specifies the number of threads for recycling and verifying page version information. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

enable_wdr_snapshot

Specifies whether to enable WDR snapshots. If this parameter is enabled, the kernel periodically takes performance snapshots.

dn:effective_cache_size

Specifies the size of the disk buffer available to the DN optimizer in a single query. Unit: 8 KB

cn:wdr_snapshot_full_backup_interval

Specifies the interval at which a full WDR snapshot is created. The interval specified by this parameter is about a number instead of time. For example, if the parameter is set to 12, a full snapshot and then 11 incremental snapshots are generated for each group. If the parameter is set to 1, all snapshots generated are full snapshots.

cn:extra_float_digits

Adjusts the number of digits displayed for floating-point values, including float4, float8, and geometric data types. The parameter value is added to the standard number of digits (FLT_DIG or DBL_DIG as appropriate). This parameter can also be set to a negative value to suppress unwanted digits.

cn:system_view_version

Determines the version of the system view. All versions are backward compatible. For example, when system_view_version is set to 3, all features of version 2 and version 1 are also supported. For details, see the product documentation.

cn:recyclebin_retention_time

Specifies how long files will be kept in the recycle bin. Files in the recycle bin will be automatically deleted after this length of time. Unit: second

cn:password_encryption_type

Specifies how user passwords are encrypted. 0: Passwords are encrypted using MD5. 1: Passwords are encrypted using SHA-256 and MD5. 2: Passwords are encrypted using SHA-256. 3: Passwords are encrypted using SM3. MD5 is not recommended because it is a weak encryption algorithm.

cn:check_disconnect_query

Specifies whether to enable the function of terminating statement execution on the server after the client is disconnected due to timeout.

password_effect_time

Specifies the validity period of an account password. Unit: day

cn:verify_log_buffers

Specifies the size or pages of verifyLog buffers in memory mode. The unit is 8 KB. For example, if the value of this parameter is 4, the requested memory is 4 x 8 KB = 32 KB. This parameter is valid only when page_version_check is set to persistence. If page_version_check is set to another value, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to persistence.

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

cn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

cn:enable_rls_match_index

Specifies whether indexes of a base table can be scanned based on target predicate conditions in row-level security scenarios. Target scenario: Row level security (RLS) policies are set and enabled in the base table, and the query predicate contains the unleakproof system function or like operator.

dn:enable_analyze_check

Specifies whether to check if statistics were collected about tables whose reltuples and relpages are displayed as 0 in pg_class during plan generation.

cn:resilience_ctrlstmt_control_iopslimit

Specifies the maximum IOPS that can be used by slow SQL statements after normal SQL statements are marked as slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. 0(None): The IOPS is not limited. 10(LOW): The limit level for IOPS is LOW. 20(MEDIUM): The limit level for IOPS is MEDIUM. 50(HIGH): The limit level for IOPS is HIGH.

dn:auto_increment_increment

Specifies the auto-increment step of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is smaller than that of auto_increment_offset, there will be an error when the values in the auto-increment column automatically increase.

cn:copy_special_character_version

Specifies whether to report an error when there are invalid characters during data import and export using COPY FROM.

cn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

dn:gs_format_behavior_compat_options

Specifies the configuration items of GaussDB internal system functions.

cn:hll_duplicate_check

Specifies whether duplicate check is enabled by default.

cn:max_standby_streaming_delay

Specifies the wait period before queries on the standby node are canceled when the queries conflict with WAL data receiving through streaming replication in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

dn:b_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

dn:check_disconnect_query

Specifies whether to enable the function of terminating statement execution on the server after the client is disconnected due to timeout.

cn:gs_format_behavior_compat_options

Specifies the configuration items of GaussDB internal system functions.

cn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

cn:static_thread_pool_num

Specifies the number of threads used to create a static thread pool (static pool). This parameter takes effect only on CNs of a distributed instance after enable_thread_pool is enabled.

dn:max_standby_streaming_delay

Specifies the wait period before queries on the standby node are canceled when the queries conflict with WAL data receiving through streaming replication in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

audit_system_object

Specifies whether to audit the CREATE, DROP, and ALTER operations on GaussDB database objects. GaussDB database objects include databases, users, schemas, and tables. You can change the value of this parameter to audit only the operations on required database objects. In the scenario where the leader node is forcibly selected, you are advised to set audit_system_object to the maximum value and audit all DDL objects. For details about the value range, see the product documentation.

cn:enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

cn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

dn:page_version_partitions

Specifies the number of hash table partitions in cached page version information in the memory. This parameter directly affects the hash query efficiency and hash conflict probability, and is valid only when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:page_version_recycler_thread_num

Specifies the number of threads for recycling and verifying page version information. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:password_encryption_type

Specifies how user passwords are encrypted. 0: Passwords are encrypted using MD5. 1: Passwords are encrypted using SHA-256 and MD5. 2: Passwords are encrypted using SHA-256. 3: Passwords are encrypted using SM3. MD5 is not recommended because it is a weak encryption algorithm.

dn:dcf_thread_effective_time

Specifies the effective time of the DCF flushing thread. This parameter is used to determine whether the disk I/O hangs. If the DCF cannot access I/O resources within the period specified by this parameter, the DCF considers that the thread I/O hangs and a primary/standby switchover is triggered. If this parameter is set to 0, I/O hang detection is disabled. Unit: second

dn:a_format_date_timestamp

Specifies whether to enable current_date, current_timestamp, and localtimestamp to return the system time, instead of the transaction start time, when a transaction starts.

dn:hll_default_log2sparse

Specifies the default threshold for switching from Sparse mode to Full mode.

cn:enable_xid_abort_check

Specifies whether to check the status of transaction ID rollback when a transaction is committed.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

cn:hll_default_log2sparse

Specifies the default threshold for switching from Sparse mode to Full mode.

dn:disable_keyword_options

Specifies the configuration items for database compatibility. Multiple items are separated by commas (,). An identifier with this parameter set will not be used as a keyword.

cn:fix_func_selection

Specifies whether to optimize the function matching policy. The value catlist indicates the catlist sequence is optimized. (The non-B-compatible mode has been optimized. In non-B-compatible mode, system functions are always preferentially selected and executed. The policy in B-compatible mode is the same as that in versions earlier than 505.1.0. An error message indicating that the function is not unique may be displayed, or a system function may be selected for execution.)

cn:support_binary_copy_version

Specifies whether to verify the binary file encoding information when data is imported using COPY FROM in BINARY mode. If forward compatibility is required, leave this parameter empty. Otherwise, retain the default value.

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

password_lock_time

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

wdr_snapshot_interval

Specifies the interval at which the background thread Snapshot automatically takes snapshots of the database monitoring data. Unit: minute

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

dn:undo_retention_time

Specifies how long undo logs are kept. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained. Unit: second

cn:recovery_parse_workers

Specifies the number of ParseRedoRecord threads in the ultimate RTO feature. This parameter must be used together with recovery_redo_workers. If both recovery_parse_workers and recovery_redo_workers are greater than 1, ultimate RTO is enabled. If you do not want to enable ultimate RTO, retain the default value 1 for recovery_parse_workers. When enabling ultimate RTO, ensure that replication_type is set to 1. If both ultimate RTO and parallel replay are enabled, only ultimate RTO takes effect, and parallel replay is ineffective. Ultimate RTO does not support column-store tables. Therefore, disable ultimate RTO in a system where column-store tables are used or are to be used. Ultimate RTO also does not support flow control. Flow control is determined by the recovery_time_target parameter.

cn:undo_space_limit_size

Specifies the threshold for forcibly recycling undo space. When the undo space usage reaches 80% of the threshold, forcible recycling starts. It is recommended that the value of this parameter be greater than or equal to the value of undo_limit_size_per_transaction. Unit: 8 KB

cn:auto_increment_cache

Specifies the number of reserved auto-increment cache values when auto-increment is triggered by batch insertion or import of auto-increment columns. When auto-increment values are reserved, the auto-increment counter value is updated to the maximum auto-increment cache value. Before the cache values are used up, the auto-increment counter value remains unchanged, and the triggered auto-increment uses the cache values. If this parameter is set to 0, the auto-increment cache values are automatically reserved. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, 2^n values are reserved each time. For example, one value is reserved in the first auto-increment, two values are reserved in the second auto-increment, four values are reserved in the third auto-increment, and eight values are reserved for in fourth auto-increment. However, if the number of reserved values exceeds 65,535, 65,535 values are reserved. If this parameter is not set to 0, the number of reserved cache values is the value of this parameter. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, the value of auto_increment_cache is the number of auto-increment values reserved each time. The reserved cache values are valid only in the statement. If the reserved auto-increment cache values are used up and subsequent INSERT statements trigger auto-increment based on the auto-increment counter, the values in the auto-increment column in the table are discontinuous. This parameter does not affect the auto-increment column in the local temporary table.

dn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

cn:enable_early_free

Specifies whether the operator memory can be released in advance.

cn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed. The theoretical maximum value is 10000. Set this parameter based on workload requirements and hardware configurations. It is recommended that this parameter be set to a value less than or equal to 1/10 of max_connections.

cn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

cn:enable_enhance_toast_table

Specifies whether to use the enhanced TOAST out-of-line storage table. The value on indicates that the enhanced TOAST out-of-line storage table is used. The value off indicates that the TOAST out-of-line storage table is used.

cn:archive_interval

Specifies the archiving interval. Log files are forcibly archived when the period specified by this parameter has elapsed. A large value of this parameter affects the RPO of PITR. The default value is recommended. Unit: second

dn:undo_limit_size_per_transaction

Specifies the maximum undo space for a single transaction. If the undo space of a transaction exceeds this parameter value, the transaction is rolled back due to an error. It is recommended that the value of this parameter be smaller than the value of undo_space_limit_size. If this parameter value is larger, the value of undo_space_limit_size will be used as the maximum undo space for a single transaction. If this undo space is greater than 1 TB, system performance and stability may be affected. Unit: 8 KB

dn:convert_illegal_char_mode

Specifies the placeholders of invalid characters that can be displayed on the client when enable_convert_illegal_char is enabled. Value range: 95 characters whose decimal codes range from 32 to 126 in the ASCII coding table.

cn:enable_analyze_check

Specifies whether to check if statistics were collected about tables whose reltuples and relpages are displayed as 0 in pg_class during plan generation.

dn:tde_index_default_encrypt

When tde_index_default_encrypt is set to on and an index is created based on an encrypted table, the database automatically converts the index to an encrypted index.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

dn:enable_bbox_dump

Specifies whether to enable the black box function. Core files can be generated even when the core mechanism is not configured in the system.

dn:enable_recyclebin

Enables or disables the recycle bin in real time.

cn:wdr_snapshot_space_threshold

Specifies the threshold for controlling the space used by snapshots. When the space used by snapshots reaches 80% of the value of this parameter, the control logic of the database is enabled to stabilize the space usage. Unit: KB

enable_global_syscache

Determines whether to enable global system cache.

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

cn:max_compile_functions

Specifies the maximum number of function compilation results stored in the server. Excessive functions and compilation results of stored procedures may occupy large memory space. Setting this parameter to an appropriate value can reduce the memory usage and improve system performance. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

dn:auto_increment_offset

Specifies the initial value of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is greater than that of auto_increment_increment, there will be an error when the values in the auto-increment column automatically increase.

cn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a session. If the number of full or slow SQL statements exceeds this value, new statements will not be traced until the flush thread flushes the cached statements to the disk to reserve free space. The default value is recommended.

cn:page_version_partitions

Specifies the number of hash table partitions in cached page version information in the memory. This parameter directly affects the hash query efficiency and hash conflict probability, and is valid only when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. The value 0 indicates that there is no timeout limit. Unit: second

dn:index_txntype

Sets the index page format to PCR or RCR. This parameter is left unconfigured during system initialization. By default, the created indexes are compatible with the index type (RCR) of earlier versions. Once this parameter is specified, it cannot be left unconfigured again.

cn:enable_recyclebin

Enables or disables the recycle bin in real time.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

max_replication_slots

Specifies the number of log replication slots on the primary node.

cn:enable_bbox_dump

Specifies whether to enable the black box function so that core files can be generated even when the core mechanism is not configured in the system.

datestyle

Specifies the display format for date and time.

dn:hll_default_log2m

Specifies the number of buckets for HLL data.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots. Currently, only archive slots and backup slots are involved. Set this parameter to on only in cloud scenarios where logical replication is enabled. In other scenarios, set this parameter to off.

cn:enable_convert_illegal_char

Specifies whether the database supports characters not included the character sets.

cn:enable_workload_rule

Specifies whether to enable SQL throttling.

cn:index_txntype

Sets the index page format to PCR or RCR. This parameter is left unconfigured during system initialization. By default, the created indexes are compatible with the index type (RCR) of earlier versions. Once this parameter is specified, it cannot be left unconfigured again.

dn:enable_enhance_toast_table

Specifies whether to use the enhanced TOAST out-of-line storage table. The value on indicates that the enhanced TOAST out-of-line storage table is used. The value off indicates that the TOAST out-of-line storage table is used.

dn:support_binary_copy_version

Specifies whether to verify the binary file encoding information when data is imported using COPY FROM in BINARY mode. If forward compatibility is required, leave this parameter empty. Otherwise, retain the default value.

enable_stream_operator

Specifies the query optimizer's use of streams. If enable_stream_operator is disabled, a large number of logs indicating that the plans cannot be pushed down are recorded. If you do not need these logs, you are advised to disable both enable_unshipping_log and enable_stream_operator. The default value is recommended.

cn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

cn:gs_perf_retention_days

Specifies how many days the flame graph files in HTML format are retained. Unit: day

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

dn:enable_rls_match_index

Specifies whether indexes of a base table can be scanned based on target predicate conditions in row-level security scenarios. Target scenario: Row level security (RLS) policies are set and enabled in the base table, and the query predicate contains the unleakproof system function or like operator.

dn:audit_internal_event

Specifies whether to audit the connections and operations of internal tools cm_agent, gs_clean, and WDRXdb, and whether to audit the logins and logouts from CNs on DNs.

cn:effective_cache_size

Specifies the size of the disk buffer available to the CN optimizer in a single query. Unit: 8 KB

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

dn:hll_default_log2explicit

Specifies the threshold for switching from the explicit mode to the sparse mode.

cn:instr_unique_sql_combination_options

Specifies the configuration items of combining unique SQL statements of the same type.

dn:undo_space_limit_size

Specifies the threshold for forcibly recycling undo space. When the undo space usage reaches 80% of the threshold, forcible recycling starts. It is recommended that the value of this parameter be greater than or equal to the value of undo_limit_size_per_transaction. Unit: 8 KB

dn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

dn:llvm_max_memory

Specifies the upper limit of the memory used by IRs (including cached and in-use IRs) generated during compilation in Codegen execution mode. The memory used by Codegen is not applied for by preoccupation. It is a part of max_dynamic_memory and is restricted by the llvm_max_memory parameter. Unit: KB

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

cn:tde_index_default_encrypt

When tde_index_default_encrypt is set to on and an index is created based on an encrypted table, the database automatically converts the index to an encrypted index.

dn:enable_workload_rule

Specifies whether to enable SQL throttling.

dn:archive_interval

Specifies the archiving interval. Log files are forcibly archived when the period specified by this parameter has elapsed. A large value of this parameter affects the RPO of PITR. The default value is recommended. Unit: second

cn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

cn:enable_codegen

Specifies whether code optimization can be enabled. Currently, code optimization uses the LLVM optimization.

max_wal_senders

Specifies the maximum number of WAL sender threads that can be created. Processes created for the following purposes occupy WAL sender threads: Standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. If this parameter is set to a value smaller than 20, scale-out may fail. The value of this parameter must be smaller than that of max_connections.

cn:undo_limit_size_per_transaction

Specifies the maximum undo space for a single transaction. If the undo space of a transaction exceeds this parameter value, the transaction is rolled back due to an error. It is recommended that the value of this parameter be smaller than the value of undo_space_limit_size. If this parameter value is larger, the value of undo_space_limit_size will be used as the maximum undo space for a single transaction. If this undo space is greater than 1 TB, system performance and stability may be affected. Unit: 8 KB

dn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed. The theoretical maximum value is 10000. Set this parameter based on workload requirements and hardware configurations. It is recommended that this parameter be set to a value less than or equal to 1/10 of max_connections.

cn:page_missing_dirty_check

Specifies whether to enable the verification for pages not marked as dirty. The verification checks whether the modified pages are not marked as dirty. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

cn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

cn:audit_dml_state_select

Determines whether to audit the SELECT operation.

cn:hll_default_log2m

Specifies the number of buckets for HLL data.

dn:use_workload_manager

Specifies whether to enable resource management. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:use_workload_manager

Specifies whether to enable resource management. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_save_datachanged_timestamp

Specifies whether to record the time when INSERT, UPDATE, DELETE, or EXCHANGE/TRUNCATE/DROP PARTITION is performed on table data. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_save_datachanged_timestamp

Specifies whether to record the time when INSERT, UPDATE, DELETE, or EXCHANGE/TRUNCATE/DROP PARTITION is performed on table data. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_process_memory

Specifies the maximum physical memory allowed for a DN. Unit: KB This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_process_memory

Specifies the maximum physical memory allowed for a CN. Unit: KB This parameter is supported in GaussDB V2.0-8.201 or later.

dn:standby_shared_buffers_fraction

Specifies the shared_buffers proportion used on the server where a standby instance is deployed. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:standby_shared_buffers_fraction

Specifies the shared_buffers proportion used on the server where a standby instance is deployed. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_sleep

Specifies the interval (in milliseconds) for the page writer thread to flush dirty pages to disks after the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:pagewriter_sleep

Specifies the interval (in milliseconds) for the page writer thread to flush dirty pages to disks after the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:xloginsert_locks

Specifies the number of locks on concurrent write-ahead logging. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:xloginsert_locks

Specifies the number of locks on concurrent write-ahead logging. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_light_proxy

Specifies whether the optimizer optimizes the execution of simple queries on CNs. This parameter does not take effect if the character set of the application side does not match that of the kernel side. You are advised to set the character set to UTF-8 when creating a database. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_light_proxy

Specifies whether the optimizer optimizes the execution of simple queries on CNs. This parameter does not take effect if the character set of the application side does not match that of the kernel side. You are advised to set the character set to UTF-8 when creating a database. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:synchronous_commit

Specifies the synchronization mode of the current transaction. Generally, logs generated by a transaction are synchronized in the following sequence: The primary node writes the logs to the local memory. The primary node writes logs in the local memory to the local file system. The primary node flushes the logs in the local file system to disks. The primary node sends the logs to the standby node. The standby node receives the logs and saves them to its local memory. The standby node writes the logs in the local memory to the local file system. The standby node flushes the logs in the local file system to disks. The standby node replays the logs to complete the incremental update of data files. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:synchronous_commit

Specifies the synchronization mode of the current transaction. Generally, logs generated by a transaction are synchronized in the following sequence: The primary node writes the logs to the local memory. The primary node writes logs in the local memory to the local file system. The primary node flushes the logs in the local file system to disks. The primary node sends the logs to the standby node. The standby node receives the logs and saves them to its local memory. The standby node writes the logs in the local memory to the local file system. The standby node flushes the logs in the local file system to disks. The standby node replays the logs to complete the incremental update of data files. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_incremental_checkpoint

Specifies whether to enable incremental checkpoint. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_incremental_checkpoint

Specifies whether to enable incremental checkpoint. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bypass_workload_manager

Enables or disables independent I/O control. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bypass_workload_manager

Enables or disables independent I/O control. This parameter is supported in GaussDB V2.0-8.201 or later.

enable_resource_track

Specifies whether to enable real-time resource monitoring. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_rt_percentile

Specifies whether to enable the function of calculating the response time of 80% and 95% of the SQL statements in the system. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_rt_percentile

Specifies whether to enable the function of calculating the response time of 80% and 95% of the SQL statements in the system. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_cpu_timer

Specifies whether to capture the CPU time consumed during SQL statement execution. In the x86-based centralized deployment scenario where 32 vCPUs and 256 GB of memory are configured, the performance tested by BenchmarkSQL 5.0 fluctuates by about 3.5% by enabling or disabling this parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_cpu_timer

Specifies whether to capture the CPU time consumed during SQL statement execution. In the x86-based centralized deployment scenario where 32 vCPUs and 256 GB of memory are configured, the performance tested by BenchmarkSQL 5.0 fluctuates by about 3.5% by enabling or disabling this parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

track_counts

Determines whether to enable collection of statistics on database activities. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_user_metric_persistent

Specifies whether to dump the historical monitoring data of user resources. If this parameter is set to on, data in the PG_TOTAL_USER_RESOURCE_INFO view is periodically sampled and saved to the system catalog GS_WLM_USER_RESOURCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_user_metric_persistent

Specifies whether to dump the historical monitoring data of user resources. If this parameter is set to on, data in the PG_TOTAL_USER_RESOURCE_INFO view is periodically sampled and saved to the system catalog GS_WLM_USER_RESOURCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:log_hostname

By default, connection log messages only show the IP address of the connecting host. The host name can be recorded when this parameter is set to on. It may take some time to parse the host name. Therefore, the database performance may be affected. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:log_hostname

By default, connection log messages only show the IP address of the connecting host. The host name can be recorded when this parameter is set to on. It may take some time to parse the host name. Therefore, the database performance may be affected. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_track_wait

Specifies whether to enable real-time collection of wait event information. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_track_wait

Specifies whether to enable real-time collection of wait event information. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instance_metric_persistent

Specifies whether to dump instance resource monitoring data. When this parameter is set to on, the instance monitoring data is saved to the system catalog GS_WLM_INSTANCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instance_metric_persistent

Specifies whether to dump instance resource monitoring data. When this parameter is set to on, the instance monitoring data is saved to the system catalog GS_WLM_INSTANCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_logical_io_statistics

Specifies whether to collect logical I/O statistics during resource monitoring. If this function is enabled, the read_kbytes, write_kbytes, read_counts, write_counts, read_speed, and write_speed columns in the PG_TOTAL_USER_RESOURCE_INFO view will collect statistics on the byte count, number of times, and speed of logical read and write. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_logical_io_statistics

Specifies whether to collect logical I/O statistics during resource monitoring. If this function is enabled, the read_kbytes, write_kbytes, read_counts, write_counts, read_speed, and write_speed columns in the PG_TOTAL_USER_RESOURCE_INFO view will collect statistics on the byte count, number of times, and speed of logical read and write. Columns related to logical read and write in the system catalogs GS_WLM_USER_RESOURCE_HISTORY and GS_WLM_INSTANCE_HISTORY will collect statistics on the logical read and write of related users and instances. This parameter is supported in GaussDB V2.0-8.201 or later.

enable_sort

Specifies whether the query optimizer uses sort methods. It is impossible to suppress explicit sorts entirely, but setting this variable to off encourages the optimizer to choose other methods if available. This parameter is supported in GaussDB V2.0-8.201 or later.

enable_material

Specifies whether the query optimizer uses materialization. It is impossible to suppress materialization entirely, but setting this variable to off prevents the optimizer from inserting materialized nodes. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:gtm_max_trans

Specifies the maximum number of connections accepted by the GTM. You are advised not to change the value. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:csn_sync_interval

Specifies the interval for synchronizing CSNs between the primary and standby GTMs, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:restore_duration

Specifies the reply interval of XIDs or CSNs on the GTM, that is, the number of XIDs or CSNs for reply. This parameter is supported in GaussDB V2.0-8.201 or later.

cms:coordinator_heartbeat_timeout

Specifies the heartbeat timeout (in seconds) that triggers the automatic removal of faulty CNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:client_min_messages

Specifies which level of messages will be sent to the client. A higher level covers the messages of all the lower levels. The lower the level is, the fewer messages are sent. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:client_min_messages

Specifies which level of messages will be sent to the client. A higher level covers the messages of all the lower levels. The lower the level is, the fewer messages are sent. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_thread_num

Specifies the number of threads for background page flushing after the incremental checkpoint is enabled. Dirty pages are flushed in sequence to disks, updating recovery points. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:pagewriter_thread_num

Specifies the number of threads for background page flushing after the incremental checkpoint is enabled. Dirty pages are flushed in sequence to disks, updating recovery points. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bgwriter_delay

Specifies the interval at which the background writer writes dirty shared buffers, in milliseconds. Each time, the background writer initiates write operations for some dirty buffers. In full checkpoint mode, the bgwriter_lru_maxpages parameter is used to control the amount of data to be written each time, and the process is restarted after the period of hibernation specified by bgwriter_delay (in milliseconds). In incremental checkpoint mode, the number of target idle buffer pages is calculated based on the value of candidate_buf_percent_target. If the number of idle buffer pages is insufficient, a batch of pages are flushed to disks at the interval specified by bgwriter_delay (in milliseconds). The number of flushed pages is calculated based on the target difference percentage. The maximum number of flushed pages is limited by max_io_capacity. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bgwriter_delay

Specifies the interval at which the background writer writes dirty shared buffers, in milliseconds. Each time, the background writer initiates write operations for some dirty buffers. In full checkpoint mode, the bgwriter_lru_maxpages parameter is used to control the amount of data to be written each time, and the process is restarted after the period of hibernation specified by bgwriter_delay (in milliseconds). In incremental checkpoint mode, the number of target idle buffer pages is calculated based on the value of candidate_buf_percent_target. If the number of idle buffer pages is insufficient, a batch of pages are flushed to disks at the interval specified by bgwriter_delay (in milliseconds). The number of flushed pages is calculated based on the target difference percentage. The maximum number of flushed pages is limited by max_io_capacity. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_files_per_process

Specifies the maximum number of files that can be opened simultaneously by each server process. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_files_per_process

Specifies the maximum number of files that can be opened simultaneously by each server process. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:candidate_buf_percent_target

Specifies the proportion (%) of available buffers in shared_buffer in the candidate buffer chain when the incremental checkpoint is enabled. If the percentage is less than the value of this parameter, the bgwriter thread starts to flush dirty pages based on the max_io_capacity parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:candidate_buf_percent_target

Specifies the proportion (%) of available buffers in shared_buffer in the candidate buffer chain when the incremental checkpoint is enabled. If the percentage is less than the value of this parameter, the bgwriter thread starts to flush dirty pages based on the max_io_capacity parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:checkpoint_timeout

Specifies the maximum time between automatic WAL checkpoints, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:checkpoint_timeout

Specifies the maximum time between automatic WAL checkpoints, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:incremental_checkpoint_timeout

Specifies the maximum interval (in seconds) between automatic WAL checkpoints when the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:incremental_checkpoint_timeout

Specifies the maximum interval (in seconds) between automatic WAL checkpoints when the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_writer_delay

Specifies the delay (in milliseconds) between activity rounds for the WAL writer. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:wal_writer_delay

Specifies the delay (in milliseconds) between activity rounds for the WAL writer. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_buffers

Specifies the shared memory for storing WALs, that is, multiples of XLOG_BLCKSZ, in increments of 8 KB. For details about the value range, see wal_buffers. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:wal_buffers

Specifies the shared memory for storing WALs, that is, multiples of XLOG_BLCKSZ, in increments of 8 KB. For details about the value range, see wal_buffers. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:walwriter_cpu_bind

Specifies the CPU core bound to the WAL write thread. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bulk_write_ring_size

Specifies the size of the ring buffer used by the operation when a large amount of data is written (for example, the copy operation). This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bulk_write_ring_size

Specifies the size of a ring buffer used for parallel data import. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:recovery_redo_workers

Specifies the number of PageRedoWorker threads corresponding to each ParseRedoRecord thread in when ultimate RTO is enabled. recovery_redo_workers must be used together with recovery_parse_workers. The value of recovery_redo_workers takes effect only when recovery_parse_workers is greater than 1. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:recovery_redo_workers

Specifies the number of PageRedoWorker threads corresponding to each ParseRedoRecord thread in when ultimate RTO is enabled. recovery_redo_workers must be used together with recovery_parse_workers. The value of recovery_redo_workers takes effect only when recovery_parse_workers is greater than 1. This parameter is supported in GaussDB V2.0-8.201 or later.

cms:phony_dead_effective_time

Specifies the maximum number of times DN processes are detected as zombie. If the number of times a process is detected as zombie is greater than the specified value, the process is considered to be a zombie process and will be restarted. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_ai_watchdog

Enables or disables the AI watchdog function. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_ai_watchdog

Enables or disables the AI watchdog function. This parameter is supported in GaussDB V2.0-8.201 or later.

enable_unshipping_log

Specifies whether to log statements that are not pushed down. The logs help locate performance issues that may be caused by statements not pushed down. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_connections

Specifies the maximum number of concurrent connections. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_connections

Specifies the maximum number of concurrent connections. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_memory_reject_percent

Specifies the dynamic memory usage for escape from memory overload. This parameter takes effect only when the GUC parameters use_workload_manager and enable_memory_limit are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:resilience_memory_reject_percent

Specifies the dynamic memory usage for escape from memory overload. This parameter is only applied when use_workload_manager and enable_memory_limit are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_threadpool_reject_cond

Specifies the proportion of accumulated sessions in the thread pool for escape from overload. This parameter is only applied when enable_thread_pool and use_workload_manager are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:resilience_threadpool_reject_cond

Specifies the proportion of accumulated sessions in the thread pool for escape from overload. This parameter is only applied when enable_thread_pool and use_workload_manager are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:vacuum_cost_delay

Specifies how long the process sleeps when vacuum_cost_limit has been exceeded, in milliseconds. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:vacuum_cost_delay

Specifies how long the process sleeps when vacuum_cost_limit has been exceeded, in milliseconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:b_format_version

Specifies the SQL behavior control configuration item in B-compatible mode. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:b_format_version

Specifies the forward compatibility in MySQL-compatible mode. This is a configuration item for controlling database platform compatibility behavior. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_control_group

Specifies whether to enable the Cgroups. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_control_group

Specifies whether to enable the Cgroups. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_stmt_parameter

After track_stmt_parameter is enabled, the executed statements recorded in statement_history are not normalized. The complete SQL statement information can be displayed to help the database administrator locate faults. For a simple query, the complete statement information is displayed. For a PBE statement, the complete statement information and information about each variable value are displayed. The format is "query string; parameters:$1=value1,$2=value2, ...". This parameter is used to display full SQL information for users and is not controlled by the track_activity_query_size parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:track_stmt_parameter

After track_stmt_parameter is enabled, the executed statements recorded in statement_history are not normalized. The complete SQL statement information can be displayed to help the database administrator locate faults. For a simple query, the complete statement information is displayed. For a PBE statement, the complete statement information and information about each variable value are displayed. The format is "query string; parameters:$1=value1,$2=value2, ...". This parameter is used to display full SQL information for users and is not controlled by the track_activity_query_size parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_activity_query_size

Specifies byte counts of the current running commands used to trace each active session. If the number of bytes in a command is greater than the value of this parameter, the command will be truncated. This parameter affects the memory usage of multiple system views. Adjusting the value of this parameter has a great impact on the memory usage. You can estimate the memory usage based on the value calculated from instr_unique_sql_count times track_activity_query_size. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:track_activity_query_size

Specifies byte counts of the current running commands used to trace each active session. If the number of bytes in a command is greater than the value of this parameter, the command will be truncated. This parameter is supported in GaussDB V2.0-8.201 or later.

recyclebin_retention_time

Specifies how long files will be kept in the recycle bin. Files in the recycle bin will be automatically deleted after this length of time. Unit: second

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. The value 0 indicates that there is no timeout limit. Unit: second

cn:page_version_partitions

Specifies the number of hash table partitions in cached page version information in the memory. This parameter directly affects the hash query efficiency and hash conflict probability, and is valid only when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:index_txntype

Sets the index page format to PCR or RCR. This parameter is left unconfigured during system initialization. By default, the created indexes are compatible with the index type (RCR) of earlier versions. Once this parameter is specified, it cannot be left unconfigured again.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

datestyle

Specifies the display format for date and time.

max_replication_slots

Specifies the number of log replication slots on the primary node.

timezone

Specifies the time zone for displaying and interpreting time stamps.

cn:auto_increment_offset

Specifies the initial value of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is greater than that of auto_increment_increment, there will be an error when the values in the auto-increment column automatically increase.

dn:enable_xid_abort_check

Specifies whether to check the status of transaction ID rollback when a transaction is committed.

cn:audit_internal_event

Specifies whether to audit the connections and operations of internal tools cm_agent, gs_clean, and WDRXdb, and whether to audit the logins and logouts from CNs on DNs.

enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots. Currently, only archive slots and backup slots are involved. Set this parameter to on only in cloud scenarios where logical replication is enabled. In other scenarios, set this parameter to off.

cn:enable_convert_illegal_char

Specifies whether the database supports characters not included the character sets.

cn:codegen_compile_thread_num

Specifies the number of Codegen compilation threads.

cn:enable_workload_rule

Specifies whether to enable SQL throttling.

dn:enable_enhance_toast_table

Specifies whether to use the enhanced TOAST out-of-line storage table. The value on indicates that the enhanced TOAST out-of-line storage table is used. The value off indicates that the TOAST out-of-line storage table is used.

dn:static_thread_pool_num

Specifies the number of threads used to create a static thread pool (static pool). This parameter takes effect only on CNs of a distributed instance after enable_thread_pool is enabled.

enable_stream_operator

Specifies the query optimizer's use of streams. If enable_stream_operator is disabled, a large number of logs indicating that the plans cannot be pushed down are recorded. If you do not need these logs, you are advised to disable both enable_unshipping_log and enable_stream_operator. The default value is recommended.

cn:auto_increment_increment

Specifies the auto-increment step of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is smaller than that of auto_increment_offset, there will be an error when the values in the auto-increment column automatically increase.

dn:support_binary_copy_version

Specifies whether to verify the binary file encoding information when data is imported using COPY FROM in BINARY mode. If forward compatibility is required, leave this parameter empty. Otherwise, retain the default value.

cn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

cn:gs_perf_retention_days

Specifies how many days the flame graph files in HTML format are retained. Unit: day

dn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

cn:sql_mode

Specifies the SQL behavior control configuration item in M-compatible mode.

dn:enable_rls_match_index

Specifies whether indexes of a base table can be scanned based on target predicate conditions in row-level security scenarios. Target scenario: Row level security (RLS) policies are set and enabled in the base table, and the query predicate contains the unleakproof system function or like operator.

dn:audit_internal_event

Specifies whether to audit the connections and operations of internal tools cm_agent, gs_clean, and WDRXdb, and whether to audit the logins and logouts from CNs on DNs.

cn:llvm_max_memory

Specifies the upper limit of the memory used by IRs (including cached and in-use IRs) generated during compilation in Codegen execution mode. The memory used by Codegen is not applied for by preoccupation. It is a part of max_dynamic_memory and is restricted by the llvm_max_memory parameter. Unit: KB

dn:fix_func_selection

Specifies whether to optimize the function matching policy. The value catlist indicates the catlist sequence is optimized. (The non-B-compatible mode has been optimized. In non-B-compatible mode, system functions are always preferentially selected and executed. The policy in B-compatible mode is the same as that in versions earlier than 505.1.0. An error message indicating that the function is not unique may be displayed, or a system function may be selected for execution.)

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

comm_no_delay

Specifies whether to use the no_delay attribute of a communication library connection.

enable_recyclebin

Enables or disables the recycle bin in real time.

dn:sql_mode

Specifies the SQL behavior control configuration item in M-compatible mode.

cn:instr_unique_sql_combination_options

Specifies the configuration items of combining unique SQL statements of the same type.

dn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

dn:verify_log_buffers

Specifies the size or pages of verifyLog buffers in memory mode. The unit is 8 KB. For example, if the value of this parameter is 4, the requested memory is 4 x 8 KB = 32 KB. This parameter is valid only when page_version_check is set to persistence. If page_version_check is set to another value, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to persistence.

cn:convert_illegal_char_mode

Specifies the placeholders of invalid characters that can be displayed on the client when enable_convert_illegal_char is enabled. Value range: 95 characters whose decimal codes range from 32 to 126 in the ASCII coding table.

dn:llvm_max_memory

Specifies the upper limit of the memory used by IRs (including cached and in-use IRs) generated during compilation in Codegen execution mode. The memory used by Codegen is not applied for by preoccupation. It is a part of max_dynamic_memory and is restricted by the llvm_max_memory parameter. Unit: KB

cn:page_version_check

Specifies the type of page version verification. off indicates that page version verification is disabled. memory indicates that page version verification in pure memory mode is enabled. The page version information will be lost after a restart. persistence indicates that persistent page version verification is enabled. The page version information will not be lost after a restart.

undo_space_limit_size

Specifies the threshold for forcibly recycling undo space. When the undo space usage reaches 80% of the threshold, forcible recycling starts. It is recommended that the value of this parameter be greater than or equal to the value of undo_limit_size_per_transaction. Unit: 8 KB

cn:tde_index_default_encrypt

When tde_index_default_encrypt is set to on and an index is created based on an encrypted table, the database automatically converts the index to an encrypted index.

recovery_time_target

Specifies the time for the standby node to write and replay logs. Unit: second

dn:enable_convert_illegal_char

Specifies whether the database supports characters not included the character sets.

dn:system_view_version

Determines the version of the system view. All versions are backward compatible. For example, when system_view_version is set to 3, all features of version 2 and version 1 are also supported. For details, see the product documentation.

dn:enable_workload_rule

Specifies whether to enable SQL throttling.

dn:archive_interval

Specifies the archiving interval. Log files are forcibly archived when the period specified by this parameter has elapsed. A large value of this parameter affects the RPO of PITR. The default value is recommended. Unit: second

cn:enable_codegen

Specifies whether code optimization can be enabled. Currently, code optimization uses the LLVM optimization.

dn:codegen_compile_thread_num

Specifies the number of Codegen compilation threads.

max_wal_senders

Specifies the maximum number of WAL sender threads that can be created. Processes created for the following purposes occupy WAL sender threads: Standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. If this parameter is set to a value smaller than 20, scale-out may fail. The value of this parameter must be smaller than that of max_connections.

dn:resilience_ctrlstmt_detect_timelimit

Specifies the execution time of a normal SQL statement that will be marked as a slow SQL statement. The value 0 indicates that slow SQL statements are not identified. A value greater than 0 indicates that a normal SQL statement whose execution time exceeds the value of this parameter is marked as a slow SQL statement. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. Unit: millisecond

cn:a_format_date_timestamp

Specifies whether to enable current_date, current_timestamp, and localtimestamp to return the system time, instead of the transaction start time, when a transaction starts.

cn:page_missing_dirty_check

Specifies whether to enable the verification for pages not marked as dirty. The verification checks whether the modified pages are not marked as dirty. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

cn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

dn:max_compile_functions

Specifies the maximum number of function compilation results stored in the server. Excessive functions and compilation results of stored procedures may occupy large memory space. Setting this parameter to an appropriate value can reduce the memory usage and improve system performance. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

dn:tde_index_default_encrypt

When tde_index_default_encrypt is set to on and an index is created based on an encrypted table, the database automatically converts the index to an encrypted index.

dn:recovery_parse_workers

Specifies the number of ParseRedoRecord threads in the ultimate RTO feature. This parameter must be used together with recovery_redo_workers. If both recovery_parse_workers and recovery_redo_workers are greater than 1, ultimate RTO is enabled. If you do not want to enable ultimate RTO, retain the default value 1 for recovery_parse_workers. When enabling ultimate RTO, ensure that replication_type is set to 1. If both ultimate RTO and parallel replay are enabled, only ultimate RTO takes effect, and parallel replay is ineffective. Ultimate RTO does not support column-store tables. Therefore, disable ultimate RTO in a system where column-store tables are used or are to be used. Ultimate RTO also does not support flow control. Flow control is determined by the recovery_time_target parameter.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

undo_retention_time

Specifies how long undo logs are kept. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained. Unit: second

cn:wdr_snapshot_space_threshold

Specifies the threshold for controlling the space used by snapshots. When the space used by snapshots reaches 80% of the value of this parameter, the control logic of the database is enabled to stabilize the space usage. Unit: KB

dn:page_version_check

Specifies the type of page version verification. off indicates that page version verification is disabled. memory indicates that page version verification in pure memory mode is enabled. The page version information will be lost after a restart. persistence indicates that persistent page version verification is enabled. The page version information will not be lost after a restart.

cn:max_compile_functions

Specifies the maximum number of function compilation results stored in the server. Excessive functions and compilation results of stored procedures may occupy large memory space. Setting this parameter to an appropriate value can reduce the memory usage and improve system performance. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

cn:page_version_max_num

Specifies the maximum number of page versions that can be cached in memory. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:auto_increment_offset

Specifies the initial value of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is greater than that of auto_increment_increment, there will be an error when the values in the auto-increment column automatically increase.

dn:auto_increment_cache

Specifies the number of reserved auto-increment cache values when auto-increment is triggered by batch insertion or import of auto-increment columns. When auto-increment values are reserved, the auto-increment counter value is updated to the maximum auto-increment cache value. Before the cache values are used up, the auto-increment counter value remains unchanged, and the triggered auto-increment uses the cache values. If this parameter is set to 0, the auto-increment cache values are automatically reserved. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, 2^n values are reserved each time. For example, one value is reserved in the first auto-increment, two values are reserved in the second auto-increment, four values are reserved in the third auto-increment, and eight values are reserved for in fourth auto-increment. However, if the number of reserved values exceeds 65,535, 65,535 values are reserved. If this parameter is not set to 0, the number of reserved cache values is the value of this parameter. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, the value of auto_increment_cache is the number of auto-increment values reserved each time. The reserved cache values are valid only in the statement. If the reserved auto-increment cache values are used up and subsequent INSERT statements trigger auto-increment based on the auto-increment counter, the values in the auto-increment column in the table are discontinuous. This parameter does not affect the auto-increment column in the local temporary table.

dn:enable_codegen

Specifies whether code optimization can be enabled. Currently, code optimization uses the LLVM optimization.

dn:instr_unique_sql_combination_options

Specifies the configuration items of combining unique SQL statements of the same type. If this feature is enabled, the IDs of unique SQL statements of the same type are normalized, and the generated unique SQL strings are normalized.

rewrite_rule

Sets query rewriting rules.

dn:page_missing_dirty_check

Specifies whether to enable the verification for pages not marked as dirty. The verification checks whether the modified pages are not marked as dirty. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:wdr_snapshot_full_backup_interval

Specifies the interval at which a full WDR snapshot is created. The interval specified by this parameter is about a number instead of time. For example, if the parameter is set to 12, a full snapshot and then 11 incremental snapshots are generated for each group. If the parameter is set to 1, all snapshots generated are full snapshots.

dn:wdr_snapshot_space_threshold

Specifies the threshold for controlling the space used by snapshots. When the space used by snapshots reaches 80% of the value of this parameter, the control logic of the database is enabled to stabilize the space usage. Unit: KB

cn:gs_perf_interval

Specifies the automatic perf data collection interval. The value 0 indicates that the collection is stopped. If the value is greater than 0 and less than 5, the value 5 is used. Unit: minute

dn:gs_perf_interval

Specifies the automatic perf data collection interval. The value 0 indicates that the collection is stopped. If the value is greater than 0 and less than 5, the value 5 is used. Unit: minute

dn:extra_float_digits

Adjusts the number of digits displayed for floating-point values, including float4, float8, and geometric data types. The parameter value is added to the standard number of digits (FLT_DIG or DBL_DIG as appropriate). This parameter can also be set to a negative value to suppress unwanted digits.

cn:disable_keyword_options

Specifies the configuration items for database compatibility. Multiple items are separated by commas (,). An identifier with this parameter set will not be used as a keyword.

track_stmt_stat_level

Controls the level of statement execution tracking.

cn:b_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

dn:copy_special_character_version

Specifies whether to report an error when there are invalid characters during data import and export using COPY FROM.

cn:page_version_recycler_thread_num

Specifies the number of threads for recycling and verifying page version information. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

cn:wdr_snapshot_full_backup_interval

Specifies the interval at which a full WDR snapshot is created. The interval specified by this parameter is about a number instead of time. For example, if the parameter is set to 12, a full snapshot and then 11 incremental snapshots are generated for each group. If the parameter is set to 1, all snapshots generated are full snapshots.

cn:extra_float_digits

Adjusts the number of digits displayed for floating-point values, including float4, float8, and geometric data types. The parameter value is added to the standard number of digits (FLT_DIG or DBL_DIG as appropriate). This parameter can also be set to a negative value to suppress unwanted digits.

cn:system_view_version

Determines the version of the system view. All versions are backward compatible. For example, when system_view_version is set to 3, all features of version 2 and version 1 are also supported. For details, see the product documentation.

support_batch_bind

Specifies whether to bind and execute PBE (Parse, Bind, Execute) statements in batches through interfaces such as JDBC, ODBC, and libpq.

cn:check_disconnect_query

Specifies whether to enable the function of terminating statement execution on the server after the client is disconnected due to timeout.

password_effect_time

Specifies the validity period of an account password. Unit: day

cn:verify_log_buffers

Specifies the size or pages of verifyLog buffers in memory mode. The unit is 8 KB. For example, if the value of this parameter is 4, the requested memory is 4 x 8 KB = 32 KB. This parameter is valid only when page_version_check is set to persistence. If page_version_check is set to another value, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to persistence.

cn:enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots. Currently, only archive slots and backup slots are involved. Set this parameter to on only in cloud scenarios where logical replication is enabled. In other scenarios, set this parameter to off.

undo_limit_size_per_transaction

Specifies the threshold for forcibly recycling undo space. When the undo space usage reaches 80% of the threshold, forcible recycling starts. It is recommended that the value of this parameter be greater than or equal to the value of undo_limit_size_per_transaction. Unit: 8 KB

dn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

cn:enable_rls_match_index

Specifies whether indexes of a base table can be scanned based on target predicate conditions in row-level security scenarios. Target scenario: Row level security (RLS) policies are set and enabled in the base table, and the query predicate contains the unleakproof system function or like operator.

dn:auto_increment_increment

Specifies the auto-increment step of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is smaller than that of auto_increment_offset, there will be an error when the values in the auto-increment column automatically increase.

cn:copy_special_character_version

Specifies whether to report an error when there are invalid characters during data import and export using COPY FROM.

behavior_compat_options

Specifies database compatibility configuration items. After the value of proc_outparam_override is changed, the database must be connected again or the instance must be rebooted. Otherwise, stored procedures and functions cannot be correctly called.

dn:gs_format_behavior_compat_options

Specifies the configuration items of GaussDB internal system functions.

dn:b_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

dn:check_disconnect_query

Specifies whether to enable the function of terminating statement execution on the server after the client is disconnected due to timeout.

cn:gs_format_behavior_compat_options

Specifies the configuration items of GaussDB internal system functions.

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

cn:static_thread_pool_num

Specifies the number of threads used to create a static thread pool (static pool). This parameter takes effect only on CNs of a distributed instance after enable_thread_pool is enabled.

audit_system_object

Specifies whether to audit the CREATE, DROP, and ALTER operations on GaussDB database objects. GaussDB database objects include databases, users, schemas, and tables. You can change the value of this parameter to audit only the operations on required database objects. In the scenario where the leader node is forcibly selected, you are advised to set audit_system_object to the maximum value and audit all DDL objects. For details about the value range, see the product documentation.

cn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

dn:enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots. Currently, only archive slots and backup slots are involved. Set this parameter to on only in cloud scenarios where logical replication is enabled. In other scenarios, set this parameter to off.

dn:page_version_partitions

Specifies the number of hash table partitions in cached page version information in the memory. This parameter directly affects the hash query efficiency and hash conflict probability, and is valid only when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:page_version_recycler_thread_num

Specifies the number of threads for recycling and verifying page version information. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:dcf_thread_effective_time

Specifies the effective time of the DCF flushing thread. This parameter is used to determine whether the disk I/O hangs. If the DCF cannot access I/O resources within the period specified by this parameter, the DCF considers that the thread I/O hangs and a primary/standby switchover is triggered. If this parameter is set to 0, I/O hang detection is disabled. Unit: second

dn:a_format_date_timestamp

Specifies whether to enable current_date, current_timestamp, and localtimestamp to return the system time, instead of the transaction start time, when a transaction starts.

cn:crypto_module_info

This is a high-risk parameter. If this parameter is incorrectly set for an instance, the verification will fail when the instance is rebooted. Before modifying this parameter, fully verify the change in a test environment to avoid unintended consequences. This parameter specifies the prerequisites for using third_kms in TDE. It is used to enable the third-party encryption library and configure parameters for using the library. For details about how to set this parameter, see the relevant section in the product documentation.

cn:enable_xid_abort_check

Specifies whether to check the status of transaction ID rollback when a transaction is committed.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

dn:disable_keyword_options

Specifies the configuration items for database compatibility. Multiple items are separated by commas (,). An identifier with this parameter set will not be used as a keyword.

cn:fix_func_selection

Specifies whether to optimize the function matching policy. The value catlist indicates the catlist sequence is optimized. (The non-B-compatible mode has been optimized. In non-B-compatible mode, system functions are always preferentially selected and executed. The policy in B-compatible mode is the same as that in versions earlier than 505.1.0. An error message indicating that the function is not unique may be displayed, or a system function may be selected for execution.)

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

password_lock_time

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

cn:support_binary_copy_version

Specifies whether to verify the binary file encoding information when data is imported using COPY FROM in BINARY mode. If forward compatibility is required, leave this parameter empty. Otherwise, retain the default value.

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

cn:recovery_parse_workers

Specifies the number of ParseRedoRecord threads in the ultimate RTO feature. This parameter must be used together with recovery_redo_workers. If both recovery_parse_workers and recovery_redo_workers are greater than 1, ultimate RTO is enabled. If you do not want to enable ultimate RTO, retain the default value 1 for recovery_parse_workers. When enabling ultimate RTO, ensure that replication_type is set to 1. If both ultimate RTO and parallel replay are enabled, only ultimate RTO takes effect, and parallel replay is ineffective. Ultimate RTO does not support column-store tables. Therefore, disable ultimate RTO in a system where column-store tables are used or are to be used. Ultimate RTO also does not support flow control. Flow control is determined by the recovery_time_target parameter.

cn:auto_increment_cache

Specifies the number of reserved auto-increment cache values when auto-increment is triggered by batch insertion or import of auto-increment columns. When auto-increment values are reserved, the auto-increment counter value is updated to the maximum auto-increment cache value. Before the cache values are used up, the auto-increment counter value remains unchanged, and the triggered auto-increment uses the cache values. If this parameter is set to 0, the auto-increment cache values are automatically reserved. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, 2^n values are reserved each time. For example, one value is reserved in the first auto-increment, two values are reserved in the second auto-increment, four values are reserved in the third auto-increment, and eight values are reserved for in fourth auto-increment. However, if the number of reserved values exceeds 65,535, 65,535 values are reserved. If this parameter is not set to 0, the number of reserved cache values is the value of this parameter. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, the value of auto_increment_cache is the number of auto-increment values reserved each time. The reserved cache values are valid only in the statement. If the reserved auto-increment cache values are used up and subsequent INSERT statements trigger auto-increment based on the auto-increment counter, the values in the auto-increment column in the table are discontinuous. This parameter does not affect the auto-increment column in the local temporary table.

cn:enable_enhance_toast_table

Specifies whether to use the enhanced TOAST out-of-line storage table. The value on indicates that the enhanced TOAST out-of-line storage table is used. The value off indicates that the TOAST out-of-line storage table is used.

cn:archive_interval

Specifies the archiving interval. Log files are forcibly archived when the period specified by this parameter has elapsed. A large value of this parameter affects the RPO of PITR. The default value is recommended. Unit: second

dn:convert_illegal_char_mode

Specifies the placeholders of invalid characters that can be displayed on the client when enable_convert_illegal_char is enabled. Value range: 95 characters whose decimal codes range from 32 to 126 in the ASCII coding table.

dn:use_workload_manager

Specifies whether to enable resource management. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:use_workload_manager

Specifies whether to enable resource management. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_save_datachanged_timestamp

Specifies whether to record the time when INSERT, UPDATE, DELETE, or EXCHANGE/TRUNCATE/DROP PARTITION is performed on table data. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_save_datachanged_timestamp

Specifies whether to record the time when INSERT, UPDATE, DELETE, or EXCHANGE/TRUNCATE/DROP PARTITION is performed on table data. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_process_memory

Specifies the maximum physical memory allowed for a DN. Unit: KB This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_process_memory

Specifies the maximum physical memory allowed for a CN. Unit: KB This parameter is supported in GaussDB V2.0-8.201 or later.

dn:standby_shared_buffers_fraction

Specifies the shared_buffers proportion used on the server where a standby instance is deployed. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:standby_shared_buffers_fraction

Specifies the shared_buffers proportion used on the server where a standby instance is deployed. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_sleep

Specifies the interval (in milliseconds) for the page writer thread to flush dirty pages to disks after the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:pagewriter_sleep

Specifies the interval (in milliseconds) for the page writer thread to flush dirty pages to disks after the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:xloginsert_locks

Specifies the number of locks on concurrent write-ahead logging. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:xloginsert_locks

Specifies the number of locks on concurrent write-ahead logging. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_light_proxy

Specifies whether the optimizer optimizes the execution of simple queries on CNs. This parameter does not take effect if the character set of the application side does not match that of the kernel side. You are advised to set the character set to UTF-8 when creating a database. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_light_proxy

Specifies whether the optimizer optimizes the execution of simple queries on CNs. This parameter does not take effect if the character set of the application side does not match that of the kernel side. You are advised to set the character set to UTF-8 when creating a database. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:synchronous_commit

Specifies the synchronization mode of the current transaction. Generally, logs generated by a transaction are synchronized in the following sequence: The primary node writes the logs to the local memory. The primary node writes logs in the local memory to the local file system. The primary node flushes the logs in the local file system to disks. The primary node sends the logs to the standby node. The standby node receives the logs and saves them to its local memory. The standby node writes the logs in the local memory to the local file system. The standby node flushes the logs in the local file system to disks. The standby node replays the logs to complete the incremental update of data files. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:synchronous_commit

Specifies the synchronization mode of the current transaction. Generally, logs generated by a transaction are synchronized in the following sequence: The primary node writes the logs to the local memory. The primary node writes logs in the local memory to the local file system. The primary node flushes the logs in the local file system to disks. The primary node sends the logs to the standby node. The standby node receives the logs and saves them to its local memory. The standby node writes the logs in the local memory to the local file system. The standby node flushes the logs in the local file system to disks. The standby node replays the logs to complete the incremental update of data files. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_incremental_checkpoint

Specifies whether to enable incremental checkpoint. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_incremental_checkpoint

Specifies whether to enable incremental checkpoint. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bypass_workload_manager

Enables or disables independent I/O control. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bypass_workload_manager

Enables or disables independent I/O control. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_resource_track

Specifies whether to enable real-time resource monitoring. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_resource_track

Specifies whether to enable real-time resource monitoring. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_rt_percentile

Specifies whether to enable the function of calculating the response time of 80% and 95% of the SQL statements in the system. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_rt_percentile

Specifies whether to enable the function of calculating the response time of 80% and 95% of the SQL statements in the system. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_cpu_timer

Specifies whether to capture the CPU time consumed during SQL statement execution. In the x86-based centralized deployment scenario where 32 vCPUs and 256 GB of memory are configured, the performance tested by BenchmarkSQL 5.0 fluctuates by about 3.5% by enabling or disabling this parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_cpu_timer

Specifies whether to capture the CPU time consumed during SQL statement execution. In the x86-based centralized deployment scenario where 32 vCPUs and 256 GB of memory are configured, the performance tested by BenchmarkSQL 5.0 fluctuates by about 3.5% by enabling or disabling this parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_counts

Determines whether to enable collection of statistics on database activities. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:track_counts

Determines whether to enable collection of statistics on database activities. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_user_metric_persistent

Specifies whether to dump the historical monitoring data of user resources. If this parameter is set to on, data in the PG_TOTAL_USER_RESOURCE_INFO view is periodically sampled and saved to the system catalog GS_WLM_USER_RESOURCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_user_metric_persistent

Specifies whether to dump the historical monitoring data of user resources. If this parameter is set to on, data in the PG_TOTAL_USER_RESOURCE_INFO view is periodically sampled and saved to the system catalog GS_WLM_USER_RESOURCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:log_hostname

By default, connection log messages only show the IP address of the connecting host. The host name can be recorded when this parameter is set to on. It may take some time to parse the host name. Therefore, the database performance may be affected. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:log_hostname

By default, connection log messages only show the IP address of the connecting host. The host name can be recorded when this parameter is set to on. It may take some time to parse the host name. Therefore, the database performance may be affected. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_track_wait

Specifies whether to enable real-time collection of wait event information. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instr_track_wait

Specifies whether to enable real-time collection of wait event information. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instance_metric_persistent

Specifies whether to dump instance resource monitoring data. When this parameter is set to on, the instance monitoring data is saved to the system catalog GS_WLM_INSTANCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_instance_metric_persistent

Specifies whether to dump instance resource monitoring data. When this parameter is set to on, the instance monitoring data is saved to the system catalog GS_WLM_INSTANCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_logical_io_statistics

Specifies whether to collect logical I/O statistics during resource monitoring. If this function is enabled, the read_kbytes, write_kbytes, read_counts, write_counts, read_speed, and write_speed columns in the PG_TOTAL_USER_RESOURCE_INFO view will collect statistics on the byte count, number of times, and speed of logical read and write. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_logical_io_statistics

Specifies whether to collect logical I/O statistics during resource monitoring. If this function is enabled, the read_kbytes, write_kbytes, read_counts, write_counts, read_speed, and write_speed columns in the PG_TOTAL_USER_RESOURCE_INFO view will collect statistics on the byte count, number of times, and speed of logical read and write. Columns related to logical read and write in the system catalogs GS_WLM_USER_RESOURCE_HISTORY and GS_WLM_INSTANCE_HISTORY will collect statistics on the logical read and write of related users and instances. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_sort

Specifies whether the query optimizer uses sort methods. It is impossible to suppress explicit sorts entirely, but setting this variable to off encourages the optimizer to choose other methods if available. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_sort

Specifies whether the query optimizer uses sort methods. It is impossible to suppress explicit sorts entirely, but setting this variable to off encourages the optimizer to choose other methods if available. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_material

Specifies whether the query optimizer uses materialization. It is impossible to suppress materialization entirely, but setting this variable to off prevents the optimizer from inserting materialized nodes. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_material

Specifies whether the query optimizer uses materialization. It is impossible to suppress materialization entirely, but setting this variable to off prevents the optimizer from inserting materialized nodes. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:gtm_max_trans

Specifies the maximum number of connections accepted by the GTM. You are advised not to change the value. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:csn_sync_interval

Specifies the interval for synchronizing CSNs between the primary and standby GTMs, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

gtm:restore_duration

Specifies the reply interval of XIDs or CSNs on the GTM, that is, the number of XIDs or CSNs for reply. This parameter is supported in GaussDB V2.0-8.201 or later.

cms:coordinator_heartbeat_timeout

Specifies the heartbeat timeout (in seconds) that triggers the automatic removal of faulty CNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:client_min_messages

Specifies which level of messages will be sent to the client. A higher level covers the messages of all the lower levels. The lower the level is, the fewer messages are sent. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:client_min_messages

Specifies which level of messages will be sent to the client. A higher level covers the messages of all the lower levels. The lower the level is, the fewer messages are sent. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_thread_num

Specifies the number of threads for background page flushing after the incremental checkpoint is enabled. Dirty pages are flushed in sequence to disks, updating recovery points. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:pagewriter_thread_num

Specifies the number of threads for background page flushing after the incremental checkpoint is enabled. Dirty pages are flushed in sequence to disks, updating recovery points. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bgwriter_delay

Specifies the interval at which the background writer writes dirty shared buffers, in milliseconds. Each time, the background writer initiates write operations for some dirty buffers. In full checkpoint mode, the bgwriter_lru_maxpages parameter is used to control the amount of data to be written each time, and the process is restarted after the period of hibernation specified by bgwriter_delay (in milliseconds). In incremental checkpoint mode, the number of target idle buffer pages is calculated based on the value of candidate_buf_percent_target. If the number of idle buffer pages is insufficient, a batch of pages are flushed to disks at the interval specified by bgwriter_delay (in milliseconds). The number of flushed pages is calculated based on the target difference percentage. The maximum number of flushed pages is limited by max_io_capacity. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bgwriter_delay

Specifies the interval at which the background writer writes dirty shared buffers, in milliseconds. Each time, the background writer initiates write operations for some dirty buffers. In full checkpoint mode, the bgwriter_lru_maxpages parameter is used to control the amount of data to be written each time, and the process is restarted after the period of hibernation specified by bgwriter_delay (in milliseconds). In incremental checkpoint mode, the number of target idle buffer pages is calculated based on the value of candidate_buf_percent_target. If the number of idle buffer pages is insufficient, a batch of pages are flushed to disks at the interval specified by bgwriter_delay (in milliseconds). The number of flushed pages is calculated based on the target difference percentage. The maximum number of flushed pages is limited by max_io_capacity. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_files_per_process

Specifies the maximum number of files that can be opened simultaneously by each server process. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_files_per_process

Specifies the maximum number of files that can be opened simultaneously by each server process. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:candidate_buf_percent_target

Specifies the proportion (%) of available buffers in shared_buffer in the candidate buffer chain when the incremental checkpoint is enabled. If the percentage is less than the value of this parameter, the bgwriter thread starts to flush dirty pages based on the max_io_capacity parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:candidate_buf_percent_target

Specifies the proportion (%) of available buffers in shared_buffer in the candidate buffer chain when the incremental checkpoint is enabled. If the percentage is less than the value of this parameter, the bgwriter thread starts to flush dirty pages based on the max_io_capacity parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:checkpoint_timeout

Specifies the maximum time between automatic WAL checkpoints, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:checkpoint_timeout

Specifies the maximum time between automatic WAL checkpoints, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:incremental_checkpoint_timeout

Specifies the maximum interval (in seconds) between automatic WAL checkpoints when the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:incremental_checkpoint_timeout

Specifies the maximum interval (in seconds) between automatic WAL checkpoints when the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_writer_delay

Specifies the delay (in milliseconds) between activity rounds for the WAL writer. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:wal_writer_delay

Specifies the delay (in milliseconds) between activity rounds for the WAL writer. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_buffers

Specifies the shared memory for storing WALs, that is, multiples of XLOG_BLCKSZ, in increments of 8 KB. For details about the value range, see wal_buffers. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:wal_buffers

Specifies the shared memory for storing WALs, that is, multiples of XLOG_BLCKSZ, in increments of 8 KB. For details about the value range, see wal_buffers. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:walwriter_cpu_bind

Specifies the CPU core bound to the WAL write thread. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bulk_write_ring_size

Specifies the size of the ring buffer used by the operation when a large amount of data is written (for example, the copy operation). This parameter is supported in GaussDB V2.0-8.201 or later.

cn:bulk_write_ring_size

Specifies the size of a ring buffer used for parallel data import. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:recovery_redo_workers

Specifies the number of PageRedoWorker threads corresponding to each ParseRedoRecord thread in when ultimate RTO is enabled. recovery_redo_workers must be used together with recovery_parse_workers. The value of recovery_redo_workers takes effect only when recovery_parse_workers is greater than 1. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:recovery_redo_workers

Specifies the number of PageRedoWorker threads corresponding to each ParseRedoRecord thread in when ultimate RTO is enabled. recovery_redo_workers must be used together with recovery_parse_workers. The value of recovery_redo_workers takes effect only when recovery_parse_workers is greater than 1. This parameter is supported in GaussDB V2.0-8.201 or later.

cms:phony_dead_effective_time

Specifies the maximum number of times DN processes are detected as zombie. If the number of times a process is detected as zombie is greater than the specified value, the process is considered to be a zombie process and will be restarted. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_ai_watchdog

Enables or disables the AI watchdog function. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_ai_watchdog

Enables or disables the AI watchdog function. This parameter is supported in GaussDB V2.0-8.201 or later.

enable_unshipping_log

Specifies whether to log statements that are not pushed down. The logs help locate performance issues that may be caused by statements not pushed down. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_connections

Specifies the maximum number of concurrent connections. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:max_connections

Specifies the maximum number of concurrent connections. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_memory_reject_percent

Specifies the dynamic memory usage for escape from memory overload. This parameter is only applied when use_workload_manager and enable_memory_limit are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:resilience_memory_reject_percent

Specifies the dynamic memory usage for escape from memory overload. This parameter is only applied when use_workload_manager and enable_memory_limit are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_threadpool_reject_cond

Specifies the proportion of accumulated sessions in the thread pool for escape from overload. This parameter is only applied when enable_thread_pool and use_workload_manager are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:resilience_threadpool_reject_cond

Specifies the proportion of accumulated sessions in the thread pool for escape from overload. This parameter is only applied when enable_thread_pool and use_workload_manager are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:vacuum_cost_delay

Specifies how long the process sleeps when vacuum_cost_limit has been exceeded, in milliseconds. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:vacuum_cost_delay

Specifies how long the process sleeps when vacuum_cost_limit has been exceeded, in milliseconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:b_format_version

Specifies the SQL behavior control configuration item in B-compatible mode. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:b_format_version

Specifies the forward compatibility in MySQL-compatible mode. This is a configuration item for controlling database platform compatibility behavior. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_control_group

Specifies whether to enable the Cgroups. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:enable_control_group

Specifies whether to enable the Cgroups. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_stmt_parameter

After track_stmt_parameter is enabled, the executed statements recorded in statement_history are not normalized. The complete SQL statement information can be displayed to help the database administrator locate faults. For a simple query, the complete statement information is displayed. For a PBE statement, the complete statement information and information about each variable value are displayed. The format is "query string; parameters:$1=value1,$2=value2, ...". This parameter is used to display full SQL information for users and is not controlled by the track_activity_query_size parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:track_stmt_parameter

After track_stmt_parameter is enabled, the executed statements recorded in statement_history are not normalized. The complete SQL statement information can be displayed to help the database administrator locate faults. For a simple query, the complete statement information is displayed. For a PBE statement, the complete statement information and information about each variable value are displayed. The format is "query string; parameters:$1=value1,$2=value2, ...". This parameter is used to display full SQL information for users and is not controlled by the track_activity_query_size parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_activity_query_size

Specifies byte counts of the current running commands used to trace each active session. If the number of bytes in a command is greater than the value of this parameter, the command will be truncated. This parameter affects the memory usage of multiple system views. Adjusting the value of this parameter has a great impact on the memory usage. You can estimate the memory usage based on the value calculated from instr_unique_sql_count times track_activity_query_size. This parameter is supported in GaussDB V2.0-8.201 or later.

cn:track_activity_query_size

Specifies byte counts of the current running commands used to trace each active session. If the number of bytes in a command is greater than the value of this parameter, the command will be truncated. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:enable_auto_clean_unique_sql

Specifies whether to enable automatic cleaning of unique SQL statements in hash tables.

dn:gs_perf_retention_days

Specifies how many days the flame graph files in HTML format are retained. Unit: day

dn:page_version_max_num

Specifies the maximum number of page versions that can be cached in memory. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

log_autovacuum_min_duration

Specifies the interval which should elapse before autovacuum operations are logged. Autovacuum operations equal to or beyond the specified interval will be logged. If it is set to 0, all autovacuum operations will be logged. If it is set to -1, no autovacuum operations will be logged.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

dn:max_connections

Specifies the maximum number of concurrent connections to DNs. The value of this parameter must be greater than that of max_wal_senders.

datestyle

Specifies the display format for date and time.

max_replication_slots

Specifies the number of log replication slots on the primary node.

timezone

Specifies the time zone for displaying and interpreting time stamps.

dn:enable_xid_abort_check

Specifies whether to check the status of transaction ID rollback when a transaction is committed.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots. Currently, only archive slots and backup slots are involved. Set this parameter to on only in cloud scenarios where logical replication is enabled. In other scenarios, set this parameter to off.

dn:enable_enhance_toast_table

Specifies whether to use the enhanced TOAST out-of-line storage table. The value on indicates that the enhanced TOAST out-of-line storage table is used. The value off indicates that the TOAST out-of-line storage table is used.

dn:support_binary_copy_version

Specifies whether to verify the binary file encoding information when data is imported using COPY FROM in BINARY mode.

dn:recovery_time_target

Specifies the time for the standby node to write and replay logs. Unit: second

plat_compat_allow_public_key_retrieval

Specifies the database configuration item in M-compatible mode. This parameter specifies whether a client can request the RSA public key. on: The kernel allows the client to request the RSA public key for password transmission encryption. off: The client is not allowed to request the RSA public key.

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

dn:enable_rls_match_index

Specifies whether indexes of a base table can be scanned based on target predicate conditions in row-level security scenarios. Target scenario: Row level security (RLS) policies are set and enabled in the base table, and the query predicate contains the unleakproof system function or like operator.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

dn:audit_internal_event

Specifies whether to audit the connections and operations of internal tools cm_agent, gs_clean, and WDRXdb, and whether to audit the logins and logouts from CNs on DNs.

dn:numa_distribute_mode

Specifies the distribution of some shared data and threads among NUMA nodes. This parameter is used to optimize the performance of large-scale Arm servers with multiple NUMA nodes. Generally, you do not need to set this parameter. This parameter cannot be set to all for instances deployed on x86-based servers.

dn:max_compile_packages

Specifies the maximum number of package compilation results stored in the server. Default values for different specifications are calculated by rounding down (max_process_memory x 2%)/4.4, in MB.

dn:fix_func_selection

Specifies whether to optimize the function matching policy. The value catlist indicates the catlist sequence is optimized. (The non-B-compatible mode has been optimized. In non-B-compatible mode, system functions are always preferentially selected and executed. The policy in B-compatible mode is the same as that in versions earlier than 505.1.0. An error message indicating that the function is not unique may be displayed, or a system function may be selected for execution.)

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

dn:sql_mode

Specifies the SQL behavior control configuration item in M-compatible mode.

dn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

dn:undo_space_limit_size

Specifies the threshold for forcibly recycling undo space. When the undo space usage reaches 80% of the threshold, forcible recycling starts. It is recommended that the value of this parameter be greater than or equal to the value of undo_limit_size_per_transaction. Unit: 8 KB

dn:umdk_enabled

Specifies whether UMDK is enabled for the primary and standby DNs of the current instance. If the UMDK protocol is used for communication between the primary and standby DNs, the related log keyword on DNs is umdk. If the TCP protocol is used for communication between the primary and standby DNs, logs are recorded.

dn:verify_log_buffers

Specifies the size or pages of verifyLog buffers in memory mode. The unit is 8 KB. For example, if the value of this parameter is 4, the requested memory is 4 x 8 KB = 32 KB. This parameter is valid only when page_version_check is set to persistence. If page_version_check is set to another value, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to persistence.

dn:llvm_max_memory

Specifies the upper limit of the memory used by IRs (including cached and in-use IRs) generated during compilation in Codegen execution mode. The memory used by Codegen is not applied for by preoccupation. It is a part of max_dynamic_memory and is restricted by the llvm_max_memory parameter. Unit: KB

dn:enable_control_group

Specifies whether to enable the Cgroups.

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

dn:enable_convert_illegal_char

Specifies whether the database supports characters not included the character sets.

dn:system_view_version

Determines the version of the system view. All versions are backward compatible. For example, when system_view_version is set to 3, all features of version 2 and version 1 are also supported. For details, see the product documentation.

track_activity_query_size

Specifies the maximum number of bytes to be logged for each SQL statement. If the number of bytes of a SQL statement exceeds the specified parameter value, the SQL statement will be truncated. Setting this parameter to a large value can lead to excessive memory usage, potentially causing the system to run out of memory. You are advised to set this parameter to a value no more than 4096.

dn:enable_workload_rule

Specifies whether to enable SQL throttling.

dn:archive_interval

Specifies the archiving interval. Log files are forcibly archived when the period specified by this parameter has elapsed. A large value of this parameter affects the RPO of PITR. The default value is recommended. Unit: second

div_precision_increment

Specifies the database configuration item in M-compatible mode. This is a session-level parameter, which is used to set the value of precision that the division result can improve. The final precision is the precision of the first operation parameter added by the value of this parameter.

lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

dn:adaptive_cost_min_time

Parameter description: Specifies the execution duration threshold of SQL statements for cardinality feedback collection. Only the feedback of statements whose execution duration is greater than the value of this parameter is collected. Unit: millisecond

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:codegen_compile_thread_num

Specifies the number of Codegen compilation threads.

dn:enable_vectordb

Specifies whether vector indexes can be created, inserted, updated, and queried.

max_wal_senders

Specifies the maximum number of WAL sender threads that can be created. Processes created for the following purposes occupy WAL sender threads: Standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. If this parameter is set to a value smaller than 20, scale-out may fail. The value of this parameter must be smaller than that of max_connections.

dn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed. The theoretical maximum value is 10000. Set this parameter based on workload requirements and hardware configurations. It is recommended that this parameter be set to a value less than or equal to 1/10 of max_connections.

dn:enable_extension

Specifies whether database extensions can be created. This parameter can be used only in specific scenarios after evaluation. Generally, this parameter does not need to be adjusted.

dn:max_compile_functions

Specifies the maximum number of function compilation results stored in the server. Excessive functions and compilation results of stored procedures may occupy large memory space. Setting this parameter to an appropriate value can reduce the memory usage and improve system performance. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

dn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: millisecond

dn:enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

dn:enable_vacuum_control

Specifies whether to move the database permanent thread autoVacuumWorker to the Vacuum Cgroup.

dn:recovery_parse_workers

Specifies the number of ParseRedoRecord threads in the ultimate RTO feature. This parameter must be used together with recovery_redo_workers. If both recovery_parse_workers and recovery_redo_workers are greater than 1, ultimate RTO is enabled. If you do not want to enable ultimate RTO, retain the default value 1 for recovery_parse_workers. When enabling ultimate RTO, ensure that replication_type is set to 1. If both ultimate RTO and parallel replay are enabled, only ultimate RTO takes effect, and parallel replay is ineffective. Ultimate RTO does not support column-store tables. Therefore, disable ultimate RTO in a system where column-store tables are used or are to be used. Ultimate RTO also does not support flow control. Flow control is determined by the recovery_time_target parameter.

dn:page_version_check

Specifies the type of page version verification. off indicates that page version verification is disabled. memory indicates that page version verification in pure memory mode is enabled. The page version information will be lost after a restart. persistence indicates that persistent page version verification is enabled. The page version information will not be lost after a restart.

dn:track_activity_query_size

Specifies the maximum number of bytes to be logged for each SQL statement. If the number of bytes of a SQL statement exceeds the specified parameter value, the SQL statement will be truncated. Setting this parameter to a large value can lead to excessive memory usage, potentially causing the system to run out of memory. You are advised to set this parameter to a value no more than 4096.

dn:gsivfflat_secondary_probes

Specifies the number of level-2 inverted indexes to be searched. If the value exceeds the total number of level-2 inverted indexes, the entire table is searched.

dn:num_internal_lock_partitions

Specifies the number of internal lightweight lock partitions. Changing the value of this parameter affects performance and memory usage. Before modifying this parameter, determine your application scenario and fully verify the change in a test environment. For details, see the reference document.

dn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

dn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions). The value 0 means to disable the function, and 1 means to enable it.

dn:auto_increment_cache

Specifies the number of reserved auto-increment cache values when auto-increment is triggered by batch insertion or import of auto-increment columns. When auto-increment values are reserved, the auto-increment counter value is updated to the maximum auto-increment cache value. Before the cache values are used up, the auto-increment counter value remains unchanged, and the triggered auto-increment uses the cache values. If this parameter is set to 0, the auto-increment cache values are automatically reserved. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, 2^n values are reserved each time. For example, one value is reserved in the first auto-increment, two values are reserved in the second auto-increment, four values are reserved in the third auto-increment, and eight values are reserved for in fourth auto-increment. However, if the number of reserved values exceeds 65,535, 65,535 values are reserved. If this parameter is not set to 0, the number of reserved cache values is the value of this parameter. When auto-increment is triggered for the first time, if the number of rows to be inserted into the auto-increment column is known, the number is the reserved value. If the number of rows is unknown, the value of auto_increment_cache is the number of auto-increment values reserved each time. The reserved cache values are valid only in the statement. If the reserved auto-increment cache values are used up and subsequent INSERT statements trigger auto-increment based on the auto-increment counter, the values in the auto-increment column in the table are discontinuous. This parameter does not affect the auto-increment column in the local temporary table.

dn:enable_codegen

Specifies whether code optimization can be enabled. Currently, code optimization uses the LLVM optimization.

dn:instr_unique_sql_combination_options

Specifies the configuration items of combining unique SQL statements of the same type. If this feature is enabled, the IDs of unique SQL statements of the same type are normalized, and the generated unique SQL strings are normalized.

plan_cache_mode

Specifies the policy of generating and caching execution plans of prepared statements.

dn:group_concat_max_len

Specifies the maximum permitted result length in bytes for the GROUP_CONCAT() function.

dn:hadr_recovery_point_target

Specifies the time allowed for the standby instance to flush logs to disks in streaming DR. This ensures that the log difference between the primary and standby instances is controlled within the period specified by hadr_recovery_point_target during the switchover and the standby instance can be promoted to primary. If this parameter is set to a small value, the performance of the primary node is affected. If it is set to a large value, the log flow is not effectively controlled. The value 0 indicates that log flow control is disabled. Unit: second

dn:behavior_compat_options

Specifies database compatibility configuration items. After the value of proc_outparam_override is changed, the database must be connected again or the instance must be rebooted. Otherwise, stored procedures and functions cannot be correctly called.

dn:page_missing_dirty_check

Specifies whether to enable the verification for pages not marked as dirty. The verification checks whether the modified pages are not marked as dirty. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

dn:wdr_snapshot_full_backup_interval

Specifies the interval at which a full WDR snapshot is created. The interval specified by this parameter is about a number instead of time. For example, if the parameter is set to 12, a full snapshot and then 11 incremental snapshots are generated for each group. If the parameter is set to 1, all snapshots generated are full snapshots.

dn:vacuum_defer_cleanup_age

Specifies the number of transactions used by VACUUM.

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:wdr_snapshot_space_threshold

Specifies the threshold for controlling the space used by snapshots. When the space used by snapshots reaches 80% of the value of this parameter, the control logic of the database is enabled to stabilize the space usage. Unit: KB

dn:m_format_behavior_compat_options

Specifies the configuration items for the M-compatible mode.

dn:gs_perf_interval

Specifies the automatic perf data collection interval. The value 0 indicates that the collection is stopped. If the value is greater than 0 and less than 5, the value 5 is used. Unit: minute

dn:extra_float_digits

Adjusts the number of digits displayed for floating-point values, including float4, float8, and geometric data types. The parameter value is added to the standard number of digits (FLT_DIG or DBL_DIG as appropriate). This parameter can also be set to a negative value to suppress unwanted digits.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

max_allowed_packet

Specifies the database configuration item in M-compatible mode. This parameter indicates the upper limit of the return value of a string function. The value must be a multiple of 1,024. Unit: KB

track_stmt_stat_level

Controls the level of statement execution tracking.

dn:max_standby_archive_delay

Specifies the wait period before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

dn:copy_special_character_version

Specifies whether to report an error when there are invalid characters during data import and export using COPY FROM.

log_temp_files

Specifies whether to log temporary file details when they are deleted. A positive value indicates that the deletion information of temporary files whose size is larger than the value specified by log_temp_files is recorded. 0 indicates that the deletion information of all temporary files is recorded. -1 indicates that the deletion information of any temporary files is not recorded. Unit: KB

enable_wdr_snapshot

Specifies whether to enable database monitoring snapshots on the standby node.

dn:m_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

password_effect_time

Specifies the validity period of an account password. Unit: day

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:immediate_analyze_threshold

Specifies the threshold for triggering ANALYZE. When the amount of inserted data reaches the original data amount multiplied by the value of immediate_analyze_threshold, and the total number of rows exceeds 100, ANALYZE is automatically triggered.

dn:auto_increment_increment

Specifies the auto-increment step of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is smaller than that of auto_increment_offset, there will be an error when the values in the auto-increment column automatically increase.

dn:gs_format_behavior_compat_options

Specifies the configuration items of GaussDB internal system functions.

enable_nestloop

Controls whether the query optimizer uses the nested-loop join plan type to fully scan inner tables.

dn:b_format_dev_version

Specifies the compatibility configuration item of database platform minor versions.

dn:check_disconnect_query

Specifies whether to enable the function of terminating statement execution on the server after the client is disconnected due to timeout.

dn:wal_keep_segments

Specifies the minimum number of transaction log files stored in the pg_xlog directory. Standby nodes obtain the logs from the primary node to perform streaming replication.

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

dn:max_standby_streaming_delay

Specifies the wait period before queries on the standby node are canceled when the queries conflict with WAL data receiving through streaming replication in hot standby mode. -1 indicates that the standby node waits until the conflicting queries are complete. Unit: millisecond

audit_system_object

Specifies whether to audit the CREATE, DROP, and ALTER operations on GaussDB database objects. GaussDB database objects include databases, users, schemas, and tables. You can change the value of this parameter to audit only the operations on required database objects. In the scenario where the leader node is forcibly selected, you are advised to set audit_system_object to the maximum value and audit all DDL objects. For details about the value range, see the product documentation.

dn:enable_copy_server_files

Specifies whether to enable the privilege to copy server files. If this parameter is set to on, users with the SYSADMIN privilege or users who inherit the built-in role privilege gs_role_copy_files are allowed to run the COPY FROM FILENAME or COPY TO FILENAME command. If it is set to off, only the initial user is allowed to run the COPY FROM FILENAME or COPY TO FILENAME statement.

checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout.

dn:page_version_partitions

Specifies the number of hash table partitions in cached page version information in the memory. This parameter directly affects the hash query efficiency and hash conflict probability, and is valid only when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off. For details about the value range, see the product documentation.

dn:password_encryption_type

Specifies how user passwords are encrypted. 0: Passwords are encrypted using MD5. 1: Passwords are encrypted using SHA-256 and MD5. 2: Passwords are encrypted using SHA-256. 3: Passwords are encrypted using SM3. MD5 is not recommended because it is a weak encryption algorithm.

dn:enable_plsql_opfusion

Specifies whether to enable SQLBypass for stored procedures.

dn:page_version_recycler_thread_num

Specifies the number of threads for recycling and verifying page version information. This parameter is only valid when page_version_check is not set to off. If page_version_check is set to off, the parameter value will still be sent to the kernel, but the relevant function does not take effect until after page_version_check is set to a value other than off.

dn:dcf_thread_effective_time

Specifies the effective time of the DCF flushing thread. This parameter is used to determine whether the disk I/O hangs. If the DCF cannot access I/O resources within the period specified by this parameter, the DCF considers that the thread I/O hangs and a primary/standby switchover is triggered. If this parameter is set to 0, I/O hang detection is disabled. Unit: second

dn:audit_login_logout

Specifies whether to audit users' logins (including successful and failed logins) and logouts. This parameter can be configured for specific PDBs. If this parameter is not specified for a PDB, the PDB inherits the global setting of this parameter. 0: Disable user login and logout auditing. 1: Audit only successful logins. 2: Audit only failed logins. 3: Audit both successful and failed logins. 4: Audit only user logouts. 5: Audit only user logouts and successful logins. 6: Audit only user logouts and failed logins. 7: Audit successful and failed logins, as well as user logouts.

dn:a_format_date_timestamp

Specifies whether to enable current_date, current_timestamp, and localtimestamp to return the system time, instead of the transaction start time, when a transaction starts.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

temp_file_limit

Specifies the maximum amount of disk space that a session can use for temporary files. The value –1 indicates there is no limit. Unit: KB

dn:disable_keyword_options

Specifies the configuration items for database compatibility. Multiple items are separated by commas (,). An identifier with this parameter set will not be used as a keyword.

dn:hadr_recovery_time_target

Specifies whether the standby DB instance completes log writing and replay in streaming DR mode. If this parameter is set to a small value, the performance of the primary node is affected. If it is set to a large value, the log flow is not effectively controlled. The value 0 indicates that log flow control is disabled. Unit: second

password_lock_time

Specifies the duration for a locked account to be automatically unlocked. Unit: day

wdr_snapshot_interval

Specifies the interval at which the background thread Snapshot automatically takes snapshots of the database monitoring data. Unit: minute

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

dn:wal_file_preinit_bounds

Specifies the maximum number of WAL segment files that can be pre-expanded by the WAL writer auxiliary thread per second during service running. The WAL segment file size is 16 MB. If this parameter is set to 0, there is no restriction.

dn:enable_feedback_cardest

Specifies whether to enable the feedback-based optimizer cardinality and cost correction functions. This parameter is used by developers to diagnose model-related problems. If enable_adaptive_cost is set to off and this parameter is set to on, the operator information is still collected and the API of cardinality estimation feedback is still called. However, in this case, the thread for background automatic model maintenance will not be enabled. You can use the gs_acm_analyze_workload_manual() function to manually train models for diagnosing problems.

dn:enable_global_plsqlcache

Specifies whether to globally cache compilation products of packages, stored procedures, and functions, and cache execution products at the session level. Enabling this function can reduce the memory usage of database nodes in high concurrency scenarios.

dn:convert_illegal_char_mode

Specifies the placeholders of invalid characters that can be displayed on the client when enable_convert_illegal_char is enabled. Value range: 95 characters whose decimal codes range from 32 to 126 in the ASCII coding table.

dn:advance_xlog_file_num

Specifies the number of Xlog files that are periodically initialized in advance in the background. This parameter is used to prevent the Xlog file initialization from affecting the performance during transaction commit. However, such a fault may occur only when the system is overloaded. Therefore, you do not need to set this parameter.

enable_mergejoin

Controls whether the query optimizer uses the merge-join plan type.

dn:tde_index_default_encrypt

When tde_index_default_encrypt is set to on and an index is created based on an encrypted table, the database automatically converts the index to an encrypted index.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

undo_retention_time

Specifies how long undo logs are kept. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained. Unit: second

enable_global_syscache

Determines whether to enable global system cache.

dn:max_io_capacity

Specifies the maximum I/O per second for the background writer process to flush pages in batches. Set this parameter based on the service scenario and the disk I/O capability. If the RTO is short or the data volume is many times that of the shared memory and the service access data volume is random, the value of this parameter cannot be too small. A small value of max_io_capacity reduces the number of pages flushed by the background writer. If a large number of pages are evicted due to service triggering, the services are affected. Unit: KB

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

index_txntype

Sets the index page format to PCR or RCR. This parameter is left unconfigured during system initialization. By default, the created indexes are compatible with the index type (RCR) of earlier versions. Once this parameter is specified, it cannot be left unconfigured again.

dn:auto_increment_offset

Specifies the initial value of an auto-increment column. The auto-increment value is calculated by the following formula: auto_increment_offset + N × auto_increment_increment. N is a positive integer. If the value of this parameter is greater than that of auto_increment_increment, there will be an error when the values in the auto-increment column automatically increase.

autovacuum_naptime

Specifies the minimum delay between autovacuum runs on any given database. If this parameter is set to a smaller value, the load is more stable but the I/O increases. If this parameter is set to a larger value, the load may fluctuate more but the I/O decreases. Unit: second

dn:diskann_probe_ncandidates

Specifies the size of the candidate set when the gsdiskann index is used to retrieve vectors.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. The value 0 indicates that there is no timeout limit. Unit: second

dn:use_workload_manager

Specifies whether to enable resource management. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_save_datachanged_timestamp

Specifies whether to record the time when INSERT, UPDATE, DELETE, or EXCHANGE/TRUNCATE/DROP PARTITION is performed on table data. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_process_memory

Specifies the maximum physical memory allowed for a DN. Unit: KB This parameter is supported in GaussDB V2.0-8.201 or later.

dn:standby_shared_buffers_fraction

Specifies the shared_buffers proportion used on the server where a standby instance is deployed. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_sleep

Specifies the interval (in milliseconds) for the page writer thread to flush dirty pages to disks after the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:xloginsert_locks

Specifies the number of locks on concurrent write-ahead logging. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_light_proxy

Specifies whether the optimizer optimizes the execution of simple queries on CNs. This parameter does not take effect if the character set of the application side does not match that of the kernel side. You are advised to set the character set to UTF-8 when creating a database. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:synchronous_commit

Specifies the synchronization mode of the current transaction. Generally, logs generated by a transaction are synchronized in the following sequence: The primary node writes the logs to the local memory. The primary node writes logs in the local memory to the local file system. The primary node flushes the logs in the local file system to disks. The primary node sends the logs to the standby node. The standby node receives the logs and saves them to its local memory. The standby node writes the logs in the local memory to the local file system. The standby node flushes the logs in the local file system to disks. The standby node replays the logs to complete the incremental update of data files. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_incremental_checkpoint

Specifies whether to enable incremental checkpoint. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bypass_workload_manager

Enables or disables independent I/O control. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_rt_percentile

Specifies whether to enable the function of calculating the response time of 80% and 95% of the SQL statements in the system. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_cpu_timer

Specifies whether to capture the CPU time consumed during SQL statement execution. In the x86-based centralized deployment scenario where 32 vCPUs and 256 GB of memory are configured, the performance tested by BenchmarkSQL 5.0 fluctuates by about 3.5% by enabling or disabling this parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_user_metric_persistent

Specifies whether to dump the historical monitoring data of user resources. If this parameter is set to on, data in the PG_TOTAL_USER_RESOURCE_INFO view is periodically sampled and saved to the system catalog GS_WLM_USER_RESOURCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:log_hostname

By default, connection log messages only show the IP address of the connecting host. The host name can be recorded when this parameter is set to on. It may take some time to parse the host name. Therefore, the database performance may be affected. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instr_track_wait

Specifies whether to enable real-time collection of wait event information. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_instance_metric_persistent

Specifies whether to dump instance resource monitoring data. When this parameter is set to on, the instance monitoring data is saved to the system catalog GS_WLM_INSTANCE_HISTORY. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_logical_io_statistics

Specifies whether to collect logical I/O statistics during resource monitoring. If this function is enabled, the read_kbytes, write_kbytes, read_counts, write_counts, read_speed, and write_speed columns in the PG_TOTAL_USER_RESOURCE_INFO view will collect statistics on the byte count, number of times, and speed of logical read and write. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:client_min_messages

Specifies which level of messages will be sent to the client. A higher level covers the messages of all the lower levels. The lower the level is, the fewer messages are sent. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:pagewriter_thread_num

Specifies the number of threads for background page flushing after the incremental checkpoint is enabled. Dirty pages are flushed in sequence to disks, updating recovery points. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bgwriter_delay

Specifies the interval at which the background writer writes dirty shared buffers, in milliseconds. Each time, the background writer initiates write operations for some dirty buffers. In full checkpoint mode, the bgwriter_lru_maxpages parameter is used to control the amount of data to be written each time, and the process is restarted after the period of hibernation specified by bgwriter_delay (in milliseconds). In incremental checkpoint mode, the number of target idle buffer pages is calculated based on the value of candidate_buf_percent_target. If the number of idle buffer pages is insufficient, a batch of pages are flushed to disks at the interval specified by bgwriter_delay (in milliseconds). The number of flushed pages is calculated based on the target difference percentage. The maximum number of flushed pages is limited by max_io_capacity. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_files_per_process

Specifies the maximum number of files that can be opened simultaneously by each server process. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:candidate_buf_percent_target

Specifies the proportion (%) of available buffers in shared_buffer in the candidate buffer chain when the incremental checkpoint is enabled. If the percentage is less than the value of this parameter, the bgwriter thread starts to flush dirty pages based on the max_io_capacity parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

checkpoint_segments

Specifies the minimum number of WAL segment files in the period specified by checkpoint_timeout. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:checkpoint_timeout

Specifies the maximum time between automatic WAL checkpoints, in seconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:incremental_checkpoint_timeout

Specifies the maximum interval (in seconds) between automatic WAL checkpoints when the incremental checkpoint is enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_writer_delay

Specifies the delay (in milliseconds) between activity rounds for the WAL writer. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:wal_buffers

Specifies the shared memory for storing WALs, that is, multiples of XLOG_BLCKSZ, in increments of 8 KB. For details about the value range, see wal_buffers. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:walwriter_cpu_bind

Specifies the CPU core bound to the WAL write thread. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:bulk_write_ring_size

Specifies the size of the ring buffer used by the operation when a large amount of data is written (for example, the copy operation). This parameter is supported in GaussDB V2.0-8.201 or later.

dn:recovery_redo_workers

Specifies the number of PageRedoWorker threads corresponding to each ParseRedoRecord thread in when ultimate RTO is enabled. recovery_redo_workers must be used together with recovery_parse_workers. The value of recovery_redo_workers takes effect only when recovery_parse_workers is greater than 1. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_ai_watchdog

Enables or disables the AI watchdog function. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:max_connections

Specifies the maximum number of concurrent connections. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_memory_reject_percent

Specifies the dynamic memory usage for escape from memory overload. This parameter is only applied when use_workload_manager and enable_memory_limit are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:resilience_threadpool_reject_cond

Specifies the proportion of accumulated sessions in the thread pool for escape from overload. This parameter is only applied when enable_thread_pool and use_workload_manager are enabled. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:vacuum_cost_delay

Specifies how long the process sleeps when vacuum_cost_limit has been exceeded, in milliseconds. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:b_format_version

Specifies the SQL behavior control configuration item in B-compatible mode. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:enable_control_group

Specifies whether to enable the Cgroups. This parameter must be applied on both CNs and DNs. This parameter is supported in GaussDB V2.0-8.201 or later.

dn:track_stmt_parameter

After track_stmt_parameter is enabled, the executed statements recorded in statement_history are not normalized. The complete SQL statement information can be displayed to help the database administrator locate faults. For a simple query, the complete statement information is displayed. For a PBE statement, the complete statement information and information about each variable value are displayed. The format is "query string; parameters:$1=value1,$2=value2, ...". This parameter is used to display full SQL information for users and is not controlled by the track_activity_query_size parameter. This parameter is supported in GaussDB V2.0-8.201 or later.

track_activity_query_size

Specifies byte counts of the current running commands used to trace each active session. If the number of bytes in a command is greater than the value of this parameter, the command will be truncated. This parameter affects the memory usage of multiple system views. Adjusting the value of this parameter has a great impact on the memory usage. You can estimate the memory usage based on the value calculated from instr_unique_sql_count times track_activity_query_size. This parameter is supported in GaussDB V2.0-8.201 or later.

audit_system_object

Determines whether to audit the CREATE, DROP, and ALTER operations on GaussDB Kernel database objects. GaussDB Kernel database objects include databases, users, schemas, and tables. You can change the parameter value to audit only the operations on required database objects. During a forcible primary/standby failover, set audit_system_object to the maximum value and audit all DDL objects. If the parameter value is incorrectly changed, DDL audit logs will be lost. Contact technical support to change it.

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. 0 indicates there is no timeout. Unit: second

cn:effective_cache_size

Specifies the size of the disk buffer available to the CN optimizer in a single query. Unit: 8 KB

cn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

cn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a CN session.

datestyle

Specifies the display format for date and time.

dn:effective_cache_size

Specifies the size of the disk buffer available to the DN optimizer in a single query. Unit: 8 KB

dn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

dn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a DN session.

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots.

enable_stream_operator

Specifies the query optimizer's use of streams. When this parameter is set to off, a large number of logs indicating that the stream plans cannot be pushed down are recorded.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

max_replication_slots

Specifies the number of log replication slots on the primary node.

max_wal_senders

The following processes occupy walsender threads: standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. This parameter specifies the maximum number of walsender threads that can be created.

password_effect_time

Specifies the validity period of the password, in days.

password_lock_time

Specifies the duration for a locked account to be automatically unlocked, in days.

recovery_time_target

Specifies the time for the standby node to write and replay logs, in seconds.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

timezone

Specifies the time zone for displaying and interpreting time stamps.

track_stmt_stat_level

Controls the level of statement execution tracking.

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

cn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

cn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

cn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

cn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

cn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

cn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

cn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

cn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

cn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

cn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

wdr_snapshot_interval

Specifies the interval (in minutes) at which the background thread Snapshot automatically performs snapshot operations on the database monitoring data.

cn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

dn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

enable_wdr_snapshot

Specifies whether to enable WDR snapshots.

cn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed.

dn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed.

cn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

dn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

cn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

cn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

cn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: ms

dn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: ms

cn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions).

dn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions).

cn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

cn:enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

dn:enable_default_ustore_table

Specifies whether to enable the Ustore storage engine by default. If this parameter is set to on, all created tables are Ustore tables.

cn:undo_space_limit_size

Specifies the undo forcible reclamation threshold. If 80% of the specified parameter value is reached, forcible reclamation is triggered. The unit is 8 KB. It is recommended that the value be at least the value of undo_limit_size_per_transaction.

dn:undo_space_limit_size

Specifies the undo forcible reclamation threshold. If 80% of the specified parameter value is reached, forcible reclamation is triggered. The unit is 8 KB. It is recommended that the value be at least the value of undo_limit_size_per_transaction.

cn:undo_limit_size_per_transaction

Specifies the maximum undo space for a single transaction. The unit is 8 KB. If the undo space of a transaction exceeds this parameter value, the transaction is rolled back due to an error. It is recommended that this parameter value be smaller than the value of undo_space_limit_size. If this parameter value is larger, the value of undo_space_limit_size will be used as the maximum undo space for a single transaction.

dn:undo_limit_size_per_transaction

Specifies the maximum undo space for a single transaction. The unit is 8 KB. If the undo space of a transaction exceeds this parameter value, the transaction is rolled back due to an error. It is recommended that this parameter value be smaller than the value of undo_space_limit_size. If this parameter value is larger, the value of undo_space_limit_size will be used as the maximum undo space for a single transaction.

cn:enable_recyclebin

Enables or disables the recycle bin in real time.

dn:enable_recyclebin

Enables or disables the recycle bin in real time.

cn:recyclebin_retention_time

Specifies how long files will be kept in the recycle bin, in seconds. Files in the recycle bin will be automatically deleted after this length of time.

dn:recyclebin_retention_time

Specifies how long files will be kept in the recycle bin, in seconds. Files in the recycle bin will be automatically deleted after this length of time.

cn:undo_retention_time

Specifies how long undo logs are kept, in seconds. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained.

dn:undo_retention_time

Specifies how long undo logs are kept, in seconds. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained.

cn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

dn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

cn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

dn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

cn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

dn:resilience_ctrlslot_available_maxpercent

Specifies the maximum percentage of threads in the thread pool that can be occupied by slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users.

cn:resilience_ctrlstmt_control_iopslimit

Specifies the maximum IOPS that can be used by slow SQL statements after normal SQL statements are marked as slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. 0(None): The IOPS is not limited. 10(LOW): The limit level for IOPS is LOW. 20(MEDIUM): The limit level for IOPS is MEDIUM. 50(HIGH): The limit level for IOPS is HIGH.

dn:resilience_ctrlstmt_control_iopslimit

Specifies the maximum IOPS that can be used by slow SQL statements after normal SQL statements are marked as slow SQL statements. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. 0(None): The IOPS is not limited. 10(LOW): The limit level for IOPS is LOW. 20(MEDIUM): The limit level for IOPS is MEDIUM. 50(HIGH): The limit level for IOPS is HIGH.

dn:resilience_ctrlstmt_detect_timelimit

Specifies the execution time of a normal SQL statement that will be marked as a slow SQL statement. The value 0 indicates that slow SQL statements are not identified. A value greater than 0 indicates that a normal SQL statement whose execution time exceeds the value of this parameter is marked as a slow SQL statement. This parameter is only suitable for SELECT statements executed by non-sysadmin/monitoradmin users. Unit: ms

audit_system_object

Determines whether to audit the CREATE, DROP, and ALTER operations on GaussDB Kernel database objects. GaussDB Kernel database objects include databases, users, schemas, and tables. You can change the parameter value to audit only the operations on required database objects. During a forcible primary/standby failover, set audit_system_object to the maximum value and audit all DDL objects. If the parameter value is incorrectly changed, DDL audit logs will be lost. Contact technical support to change it.

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. 0 indicates there is no timeout. Unit: second

datestyle

Specifies the display format for date and time.

dn:wal_keep_segments

Specifies the minimum number of transaction log files stored in the pg_xlog directory. Standby nodes obtain the logs from the primary node to perform streaming replication.

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

max_replication_slots

Specifies the number of log replication slots on the primary node.

max_wal_senders

The following processes occupy walsender threads: standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. This parameter specifies the maximum number of walsender threads that can be created.

password_effect_time

Specifies the validity period of the password, in days.

password_lock_time

Specifies the duration for a locked account to be automatically unlocked, in days.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

timezone

Specifies the time zone for displaying and interpreting time stamps.

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:vacuum_defer_cleanup_age

Specifies the number of transactions used by VACUUM.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

wdr_snapshot_interval

Specifies the interval (in minutes) at which the background thread Snapshot automatically performs snapshot operations on the database monitoring data.

undo_retention_time

Specifies how long undo logs are kept, in seconds. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained.

track_stmt_stat_level

Controls the level of statement execution tracking.

dn:enable_auto_explain

Specifies whether to automatically print execution plans. This parameter is used to locate slow stored procedures or slow queries.

enable_wdr_snapshot

Specifies whether to enable WDR snapshots.

dn:max_concurrent_autonomous_transactions

Specifies the maximum number of autonomous transaction connections, that is, the maximum number of concurrent autonomous transactions executed at the same time. If this parameter is set to 0, autonomous transactions cannot be executed.

dn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:auto_explain_log_min_duration

Specifies the minimum duration of execution plans that are automatically printed. Only execution plans whose duration is greater than the value of auto_explain_log_min_duration will be printed. Unit: ms

dn:recovery_time_target

Specifies the time for the standby node to write and replay logs. Unit: second

dn:audit_function_exec

Specifies whether to record the audit information during the execution of the stored procedures, anonymous blocks, or user-defined functions (excluding system functions).

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

dn:undo_space_limit_size

Specifies the undo forcible reclamation threshold. If 80% of the specified parameter value is reached, forcible reclamation is triggered. The unit is 8 KB. It is recommended that the value be at least the value of undo_limit_size_per_transaction.

dn:group_concat_max_len

Specifies the maximum permitted result length in bytes for the GROUP_CONCAT() function.

dn:enable_extension

Specifies whether database extensions can be created. The extension plug-in is a lab feature and is not recommended.

dn:cost_model_version

Specifies the version of the optimizer cost model. It is a protective parameter. It prevents new optimizer cost models from being applied, so you can keep the current model consistent with the plan of an existing version. If the value of this parameter is changed, many SQL plans may be changed. Exercise caution when modifying this parameter.

dn:immediate_analyze_threshold

Specifies the threshold for triggering ANALYZE. When the amount of inserted data reaches the original data amount multiplied by the value of immediate_analyze_threshold, and the total number of rows exceeds 100, ANALYZE is automatically triggered.

dn:enable_dynamic_samplesize

Specifies whether to dynamically adjust the number of sampled rows. For a large table with more than one million rows, the number of sampled rows is dynamically adjusted during statistics collection to improve statistics accuracy.

dn:max_io_capacity

Specifies the maximum I/O per second for the background writer process to flush pages in batches. Set this parameter based on the service scenario and the disk I/O capability. If the RTO is short or the data volume is many times that of the shared memory and the service access data volume is random, the value of this parameter cannot be too small. A small value of max_io_capacity reduces the number of pages flushed by the background writer process. If a large number of pages are evicted due to service triggering, the services are affected. Unit: KB

dn:max_connections

Specifies the maximum number of concurrent connections to DNs.

log_autovacuum_min_duration

Specifies the interval which should elapse before autovacuum operations are logged. Autovacuum operations equal to or beyond the specified interval will be logged. If it is set to 0, all autovacuum operations will be logged. If it is set to -1, no autovacuum operations will be logged.

audit_system_object

Determines whether to audit the CREATE, DROP, and ALTER operations on GaussDB Kernel database objects. GaussDB Kernel database objects include databases, users, schemas, and tables. You can change the parameter value to audit only the operations on required database objects. During a forcible primary/standby failover, set audit_system_object to the maximum value and audit all DDL objects. If the parameter value is incorrectly changed, DDL audit logs will be lost. Contact technical support to change it.

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. 0 indicates there is no timeout. Unit: second

cn:effective_cache_size

Specifies the size of the disk buffer available to the CN optimizer in a single query. Unit: 8 KB

cn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

cn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a CN session.

datestyle

Specifies the display format for date and time.

dn:effective_cache_size

Specifies the size of the disk buffer available to the DN optimizer in a single query. Unit: 8 KB

dn:enable_hotkeys_collection

Specifies whether to collect statistics on accessed key values in databases.

dn:track_stmt_session_slot

Specifies the maximum number of full or slow SQL statements that can be cached in a DN session.

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots.

enable_stream_operator

Specifies the query optimizer's use of streams. When this parameter is set to off, a large number of logs indicating that the stream plans cannot be pushed down are recorded.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

max_replication_slots

Specifies the number of log replication slots on the primary node.

max_wal_senders

The following processes occupy walsender threads: standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. This parameter specifies the maximum number of walsender threads that can be created.

password_effect_time

Specifies the validity period of the password, in days.

password_lock_time

Specifies the duration for a locked account to be automatically unlocked, in days.

recovery_time_target

Specifies the time for the standby node to write and replay logs, in seconds.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

timezone

Specifies the time zone for displaying and interpreting time stamps.

track_stmt_stat_level

Controls the level of statement execution tracking.

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

cn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

cn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

cn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

cn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

cn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

cn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

cn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

cn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

cn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

cn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

wdr_snapshot_interval

Specifies the interval (in minutes) at which the background thread Snapshot automatically performs snapshot operations on the database monitoring data.

enable_wdr_snapshot

Specifies whether to enable WDR snapshots.

cn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

dn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

cn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

cn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

cn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

audit_system_object

Determines whether to audit the CREATE, DROP, and ALTER operations on GaussDB Kernel database objects. GaussDB Kernel database objects include databases, users, schemas, and tables. You can change the parameter value to audit only the operations on required database objects. During a forcible primary/standby failover, set audit_system_object to the maximum value and audit all DDL objects. If the parameter value is incorrectly changed, DDL audit logs will be lost. Contact technical support to change it.

autoanalyze

Specifies whether to automatically collect statistics on tables that have no statistics when a plan is generated.

autoanalyze_timeout

Specifies the autoanalyze timeout period. If the duration of autoanalyze on a table exceeds the value of autoanalyze_timeout, the autoanalyze operation is automatically canceled. 0 indicates there is no timeout. Unit: second

datestyle

Specifies the display format for date and time.

dn:wal_keep_segments

Specifies the minimum number of transaction log files stored in the pg_xlog directory. Standby nodes obtain the logs from the primary node to perform streaming replication.

enable_seqscan

Specifies whether to enable the optimizer's use of sequential scan plan types. It is impossible to completely suppress sequential scans, but setting this parameter to off allows the optimizer to choose other methods if available.

enable_slot_log

Specifies whether to enable primary/standby synchronization for logical replication slots.

failed_login_attempts

Specifies the maximum number of incorrect password attempts before an account is locked. The account will be automatically unlocked after the time specified in password_lock_time elapses. Only the sysadmin user can set this parameter.

log_min_duration_statement

Specifies the threshold for logging the duration of a completed statement. If a statement runs for a period greater than or equal to the specified value, its duration will be logged. The value -1 disables logging statement durations. If this parameter is set to a small value, the load throughput may be affected. Unit: millisecond

max_replication_slots

Specifies the number of log replication slots on the primary node.

max_wal_senders

The following processes occupy walsender threads: standby DNs connect to primary DNs to obtain physical logs, and logical replication tools connect to primary DNs to obtain logical logs. This parameter specifies the maximum number of walsender threads that can be created.

password_effect_time

Specifies the validity period of the password, in days.

password_lock_time

Specifies the duration for a locked account to be automatically unlocked, in days.

session_timeout

Specifies how long to wait before a server connection is disconnected due to inactivity. The value 0 indicates there is no time limit. Unit: second

timezone

Specifies the time zone for displaying and interpreting time stamps.

update_lockwait_timeout

Specifies the maximum duration that a lock waits for concurrent updates on a row to complete when the concurrent update feature is enabled. If the lock wait time exceeds this value, the system will report an error. Unit: millisecond

wal_level

Specifies the level of information to be written to the WAL. This is a required value and cannot be commented out. Determines how much information is written to the WAL. When this parameter is set to logical, logical logs are extracted and primary key information is recorded in Xlogs.

dn:audit_thread_num

Specifies the number of audit threads. Value range: 1 to 48.

dn:qrw_inlist2join_optmode

Specifies whether to enable inlist-to-join query rewriting.

dn:audit_xid_info

Determines whether to record the transaction IDs of SQL statements in detail_info. 0: The transaction IDs are not recorded. 1: The transaction IDs are recorded.

dn:default_limit_rows

Specifies the estimated number of rows to return by default for generating a generic plan, that is, the default value for the LIMIT clause. If this parameter is set to a negative number, the value is converted to a percentage, for example, -5 is equivalent to 5%, indicating that 5% of the total rows will be returned.

dn:audit_dml_state_select

Determines whether to audit the SELECT operation.

dn:audit_dml_state

Determines whether to audit the INSERT, UPDATE, and DELETE operations on a specific table. 0: These operations are not audited. 1: These operations are audited.

dn:random_page_cost

Specifies the estimated cost for the optimizer to fetch an out-of-sequence disk page.

dn:enable_security_policy

Controls whether unified auditing and dynamic data masking policies are applied.

dn:audit_set_parameter

Determines whether to audit the SET operation. 0: The SET operation is not audited. 1: The SET operation is audited.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:vacuum_defer_cleanup_age

Specifies the number of transactions used by VACUUM.

dn:enable_pbe_optimization

Specifies whether the optimizer optimizes the query plan for statements executed in Parse Bind Execute (PBE) mode.

wdr_snapshot_interval

Specifies the interval (in minutes) at which the background thread Snapshot automatically performs snapshot operations on the database monitoring data.

undo_retention_time

Specifies how long undo logs are kept, in seconds. This parameter is only used for flashback query. Note: 1. The undo space of the local disk increases. 2. In subsequent incremental backups, the size of the backup set increases, because extra undo content is retained.

track_stmt_stat_level

Controls the level of statement execution tracking.

enable_wdr_snapshot

Specifies whether to enable WDR snapshots.

dn:max_standby_archive_delay

Specifies the wait period (in milliseconds) before queries on a standby node are canceled when the queries conflict with WAL processing and archiving in hot standby mode.

dn:max_standby_streaming_delay

Specifies how long a standby node waits before canceling queries, in milliseconds.

dn:recovery_max_workers

Specifies the number of concurrent replayer threads.

dn:recovery_time_target

Specifies the time for the standby node to write and replay logs. Unit: second

dn:local_syscache_threshold

Specifies the size of system catalog cache in a session. Unit: KB

cms:datastorage_threshold_value_check

Specifies the disk usage threshold to put a database node into read-only mode. If the disk usage of a data directory exceeds this threshold, the database node is automatically changed to read-only. Unit: percentage (%)

wdr_snapshot_retention_days

Specifies how many days database monitoring snapshots are saved for.

log_autovacuum_min_duration

Specifies the interval which should elapse before autovacuum operations are logged. Autovacuum operations equal to or beyond the specified interval will be logged. If it is set to 0, all autovacuum operations will be logged. If it is set to -1, no autovacuum operations will be logged.

dn:max_connections

Specifies the maximum number of concurrent connections to DNs.