Version 8.1.0
Release date: May 15, 2021
Cluster version: 8.1.0.100
Feature Changes
Feature |
Description |
---|---|
Enhanced ANALYZE for temporary tables and single-transaction operations. |
Transactions and stored procedures support the ANALYZE command. No error is reported when ANALYZE and ALTER TABLE are executed concurrently. The automatic analyze function supports temporary tables. |
Improved performance of the vectorized executor. |
|
Partition pruning with non-constant partition column filter criteria |
Partition pruning is optimized. Currently, partition pruning can be performed only for constants. Variables that need to be calculated cannot be pruned and need to be optimized. |
Schema-level full backup + Table- and schema-level fine-grained restoration |
Schema-level full backup + Table- and schema-level fine-grained restoration |
Supported third-party user-defined data conversion of GDS. |
Supported third-party user-defined data conversion. |
Optimized RTO. |
RPO = 0 and RTO < 60s in intra-cluster HA scenarios; RTO < 30s in POC scenarios |
SQL on HDFS |
Supported ORC data export using OBS foreign tables. |
Supported XML function parsing. |
Supported XML function parsing. |
Modified distribution columns. |
Modified distribution columns. |
Supported automatic job migration when a CN fails. |
If the current CN is faulty, its jobs can be automatically migrated. |
Supported smooth upgrade of C functions. |
Supported smooth upgrade of C functions. |
Supported Teradata DSQL compatibility of gsql. |
|
Synchronized the time zone with the IANA Time Zone Database (tz). |
Synchronized the time zone with the IANA Time Zone Database (tz). |
SQL Syntax
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
Column-store partitioned tables support the split operation. |
Column-store partitioned tables support the split operation. |
2 |
CREATE TABLE tbl(LIKE tbl2 INCLUDING DROPCOLUMNS ) |
The LIKE DROPCOLUMNS syntax is supported. |
|
3 |
ALTER TABLE tbl DISTRIBUTE BY |
Added the syntax for modifying the distribution column of a table. |
|
4 |
ALTER TABLE set version |
You can switch between the old and new formats of column-store tables. |
|
5 |
CREATE/ALTER FOREIGN TABLE Write Only |
Data can be written in OBS ORC format. |
|
6 |
CREATE/ALTER FOREIGN TABLE Write Only |
Data can be written in OBS HDFS format. |
|
7 |
XML |
The XML data type can be used in tables and functions. |
|
8 |
ALTER VIEW |
alter view viewname REBUILD; |
|
Modify |
9 |
VACUUM FULL |
A deadlock occurs when concurrent VACUUM FULL operations are performed on the same table. |
10 |
DROP TEXT SEARCH CONFIGURATION |
If the text search configuration to be deleted is the current default_text_search_config, an error is reported and the deletion fails. |
|
11 |
CREATE/ALTER ROLE/USER AUTHINFO 'ldap' |
The syntax for specifying the user authentication type as LDAP is modified to remove the 64 character length limit. |
|
12 |
CREATE USER/ROLE with authinfo 'ldap' password disable |
Modified the syntax for creating an LDAP user. |
|
13 |
ALTER USER/ROLE with authinfo 'ldap' password disable |
Modified the syntax for creating an LDAP user. |
|
Delete |
14 |
ALTER TABLEPartitioned tableexchange with unloggTable |
Forbid partitioned table exchange with unlogged tables. |
15 |
Inheritance can be modified using ALTER TABLE ... inherit. |
The optimizer is not adapted to table inheritance. The alter table inherit interface needs to be disabled. |
Keywords
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
xmltable |
The COL_NAME_KEYWORD keyword has been adapted for inspection. |
2 |
xmlnamespaces |
The COL_NAME_KEYWORD keyword has been adapted for inspection. |
System Catalogs
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
dbms_om.gs_wlm_session_info |
New system catalog |
2 |
dbms_om.gs_wlm_operator_info |
New system catalog |
|
Modify |
3 |
pg_jobs |
Same as the old system catalog pg_job, the metadata relationship is associated using OIDs. If the shared system catalog is full, and you cannot add a transaction after deleting one, pg_job will be renamed as pg_job_proc. |
4 |
pg_proc |
The textanycat and anytextcat functions are changed to non-STRICT functions. |
|
5 |
pg_proc |
The database_to_xml, database_to_xmlschema, and database_to_xml_and_xmlschema functions are changed to non-STRICT functions. |
|
6 |
pg_catalog.gs_wlm_session_info |
It is changed to a view. |
|
7 |
pg_catalog.gs_wlm_operator_info |
|
|
8 |
pg_catalog.gs_wlm_ec_operator_info |
It is changed to a view. |
|
9 |
pg_catalog.gs_wlm_ec_operator_info |
|
System Functions
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
pgxc_get_residualfiles |
Unified CN query function of pg_get_residualfiles. |
2 |
pgxc_rm_residualfiles |
Unified CN query function of pg_rm_residualfiles. |
|
3 |
pgxc_verify_residualfiles |
Unified CN query function of pg_verify_residualfiles. |
|
4 |
query_to_xmlschema |
It maps the query content to an XML schema document. |
|
5 |
query_to_xml |
Maps the query result to an XML file. |
|
6 |
query_to_xml_and_xmlschema |
It maps the contents of a query to XML values and schema documents. |
|
7 |
table_to_xmlschema |
It maps the contents of a relational table to an XML schema document. |
|
8 |
table_to_xml |
It maps the contents of the relation table to XML values. |
|
9 |
table_to_xml_and_xmlschema |
It maps the contents of a relational table to XML values and schema documents. |
|
10 |
cursor_to_xmlschema |
It maps the contents of a cursor query to an XML schema document. |
|
11 |
cursor_to_xml |
It maps the contents of a cursor query to an XML document. |
|
12 |
schema_to_xmlschema |
It maps the contents of the entire schema into an XML schema document. |
|
13 |
schema_to_xml |
It maps the contents of the entire schema to an XML document. |
|
14 |
schema_to_xml_and_xmlschema |
It maps the contents of the entire schema to XML values and schema documents. |
|
15 |
database_to_xmlschema |
It maps the contents of the entire database into XML schema documents. |
|
16 |
database_to_xml |
It maps the contents of the entire database to XML documents. |
|
17 |
database_to_xml_and_xmlschema |
It maps the contents of the entire database to XML values and schema documents. |
|
18 |
xmlpi |
It creates an XML processing instruction. |
|
19 |
xmlcomment |
It creates an XML comment with the specified text as the content. |
|
20 |
xmlserialize |
It generates a string from a value of the xml type. |
|
21 |
xmlparse |
It generates a value of the xml type from character data. |
|
22 |
xpath |
It returns an array of XML values corresponding to the node set generated by the XPath expression. |
|
23 |
get_instr_unique_sql_remote_cns |
It obtains the unique sql data on all CNs except the current CN. |
|
24 |
xml_is_well_formed |
It checks whether the text string is a well-formed XML value. |
|
25 |
xml_is_well_formed_content |
It checks whether the text string is well-formed. |
|
26 |
xml_is_well_formed_document |
It checks whether the text string is well-formed. |
|
27 |
xmlconcat |
It concatenates a list of XML values into a single value. |
|
28 |
xmlagg |
It combines a function that joins input values together. |
|
29 |
IS DOCUMENT |
It determines whether the parameter value is a correct XML file. |
|
30 |
IS NOT DOCUMENT |
It determines whether the parameter value is an incorrect XML file. |
|
31 |
xmlexists |
It determines whether the XPath expression in the parameter returns any node. |
|
32 |
xpath_exists |
It determines whether the XPath expression in the parameter returns any node. |
|
33 |
xmlelement |
It produces an XML element with the given name, attributes, and content. |
|
34 |
xmlforest |
It produces an XML forest (sequence) of an element using the given name and content. |
|
35 |
pg_xlog_replay_completion |
You can query the Xlog redo progress of the current DN. |
|
36 |
pg_data_sync_from_dummy_completion |
It displays the progress of transferring the current DN data page from the dummystandby. |
|
37 |
pg_stat_xlog_space |
It displays the Xlog space usage on the current DN. |
|
38 |
pgxc_stat_xlog_space |
It displays the Xlog space usage on all primary DNs. |
|
39 |
xmlroot |
It modifies the attributes of the root node of an XML value. |
|
40 |
pg_get_residualfiles |
It queries the list of all residual files on the current node. |
|
41 |
pg_get_running_jobs |
It queries jobs that are running on the current node. |
|
42 |
pg_is_residualfiles |
It can be used to check whether a specified relfilenode is a residual file in the current database. |
|
43 |
pg_rm_residualfiles |
It deletes files from a specified residual file list. |
|
44 |
pg_verify_residualfiles |
It can verify if the list of remaining files recorded in a specific file are actually residual files. |
|
45 |
table_distribution |
It adds a table_distribution function whose parameter type is regclass. |
|
46 |
XMLTABLE |
The XMLTABLE function can be used to parse XML data. |
|
Modify |
47 |
pgxc_get_senders_catchup_time |
The catchup process information field is added. |
48 |
pg_stat_get_data_senders |
The catchup process information field is added. |
|
49 |
pgxc_get_thread_wait_status |
The num_node_display parameter is added. |
|
50 |
pg_stat_get_status |
The num_node_display parameter is added. |
|
51 |
pg_catalog.pgxc_get_wlm_session_info_bytime |
The first parameter is changed from case-sensitive to case-insensitive. |
System Views
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
pg_job_single |
Job information of a single node. |
2 |
pg_job |
Used for forward compatibility with the pg_job system catalog and distributed pg_job_single collection. |
|
3 |
pgxc_get_stat_all_partitions |
Obtain the number of insert/update/delete/live/dead tuples and dirty page rate of each partition in all partitioned tables. |
|
4 |
gs_view_dependency |
The dependency of the cascading query view is the union of the preceding two functions. |
|
Modify |
5 |
pgxc_get_instr_unique_sql |
Run the get_instr_unique_sql_remote_cns statement to obtain the unique SQL data on other CNs. |
6 |
pg_stat_replication |
Debugged the function. |
|
7 |
pg_get_senders_catchup_time |
The catchup process information field is added. |
|
8 |
pg_catalog.gs_wlm_operator_history |
The nodename field is added. |
|
9 |
pg_catalog.pgxc_wlm_operator_history |
The nodename field is added. |
|
10 |
pg_catalog.pgxc_wlm_operator_info |
The nodename field is added. |
Behavior Changes
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Modify |
1 |
CN retry supports the stored procedure that affects real-time printing. |
The output of the stored procedure is printed in real time. If cn retry occurs, a notice is displayed, indicating that the output message is repeated. If cn retry occurs before printing, no notice is output. |
GUC Parameters
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Add |
1 |
wal_compression_level |
Zlib compression level of the PFI log compression function. The default value is 9. |
2 |
wal_compression |
Whether to enable the PFI log compression function. This function is disabled by default. |
|
3 |
max_xlog_backup_size |
Size of the Xlog backup file. When the size of a backup Xlog file exceeds the value of this parameter, the earliest backup Xlog file is automatically deleted until the size of the backup Xlog file is less than 90% of the value of this parameter. |
|
4 |
max_cache_partition_num |
Maximum number of partitions reserved in the memory during redistribution. If the number of partitions exceeds the value of this parameter, the earliest partition will be written to disks in CU format. |
|
5 |
password_encryption_type |
Encryption type of user passwords. The value 2 is added, indicating that the password is encrypted using SHA256. If the current cluster is upgraded from 8.0.0 or an earlier version, the default value of this parameter is 1 for forward compatibility. The default value is 2 for a newly installed cluster. |
|
6 |
join_num_distinct |
Default distinct value of the join column. The default value is -20. |
|
7 |
cost_model_version |
Whether the cost estimation optimization takes effect. The default value is 1, indicating that the optimization takes effect. |
|
8 |
qual_num_distinct |
Default distinct value of the filter column. The default value is 200. |
|
9 |
behavior_compat_options |
The varray_verification option is added to roll back the new verification. |
|
10 |
behavior_compat_options |
The check_function_conflicts option is added to check whether the IMMUTABLE function has non-IMMUTABLE behavior. |
|
11 |
auto_process_residualfile |
Switch for automatically recording residual files. The default value is true, indicating that the function is enabled. |
|
12 |
default_colversion |
Default format for creating a column-store table. The default format is 1.0. |
|
13 |
enable_partition_dynamic_pruning |
Whether to support dynamic pruning during partition table scanning. Dynamic pruning is enabled by default. |
|
14 |
enable_join_pseudoconst |
Whether to generate a join expression on the join expression that is equal to a constant. Similar to the t1 inner join t2 on t1.a=t2.a where t1.a = 1 scenario, the join expression can be generated based on t1.a=t2.a. (In the earlier version, such join conditions cannot be used for join cond.) |
|
15 |
view_independent |
Whether to enable the view dependency decoupling function. Default value: off |
|
16 |
enable_view_update |
Whether to enable single view update. |
|
Modify |
17 |
enable_index_nestloop |
If the system is upgraded from C80 or an earlier version to the latest patch version, set enable_index_nestloop to off. Enabling the upgrade of enable_index_nestloop from 6.5.0 or 6.5.0 will maintain forward compatibility. The default value of enable_index_nestloop is on. |
18 |
archive_mode |
The default value of the xlog archive switch is changed from off to on. In some scenarios, the performance is slightly affected. In POC scenarios such as performance comparison, you are advised to manually avoid this problem. |
|
19 |
cost_param |
The default value is changed to 16, which corresponds to the optimized cost estimation. |
|
20 |
rewrite_rule |
The partialpush option is removed. |
|
21 |
behavior_compat_options |
Added the strict_concat_functions parameter for forward compatibility of the textanycat and anytextcat functions. |
|
22 |
behavior_compat_options |
The 'strict_text_concat_td' option has been added to change textcat, textanycat, and anytextcat functions to strict functions in TD mode. |
|
23 |
behavior_compat_options |
The strict_text_concat_td option is added to be compatible with the null stitching behavior in T/D mode. |
|
24 |
behavior_compat_options |
The bpchar_text_without_rtrim option is added to be compatible with the style of processing spaces at the end of the bpchar string in TD mode. |
Time Zone
Change Type |
No. |
Name |
Change Description |
---|---|---|---|
Modify |
1 |
Asia/Beijing: The definition of the Beijing time zone is changed to be the same as that of Asia/Shanghai. |
|
2 |
Modified the definition of Asia/Shanghai. |
|
|
3 |
Asia/Urumqi |
After May 1980, the time zone offset is changed from GMT+8 to GMT+6. |
|
4 |
The five time zones defined based on the physical time zone are changed to two time zones. |
|
|
5 |
The initial default time zone may change. |
During database initialization, a time zone that best matches the operating system time zone is used as the default time zone. The time zone definition supported by the database is changed. As a result, the initial default time zone may be changed. |
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot