Hudi System Functions
Hudi system functions are supported only by 8.2.1.100 and later versions.
pg_show_custom_settings()
Description: Queries details about Hudi foreign table parameter settings.
Return type: SETOF record
Example:
1 2 3 4 5 6 7 8 9 10 11 12 |
SELECT * FROM pg_show_custom_settings(); name | setting | unit | category | short_desc | extra_desc | context | vartype | source | min_val | max_val | enumvals | boot_val | reset_val | sourcefile | sourceline ----------------------------------------------------+-------------------+------+--------------------+--------------------------+------------+---------+---------+---------+---------+---------+ ----------+----------+-----------+------------+------------ hoodie.public.hudi_mor_ft.consume.ending.timestamp | 20230404172329544 | | Customized Options | GUC placeholder variable | | user | string | session | | | | | | | hoodie.public.hudi_mor_ft.consume.mode | incremental | | Customized Options | GUC placeholder variable | | user | string | session | | | | | | | hoodie.public.hudi_mor_ft.consume.start.timestamp | 20230404172329543 | | Customized Options | GUC placeholder variable | | user | string | session | | | | | | | (3 rows) |
hudi_get_options(regclass)
Description: Queries the properties of a Hudi foreign table (hoodie.properties). It is represented by a key-value pair.
Return type: SETOF record
Example:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 |
select * from hudi_get_options('public.hudi_mor_ft'); key | value -------------------------------------------------+--------------------------------------------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -------------------------------------------------------- hoodie.table.precombine.field | col_int hoodie.datasource.write.drop.partition.columns | false hoodie.table.partition.fields | hoodie.table.type | MERGE_ON_READ hoodie.archivelog.folder | archived hoodie.compaction.payload.class | org.apache.hudi.common.model.OverwriteWithLatestAvroPayload hoodie.timeline.layout.version | 1 hoodie.table.version | 4 hoodie.table.recordkey.fields | col_bigint hoodie.database.name | default hoodie.datasource.write.partitionpath.urlencode | false hoodie.table.name | lt_test_mor_014 hoodie.table.keygenerator.class | org.apache.hudi.keygen.ComplexKeyGenerator hoodie.datasource.write.hive_style_partitioning | true hoodie.table.create.schema | {"type"\:"record","name"\:"lt_test_mor_014_record","namespace"\:"hoodie.lt_test_mor_014","fields"\:[{"name"\:"_hoodie_commit_time","type"\:[ "string","null"]},{"name"\:"_hoodie_commit_seqno","type"\:["string","null"]},{"name"\:"_hoodie_record_key","type"\:["string","null"]},{"name"\:"_hoodie_partition_path","type"\:["string","null "]},{"name"\:"_hoodie_file_name","type"\:["string","null"]},{"name"\:"col_bigint","type"\:["long","null"]},{"name"\:"col_int","type"\:["int","null"]},{"name"\:"col_text","type"\:["string","nu ll"]},{"name"\:"col_text2","type"\:["string","null"]}]} hoodie.table.checksum | 515660817 (16 rows) |
hudi_get_max_commit(regclass)
Description: This function retrieves the most recent commit timestamp and data write time for the current Hudi external table.
Return type: record
Example:
1 2 3 4 5 |
SELECT * FROM hudi_get_max_commit('public.hudi_mor_ft'); max_commit | write_time ----------------+------------------------ 20221207141822 | 2022-12-07 14:18:30+08 (1 row) |
hudi_get_commit(regclass, cstring, int)
Description: This function retrieves the timestamp and commit data write time of the current Hudi foreign table, starting from the specified commit up to the Nth commit. If the Nth commit does not exist, the latest commit and the corresponding data write time are returned. This function is supported only by 9.1.0.100 and later versions.
Return type: record
Example:
1 2 3 4 5 |
SELECT * FROM hudi_get_commit('public.hudi_mor_ft', '20230329174744657', 3); end_commit | write_time -------------------+------------------------ 20230329174808908 | 2023-08-31 15:43:08+08 (1 row) |
hudi_sync_task_submit(regclass, regclass)
Description: Submits an automatic Hudi synchronization task. The first input parameter is the synchronization target table, and the second input parameter is the Hudi foreign table. If the task is submitted successfully, the task ID is returned.
Return type: text
- The synchronization target table must contain a primary key, which must be the same as the value of hudi recordkey.
- If the hudi table contains the precombine field, the synchronization target table must also contain the corresponding field.
- If the synchronization target table contains only the primary key (no other fields except the primary key), the synchronization task cannot be submitted normally.
- The user must have the INSERT and UPDATE permissions on the target table and the SELECT permission on the Hudi foreign table. Otherwise, the synchronization task cannot be submitted.
Example:
1 2 3 4 5 |
SELECT hudi_sync_task_submit('public.hudi_sync_i','public.hudi_mor_ft'); hudi_sync_task_submit -------------------------------------- 6465efe2-3ea1-0b00-dde5-b57dfb30fffe (1 row) |
hudi_sync_task_submit(regclass, regclass, interval)
Description: The function is the same as that of hudi_sync_task_submit(regclass, regclass). The difference is that, in this function, you can specify an input parameter of the interval type to specify the task scheduling period. The value ranges from 5 seconds to 24 hours. If the task is submitted successfully, the task ID is returned. This function is supported only by 8.2.1.300 and later versions.
Return type: text
- The synchronization target table must contain a primary key, which must be the same as the value of hudi recordkey.
- If the hudi table contains the precombine field, the synchronization target table must also contain the corresponding field.
- If the synchronization target table contains only the primary key (no other fields except the primary key), the synchronization task cannot be submitted normally.
- The user must have the INSERT and UPDATE permissions on the target table and the SELECT permission on the Hudi foreign table. Otherwise, the synchronization task cannot be submitted.
Example:
1 2 3 4 5 |
SELECT hudi_sync_task_submit('public.hudi_sync_i','public.hudi_mor_ft','1 hour'); hudi_sync_task_submit -------------------------------------- 6465efe2-3ea1-0b00-dde5-b57dfb30fffe (1 row) |
hudi_sync_task_submit(regclass, regclass, text, text)
Description: The function is the same as that of hudi_sync_task_submit(regclass, regclass). The difference is that you can specify two additional text input parameters to indicate the fields that you expect to be synchronized. The fields are separated by commas (,). Quotation marks and escape characters can be parsed. The number and sequence of fields in the two text parameters must be the same, because the synchronization will be performed based on the mapping between the fields in the two text parameters. If the task is submitted successfully, the task ID is returned.
Return type: text
- The synchronization target table must contain a primary key, which must be the same as the value of hudi recordkey.
- If the hudi table contains the precombine field, the synchronization target table must also contain the corresponding field.
- If the synchronization target table contains only the primary key (no other fields except the primary key), the synchronization task cannot be submitted normally.
- The user must have the INSERT and UPDATE permissions on the target table and the SELECT permission on the Hudi foreign table. Otherwise, the synchronization task cannot be submitted.
Example:
1 2 3 4 5 |
SELECT hudi_sync_task_submit('public.hudi_sync_i','public.hudi_mor_ft','_hoodie_commit_time, col_bigint, col_text', '_hoodie_commit_time, col_bigint, col_text'); hudi_sync_task_submit -------------------------------------- 646610bc-cdd1-0d00-d07d-b57e89a0fffe (1 row) |
hudi_sync_task_submit(regclass, regclass, text, text, interval)
Description: The function is the same as that of hudi_sync_task_submit(regclass, regclass, text, text). The difference is that you can specify an additional input parameter of the interval type to specify the task scheduling period. The value ranges from 5 seconds to 24 hours. This function is supported only by 8.2.1.300 and later versions.
Return type: text
- The synchronization target table must contain a primary key, which must be the same as the value of hudi recordkey.
- If the hudi table contains the precombine field, the synchronization target table must also contain the corresponding field.
- If the synchronization target table contains only the primary key (no other fields except the primary key), the synchronization task cannot be submitted normally.
- The user must have the INSERT and UPDATE permissions on the target table and the SELECT permission on the Hudi foreign table. Otherwise, the synchronization task cannot be submitted.
Example:
1 2 3 4 5 |
SELECT hudi_sync_task_submit('public.hudi_sync_i','public.hudi_mor_ft','_hoodie_commit_time, col_bigint, col_text', '_hoodie_commit_time, col_bigint, col_text', '10 minute 30second'); hudi_sync_task_submit -------------------------------------- 646610bc-cdd1-0d00-d07d-b57e89a0fffe (1 row) |
hudi_show_sync_state()
Description: Obtains the synchronization status of the Hudi automatic synchronization task.
Return type: SETOF record
Example:
1 2 3 4 5 |
SELECT * FROM hudi_show_sync_state(); target_tbl | source_ftbl | payload_type | precombine_key | latest_commit --------------------+--------------------+-------------------------------------------------------------+----------------+------------------- public.hudi_sync_i | public.hudi_mor_ft | org.apache.hudi.common.model.OverwriteWithLatestAvroPayload | col_int | 20230511114021573 (1 row) |
hudi_sync(regclass, regclass)
Description: Stored procedure, which is invoked by the Hudi automatic synchronization task. Tasks submitted using pg_catalog.hudi_sync_task_submit(regclass, regclass) will execute the stored procedure. If the command is executed successfully, the number of synchronized rows and timestamp are displayed.
Return type: text
Example:
1 2 3 4 5 6 7 |
CALL hudi_sync('public.hudi_sync_i', 'public.hudi_mor_ft'); NOTICE: execute full sync CONTEXT: PL/pgSQL function hudi_sync(regclass,regclass) line 11 at RETURN hudi_sync -------------------------------------- sync 1 rows up to 20230511114021573. (1 row) |
hudi_sync_custom(regclass, regclass, text)
Description: A stored procedure, which is the entry for invoking the Hudi automatic synchronization task. Users can customize the mapping between the fields in the target table and those in the data source table. Tasks submitted using pg_catalog.hudi_sync_task_submit(regclass, regclass, text, text) will execute the stored procedure. In the preceding information, text is a JSON string, indicating the synchronization mapping between two table fields. If the command is executed successfully, the number of synchronized rows and timestamp are displayed.
Return type: text
Example:
1 2 3 4 5 6 7 |
CALL hudi_sync_custom('public.hudi_sync_i', 'public.hudi_mor_ft', '{"_hoodie_commit_time" : "_hoodie_commit_time", "col_bigint" : "col_bigint", "col_text" : "col_text"}'); NOTICE: execute full sync CONTEXT: PL/pgSQL function hudi_sync_custom(regclass,regclass,text) line 14 at RETURN hudi_sync_custom -------------------------------------- sync 1 rows up to 20230511114021573. (1 row) |
hudi_set_sync_commit(regclass, regclass, text)
Description: Sets the start timestamp of the first synchronization of the Hudi automatic synchronization task to prevent resynchronization. The first parameter is the synchronization target table, the second parameter is the Hudi foreign table, and the third parameter is the expected synchronization start point. This function must be used before a synchronization task is submitted. This function is supported only by 8.2.1.210 and later versions.
Return type: text
Example:
1 2 3 4 5 6 7 |
select hudi_set_sync_commit('public.hudi_sync_i', 'public.hudi_mor_ft', '20230511114021573'); NOTICE: set sync commit successfully, the next synchronization will start from 20230511114021573 CONTEXT: referenced column: hudi_set_sync_commit hudi_set_sync_commit ---------------------- 20230511114021573 (1 row) |
You must have the INSERT and UPDATE permissions on the target table and the SELECT permission on the Hudi foreign table. Otherwise, you cannot set the synchronization progress.
hudi_set_sync_commit(text, text)
Description: Sets the start timestamp of the next synchronization of a Hudi automatic synchronization task. You can use it to sync historical data again or to skip some data. The first parameter is the task ID, and the second parameter is the expected start time of the next synchronization. This function can be used only after a synchronization task is submitted. Before using this function, you need to pause the task. This function is supported only by 8.2.1.210 and later versions.
Return type: text
Example:
1 2 3 4 5 6 7 |
select hudi_set_sync_commit('6524c8e3-aae9-0000-5a14-be8ec000fffe', '20230511114021573'); NOTICE: set sync commit successfully, the next synchronization will start from 20230511114021573 CONTEXT: referenced column: hudi_set_sync_commit hudi_set_sync_commit ---------------------- 20230511114021573 (1 row) |
- Only users who have the permission on the target task can invoke this function.
- Before invoking this function, ensure that the target task is paused and has been successfully executed at least once.
pg_task_show(text)
Description: Queries the information about the current automatic scheduling task. For a Hudi synchronization task, the input parameter should be SQLonHudi.
Return type: SETOF record
Example:
1 2 3 4 5 |
SELECT * FROM pg_task_show('SQLonHudi'); task_id | what | category_id | userid | is_broken | interval | time_cons | start_time | end_time | parameter | task_rank | next_start_time | next_end_time | last_log | failure_times --------------------------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------------+--------+-----------+------------+-----------+-------------------------------+----------+-----------+-----------+-------------------------------+-------------------------------+----------+--------------- 64d257e9-1e9b-0d00-3ce3-7e61b5e0fffe | call pg_catalog.hudi_sync_custom('public.hudi_read_target', 'public.hudi_read101', '{"_hoodie_commit_seqno" : "_hoodie_commit_seqno", "id" : "id", "ts" : "ts", "long_field" : "ts"}'); | SQLonHudi | 10 | f | '00:00:10' | | 2023-08-08 22:58:15.846903+08 | | | 5 | 2023-08-08 22:58:15.846903+08 | 2023-08-08 22:58:24.846903+08 | | 0 (1 row) |
The last_log and failure_times fields are used to record the status of the last task.
- The value of last_log is updated when the task is complete. If the task is successful, the content is cleared. If the task fails, the task failure log is recorded.
- The value of failure_times is updated at the end of the time window. If the task is successful, the value of failure_times is set to 0. If the task fails, the value of failure_times increases by 1. The value of failure_times can be used to infer the time when the first failure occurs.
pg_task_remove(text)
Description: Deletes an automatic scheduling task. The input parameter is a task ID. The function returns the number of deleted tasks.
Return type: integer
Example:
1 2 3 4 5 |
SELECT pg_task_remove('64661705-8ada-0100-d07f-b57e89a0fffe'); pg_task_remove ---------------- 1 (1 row) |
pg_task_pause(text)
Description: Suspend an automatic scheduling task. The input parameter is a task ID. The function returns the number of suspended tasks.
Return type: integer
Example:
1 2 3 4 5 |
SELECT pg_task_pause('64661705-8ada-0100-d07f-b57e89a0fffe'); pg_task_pause --------------- 1 (1 row) |
pg_task_resume(text)
Description: Resumes an automatic scheduling task. The input parameter is the ID of a suspended task. The function returns the number of resumed tasks. This function is supported only by 8.2.1.300 and later versions.
Return type: integer
Example:
1 2 3 4 5 |
SELECT pg_task_resume('64661705-8ada-0100-d07f-b57e89a0fffe'); pg_task_resume ---------------- 1 (1 row) |
pg_task_reset_interval(text, interval)
Description: Modifies the scheduling period of a synchronization task. The first input parameter is task_id, and the second input parameter the scheduling period, of which the value ranges from 5 seconds to 24 hours. The function returns the number of tasks whose periods are modified. This function is supported only by 8.2.1.300 and later versions.
Return type: integer
Example:
1 2 3 4 5 |
select pg_task_reset_interval('64bfd69c-a016-0000-120e-1e802978fffe', '10 hours 30 minutes'); pg_task_reset_interval ------------------------ 1 (1 row) |
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.
