Note
This view is available starting with V4.0.0.
Purpose
The GV$OB_TABLET_COMPACTION_HISTORY view displays the history of tablet-level major compactions.
Columns
Column |
Type |
Nullable |
Description |
|---|---|---|---|
| SVR_IP | varchar(46) | NO | The IP address of the server. |
| SVR_PORT | bigint(20) | NO | The port number of the server. |
| TENANT_ID | bigint(20) | NO | The tenant ID. |
| LS_ID | bigint(20) | NO | The log stream ID. |
| TABLET_ID | bigint(20) | NO | The tablet ID. |
| TYPE | varchar(64) | NO | The type of compaction. Valid values:
|
| COMPACTION_SCN | bigint(20) unsigned | NO | The compaction version. |
| START_TIME | timestamp(6) | NO | The start time. |
| FINISH_TIME | timestamp(6) | NO | The end time. |
| TASK_ID | varchar(64) | NO | The trace ID of the task. |
| OCCUPY_SIZE | bigint(20) | NO | The size of the data. |
| MACRO_BLOCK_COUNT | bigint(20) | NO | The number of macroblocks. |
| MULTIPLEXED_MACRO_BLOCK_COUNT | bigint(20) | NO | The number of reused macroblocks. |
| NEW_MICRO_COUNT_IN_NEW_MACRO | bigint(20) | NO | The number of new microblocks in new macroblocks. |
| MULTIPLEXED_MICRO_COUNT_IN_NEW_MACRO | bigint(20) | NO | The number of reused microblocks in new macroblocks. |
| TOTAL_ROW_COUNT | bigint(20) | NO | The total number of rows. |
| INCREMENTAL_ROW_COUNT | bigint(20) | NO | The number of new rows. |
| COMPRESSION_RATIO | double | NO | The compression ratio of new data. The compression ratio is the ratio of the size of compressed data to that of the original data in new macroblocks. |
| NEW_FLUSH_DATA_RATE | bigint(20) | NO | The output speed of new data, in KB/s. |
| PROGRESSIVE_COMPACTION_ROUND | bigint(20) | NO | The current round of progressive compaction. If the compaction is a full compaction, this column is -1. |
| PROGRESSIVE_COMPACTION_NUM | bigint(20) | NO | The total number of rounds of progressive compaction. |
| PARALLEL_DEGREE | bigint(20) | NO | The parallelism. |
| PARALLEL_INFO | varchar(512) | NO | The information about parallel tasks. It displays the statistics of the amount of data scanned, the running time, and the amount of data output by parallel tasks (minimum, maximum, and average). |
| PARTICIPANT_TABLE | varchar(512) | NO | The information about the tables involved in the current compaction. |
| MACRO_ID_LIST | varchar(256) | NO | The list of output macroblocks. If the list is too long, it will not be displayed. |
| COMMENTS | varchar(1024) | NO | The remarks. It displays the problems encountered during the previous scheduling or execution. |
| START_CG_ID | bigint(20) | NO | The start ID of the column group
NoteThis column is available starting with V4.3.0. |
| END_CG_ID | bigint(20) | NO | The end ID of the column group
NoteThis column is available starting with V4.3.0. |
| KEPT_SNAPSHOT | varchar(128) | NO | The information about the multi-version snapshot retained by the current compaction
NoteThis column is available starting with V4.3.0. |
| MERGE_LEVEL | varchar(64) | NO | The information about whether the merge or microblock reuse is used during the current execution
NoteThis column is available starting with V4.3.0. |
| EXEC_MODE | varchar(64) | NO | The mode of the compaction trace. Valid values:
NoteThis column is available starting with V4.3.3. |
| IS_FULL_MERGE | varchar(5) | NO | Indicates whether the compaction is a full compaction.
NoteThis column is available starting with V4.3.3. |
| IO_COST_TIME_PERCENTAGE | bigint(20) | NO | The percentage of the I/O time in the major compaction time.
NoteThis column is available starting with V4.3.3. |
| MERGE_REASON | varchar(32) | NO | The reason for initiating the major compaction. Valid values:
NoteThis column is available starting with V4.3.3. |
| BASE_MAJOR_STATUS | varchar(64) | NO | The type of the major SSTable before the major compaction:
NoteThis column is available starting with V4.3.3. |
| CO_MERGE_TYPE | varchar(64) | NO | The type of the column store compaction:
NoteThis column is available starting with V4.3.3. |
| MDS_FILTER_INFO | varchar(1024) | NO | The MDS filter information used for the major compaction.
Note
|
| EXECUTE_TIME | bigint(20) | NO | The CPU time consumed by the major compaction task, in microseconds. |
| FILTER_ROW_COUNT | bigint(20) | NO | The number of rows that are recycled by using the major compaction.
NoteThis column is available starting with V4.6.0. |
Sample query
Query the tablet-level compaction history information on all OBServer nodes in the user tenant.
obclient(root@mysql001)[oceanbase]> SELECT * FROM oceanbase.GV$OB_TABLET_COMPACTION_HISTORY LIMIT 1\G
The query result is as follows:
*************************** 1. row ***************************
SVR_IP: 172.xx.xxx.xxx
SVR_PORT: 2882
TENANT_ID: 1002
LS_ID: 1
TABLET_ID: 60379
TYPE: MAJOR_MERGE
COMPACTION_SCN: 1770746402543355000
START_TIME: 2026-02-11 02:03:26.743059
FINISH_TIME: 2026-02-11 02:03:26.751584
TASK_ID: YB42AC1E87C4-00064A6F20227C22-0-0
OCCUPY_SIZE: 0
MACRO_BLOCK_COUNT: 0
MULTIPLEXED_MACRO_BLOCK_COUNT: 0
NEW_MICRO_COUNT_IN_NEW_MACRO: 0
MULTIPLEXED_MICRO_COUNT_IN_NEW_MACRO: 0
TOTAL_ROW_COUNT: 0
INCREMENTAL_ROW_COUNT: 0
COMPRESSION_RATIO: 1
NEW_FLUSH_DATA_RATE: 0
PROGRESSIVE_COMPACTION_ROUND: 1
PROGRESSIVE_COMPACTION_NUM: 0
PARALLEL_DEGREE: 1
PARALLEL_INFO: -
PARTICIPANT_TABLE: table_cnt=1,[MAJOR]snapshot_version=1;
MACRO_ID_LIST:
COMMENTS:
START_CG_ID: 0
END_CG_ID: 0
KEPT_SNAPSHOT: {type:"LS_RESERVED", snapshot:1770744792640865000}
MERGE_LEVEL: MICRO_BLOCK_LEVEL
EXEC_MODE: EXEC_MODE_LOCAL
IS_FULL_MERGE: FALSE
IO_COST_TIME_PERCENTAGE: 0
MERGE_REASON: TENANT_MAJOR
BASE_MAJOR_STATUS:
CO_MERGE_TYPE:
MDS_FILTER_INFO:
EXECUTE_TIME: 827
FILTER_ROW_COUNT: 0
1 row in set
References
Query the detailed history information about tablet-level major compactions on the current node: V$OB_TABLET_COMPACTION_HISTORY
Query the progress information about tablet-level compactions on all nodes: GV$OB_TABLET_COMPACTION_PROGRESS