Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: ddl卡住，show执行不出来
|
username: xie123
[TiDB Usage Environment] Production Environment / Test / Poc
[TiDB Version]
V6.5.5
[Reproduction Path] What operations were performed when the issue occurred
[Encountered Issues: Issue Phenomenon and Impact]
- Dropping a table gets stuck
- ADMIN SHOW DDL; shows a DDL SQL that has been stuck for a long time (saw two entries, first add index running, then drop table)
|404 | 2be432ce-525a-49df-8e46-f53c1f10888d |xx:4000 |
ID:528, Type:drop table, State:queueing, SchemaState:public, SchemaID:90, TableID:444, RowCount:0, ArgLen:0, start time: 2024-03-06 16:29:56.586 +0800 CST, Err:<nil>, ErrCount:0, SnapshotVersion:0
ID:526, Type:add index, State:running, SchemaState:write reorganization, SchemaID:90, TableID:444, RowCount:0, ArgLen:0, start time: 2024-03-06 16:12:13.186 +0800 CST, Err:<nil>, ErrCount:0, SnapshotVersion:448193549109821543, UniqueWarnings:0
| 3841a51f-6760-4dc0-a260-8ce125d06f6f
| DROP TABLE IF EXISTS `table_name`
/* ApplicationName=DataGrip 2023.3.4 */ CREATE INDEX xx ON table_name (xx, xx)
- ADMIN SHOW DDL JOBS; reports an error
ERROR 1105 (HY000): tikv aborts txn: Error(InvalidKeyRangeMode { cmd: scan, storage_api_version: V2, range: (Some("6D44444C4A6F6248FF69FF73746F727900FF0000FC0000000000FF0000690000000000FA"), None) })
- Find the job ID related to this table in all TiDB logs and cancel the job
ADMIN CANCEL DDL JOBS 521,522,523,524,525,526,528,529,533,534,535,536,537,538,541,542,543;
±-------±---------------------------------------+
| JOB_ID | RESULT |
±-------±---------------------------------------+
| 521 | error: [ddl:8224]DDL Job:521 not found |
| 522 | error: [ddl:8224]DDL Job:522 not found |
| 523 | error: [ddl:8224]DDL Job:523 not found |
| 524 | error: [ddl:8224]DDL Job:524 not found |
| 525 | error: [ddl:8224]DDL Job:525 not found |
| 526 | successful |
| 528 | successful |
| 529 | successful |
| 533 | successful |
| 534 | successful |
| 535 | successful |
| 536 | successful |
| 537 | successful |
| 538 | successful |
| 541 | successful |
| 542 | successful |
| 543 | successful |
±-------±---------------------------------------+
ps: Previously, I didn't notice there were two jobs, canceled the drop table DDL job, the stuck add index is still running, rename is stuck
5. After canceling all the above jobs, continue to rename this table, succeeded after about 8 minutes
Seeking help:
- Want to know why adding an index with only 2 rows of data got stuck
- Is ADMIN SHOW DDL JOBS stuck because of storage_api_version v2, can it be rolled back?
[Resource Configuration] Go to TiDB Dashboard - Cluster Info - Hosts and take a screenshot of this page
[Attachments: Screenshots/Logs/Monitoring]
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: 
|
username: