pt-table-checksum reports diffs not found with pt-table-sync
Bug #1059680 reported by
Mike Benshoof
This bug affects 3 people
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Percona Toolkit moved to https://jira.percona.com/projects/PT |
Invalid
|
Undecided
|
Unassigned | ||
Bug Description
Running pt-table-checksum against the master repeatedly shows different chunks out of sync on the slave that are not found by pt-table-sync.
Details:
- Table has timestamp column with ON UPDATE CURRENT_TIMESTAMP
- Updates to table are done via an explicit transaction
- Max timestamp from the ON UPDATE field generally matches the ts column from the checksum table (give or take a second) for the out of sync chunks
- tx_isolation is REPEATABLE-READ on master/slave
| tags: | added: pt-table-checksum tx-isolation wrong-output |
| Changed in percona-toolkit: | |
| status: | New → Triaged |
| importance: | Undecided → High |
| assignee: | nobody → Daniel Nichter (daniel-nichter) |
| tags: | added: percona-26500 |
Revision history for this message
| Daniel Nichter (daniel-nichter) wrote | #1 |
Revision history for this message
| Daniel Nichter (daniel-nichter) wrote | #2 |
Revision history for this message
| Daniel Nichter (daniel-nichter) wrote | #3 |
| Changed in percona-toolkit: | |
| status: | Triaged → Invalid |
| assignee: | Daniel Nichter (daniel-nichter) → nobody |
| importance: | High → Undecided |
Revision history for this message
| Shahriyar Rzayev (rzayev-sehriyar) wrote | #4 |
To post a comment you must log in.
Michael, have you been able to produce a test case for this? Or, was I supposed to to something and forgot? :-)