pt-table-checksum Illegal division by zero at line 7950
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Percona Toolkit moved to https://jira.percona.com/projects/PT |
Fix Released
|
Medium
|
Brian Fraser |
Bug Description
While running pt-table-checksum on a machine with singe-master, single-slave setup, I *sometimes* run into a bunch of these errors:
Error checksumming table scheme_
11-06T17:47:56 1 0 141 1 0 0.004 scheme_
In a test/lab setup, copied over the whole database, I was unable to reproduce this. It seems the production enviroment with older/slower hardware with more load as a result triggers this. Re-running on the same server doesn't always trigger it, so it seems. Because it is usually a meaningless error for the sysadmin and it will be checksummed in the next run anyway, it generates a lot of unwanted noise in the cron output.
Installation details:
Ubuntu 12.04.1 running plain MySQL server 5.5.28-
The simplest table affected here:
CREATE TABLE `table_name` (
`col1` int(10) unsigned NOT NULL DEFAULT '0',
`col2` int(10) unsigned NOT NULL DEFAULT '0',
KEY `key1` (`col1`),
KEY `key2` (`col2`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1 ROW_FORMAT=DYNAMIC
Related branches
- Daniel Nichter: Approve
-
Diff: 77 lines (+40/-5)2 files modifiedbin/pt-table-checksum (+6/-5)
t/pt-table-checksum/bugs.t (+34/-0)
tags: | added: crash pt-table-checksum |
Changed in percona-toolkit: | |
assignee: | nobody → Brian Fraser (fraserbn) |
status: | Triaged → In Progress |
Changed in percona-toolkit: | |
milestone: | none → 2.1.7 |
Changed in percona-toolkit: | |
milestone: | 2.1.7 → 2.1.8 |
Changed in percona-toolkit: | |
importance: | Undecided → Medium |
Changed in percona-toolkit: | |
status: | In Progress → Fix Committed |
summary: |
- Illegal division by zero at pt-table-checksum line 7950 + pt-table-checksum Illegal division by zero at line 7950 |
Changed in percona-toolkit: | |
status: | Fix Committed → Fix Released |
While probably caused by something entirely different, this and the ptc part of https:/ /bugs.launchpad .net/percona- toolkit/ +bug/1050737 hit the same error in the same part of the code: When the time to nibble a table is (reported as) zero, ptc will start dividing by zero.