pt-slave-delay won't restart after mysql IO error

[Jarrod Cuzens]

More of a feature request than a bug but we run a slave delayed two days in our office from our production system. autossh is used to ensure a tunnel to production exists but there are still occasions when the slave has an issue connecting and we get an IO error 2013 'error reconnecting to master 'repl@localhost:2306' - retry-time: 60 retries: 86400'.

This IO error appears to cause pt-slave-delay to never restart unless we manually issue a "slave start" so we have this in an hourly cron. It would be nice if pt-slave-delay would automatically try to restart regardless of the IO issue.

pt-slave-delay version 2.0.3

2012-02-29T11:18:19 slave stopped at master position mysql-bin.000922/298211578
2012-02-29T11:19:19 slave stopped at master position mysql-bin.000922/298338346
2012-02-29T11:20:19 slave stopped at master position mysql-bin.000922/298452192
2012-02-29T11:21:19 slave stopped at master position mysql-bin.000922/298589990
2012-02-29T11:22:19 slave stopped at master position mysql-bin.000922/298673234
2012-02-29T11:23:19 slave stopped at master position mysql-bin.000922/298771151
2012-02-29T11:24:19 slave stopped at master position mysql-bin.000922/298889517
2012-02-29T11:25:19 slave stopped at master position mysql-bin.000922/298989539
2012-02-29T11:26:19 slave stopped at master position mysql-bin.000922/299186614
2012-02-29T11:27:19 slave stopped at master position mysql-bin.000922/299327914

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PT-944