Schedule periodic maintenance on acralyzer host

[mikehardy]

Went to clean up the acralyzer instance and it was a bit of a mess - these items need periodic scheduling:

• purge ACRA reports older than X days (30 is the default, 90 seems to fit us better, not sure how to do this via an API or command line)
• mysql needs maintenance to reclaim space post-purge (mysqlcheck -u root -p --auto-repair --optimize --all-databases) - requires a root mysql password but it's empty in our instance 😬 so anyway just hitting return at the password prompt works)

Consequence of not doing this is that the 25GB of available storage fills - that's important to mention.

The disk appears to fill about once every 6 months

[david-allison]

30 days is a little too aggressive for us. Could we up this to 90 when this is implemented

[github-actions[bot]]

Hello 👋, this issue has been opened for more than 3 months with no activity on it. If the issue is still here, please keep in mind that we need community support and help to fix it! Just comment something like still searching for solutions and if you found one, please open a pull request! You have 7 days until this gets closed automatically

[github-actions[bot]]

Hello 👋, this issue has been opened for more than 3 months with no activity on it. If the issue is still here, please keep in mind that we need community support and help to fix it! Just comment something like still searching for solutions and if you found one, please open a pull request! You have 7 days until this gets closed automatically

[github-actions[bot]]

Hello 👋, this issue has been opened for more than 3 months with no activity on it. If the issue is still here, please keep in mind that we need community support and help to fix it! Just comment something like still searching for solutions and if you found one, please open a pull request! You have 7 days until this gets closed automatically

[mikehardy]

There's no API to do this, but now that I know delete of bug table entries cascades, I believe it's possible to write some SQL that does this and runs a nightly purge

rough first thought on sql taking into account that we have a bug -> report -> stacktrace cascade but the date information is only available on the report table

1- delete from bug where no reports meet criteria that reports.bug_id = bug.id and reports.date is newer than 2- delete from reports where date is older than

basic idea being first delete whole data tree (bug is root) where no reports are new enough, so we don't orphan bug table entries, then after making sure we won't create orphan bug entries, clean up reports table in general

stacktrace table entries will cascade delete from reports, so those won't orphan