|
|
There's a bug you can hit with pgAgent after an upgrade (like from
postgres-9.5 to postgres-9.6) that causes your jobs to get stuck in
a "Running" state. Since pgAgent thinks the job is running, it won't
start the job again the next time it's scheduled, even though the
job has actually completed successfully. To fix that issue, we apply
four new patches that add recent improvements from upstream:
* Fix a crash from DBconn:GetLastError().
* Save the stderr output in the jobstep log.
* Add timestamps to the pgAgent logs.
* And finally, fix the "job succeeded" check.
The intermediate patches are not strictly necessary, but they are
helpful, and most importantly, let us apply the upstream patches
cleanly without having to re-roll them.
Package-Manager: Portage-2.3.6, Repoman-2.3.1
|
Thomas Deutschmann
Mikle Kolyada
David Seifert
Michael Orlitzky
GitWeb