sql_error_code = 01000 WARNING: you don't own a lock of type ShareLock about oban HOT 4 CLOSED

The warning is coming from a producer attempting to unlock jobs that weren't locked by the current session. It is noisy and annoying, but can safely be ignored.

This is a shortcoming of using advisory locks and is among the primary reasons that I'm investigating an alternative tracking mechanism for executing jobs.

from oban.

I currently get thousands of lines of WARNING: you don't own a lock of type ShareLock per second, which leads to the log file growing rapidly to unhealthy sizes.

I can reproduce this on my local machine and it also happens on staging, where the disk ran out of space. Is there anything I can do to mitigate this besides disabling logs or reducing log retention?

from oban.

There are changes in master which eliminate the use of advisory locks entirely. I’ve been running master for the past week without any issue, but I haven’t released 0.8 yet.

Feel free to test master on your staging server and report back. I’d really appreciate any feedback.

If that isn’t an option and you’d prefer to wait, you can reduce your Ecto pool size which will reduce the number of warnings.

from oban.

I had an issue with the migration (#61) but apart from that it's running smoothly on my local machine so far 👍🏻

from oban.