I received an email at ~10pm EST that our
rp5-cpbtrue-rinkeby-postgres application hit an OOM and died, with the following log:
Out of memory: Killed process 4141 (postgres) total-vm:248432kB, anon-rss:89664kB, file-rss:4kB, shmem-rss:7756kB, UID:1000 pgtables:340kB oom_score_adj:0
It appears to have restarted successfully, which I appreciate.
However, I am not sure what could have caused it - we didn’t have a spike in traffic or such. I am also having trouble finding trails of evidence in the metrics tab, although the monitoring tab shows it.
Any advice on how to track down what could have caused it? Thanks.