app/worker.1: production.ERROR: SQLSTATE[25006]: Read only sql transaction: 7 ERROR: cannot execute SELECT FOR UPDATE in a read-only transaction (SQL: select * from "jobs" where "queue" = default and (("reserved_at" is null and "available_at" <= 1545253197) or ("reserved_at" <= 1545253107)) order by "id" asc limit 1 for update)
It happened every 3 seconds over slightly more than a minute.
This is from papertrail on heroku.
I know the queue worker polls the database (I'm using db queue driver) about every 3 seconds, so that's the cause. Simply interested in explanation of what is happening.
Does this mean the (heroku) postgres became temporarily unwritable for some reason?
Any ideas on likely causes / what this means? Have you seen this before on your apps?
Thanks much.
Are you running out of storage? Sometimes write access can be disabled if there is not enough space to write anything.
Other plausible reason are:
being throttled somehow?
trying update statements on a read-only replica (e.g. the entire instance is read-only/is a follower)?
Try manually running SELECT pg_is_in_recovery() and logging the response (in the jobs) to see if when it does throw that error if it's in recovery mode. If it's true, then it's likely connected to a read-only replica server in hot_standby mode.
