Comment by raimue
7 hours ago
You could just wrap cronic around the command and immediately receive the full output per mail when it fails. I don't see how a heartbeat alone without logs would help to identify temporary failures.
Your service could accomplish something similar if it had such a wrapper to report both success and failures with logs to a remote server. That would take away the need to run a local MTA, while also detecting with the heartbeat whether the job ran at all.
so cronic combined with a healthcheck, immediate failure alerting and integration for e.g. graylog or fluentd - would be a killer :-)
Thanks for the suggestion!
There's currently no explicit "fail" ping, as that's not the main use case I care about. It's valid, though, and I've thought about adding it.
Being able to pipe logs to curl would also be useful. I'm concerned that people might accidentally send me private/sensitive data, but I'll consider it.
> I don't see how a heartbeat alone without logs would help to identify temporary failures.
It does help. Sometimes jobs unexpectedly don't run at all.
A temporary failure means that the job failed only once and then starts running again. If you want to investigate why it failed, you will need the output of that particular failed run.