Hello Aaron,
you see what monit will use by using "monit -v".
> The more I think about this, the more I think I might consider it a bug one
> way or the other: Maybe I'm missing something, but it seems to me like
> monit should either raise an error when you supply multiple 'every'
> statements (at least for `monit -t`), *or* use the combination of all the
> 'every' statements. Allowing multiple 'every' statements but ignoring some
> of them seems it's a trap for people to fall into, especially since this
> behavior isn't mentioned in the documentation.
You are right, sometimes some more error/warning or info messages are
helpful. But in general the config parser works well from my point of view.
A snippet from the Monit documentation.
> We will address this limitation in a future release and convert the
> scheduler from ...
Ah, I interpreted that section to be referring to "The cron specification does not guarantee when exactly the test will run" in the previous paragraph as the limitation.
Feel free to add a ticket (see
https://bitbucket.org/tildeslash/monit/issues), suggestions are welcome
in general.
Okay, I can certainly add an issue. Is it realistic to put in an issue for allowing/interpreting multiple 'every' statements, or would this issue be for a documentation/`monit -t` change?
It sounds like there isn't currently a way for me to control both the 'cron' and cycle-based timing of a service, is that correct?
-Aaron