Issues with post-start stopping

[James Hunt]

Hi Jens,

2014-03-20 17:41 GMT+00:00 Jens Rantil <jens.rantil at tink.se>:

> Hi,
>
> I have an HTTP server daemon process that I start using upstart. When I
> start it I'd like it to wait for at most 60 seconds for an HTTP check to
> return 200. If it doesn't return within 60 I'd like to fail "start myapp",
> otherwise I'd like to return "start myapp" as soon as I get a correct HTTP
> response.
>
> I've been trying to use
>
> post-start script
>   /bin/my-healthcheck --max-wait 60
>   if [ $? -ne 0 ]; then
>     stop
>     exit 1
>   fi
> end script
>
> But "stop; exit 1" doesn't seem to stop my service.
>
Your script is subtly incorrect; Upstart uses '/bin/sh -e' to run all
script sections so if my-healthcheck fails, that post-start script shell
will *immediately* exit 1, and stop will never be called.

See:

- Behaviour of 'set -e' in sh(1).
- http://upstart.ubuntu.com/cookbook/#develop-scripts-using-bin-sh

To handle this, you could do something like:

post-start script
  /bin/my-healthcheck --max-wait 60 || { stop; exit 1; }
end script


>
> Can this be done using upstart? Or will I need to `exec myservice_safe`
> that initiates a check and my main script? Or use events?
>
> Thanks,
> Jens
>
> PS. I've cross posted my question here:
> https://serverfault.com/questions/583426/why-is-this-upstart-script-not-stopping-my-process
>
> --
> upstart-devel mailing list
> upstart-devel at lists.ubuntu.com
> Modify settings or unsubscribe at:
> https://lists.ubuntu.com/mailman/listinfo/upstart-devel



Kind regards,

James.
--
James Hunt
____________________________________
#upstart on freenode
http://upstart.ubuntu.com/cookbook
https://lists.ubuntu.com/mailman/listinfo/upstart-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/upstart-devel/attachments/20140324/bdd1ddd5/attachment.html>