I’ve had issues where a deploy becomes hung during the weekend and ends up running 4 days, buried in the task log, before anyone becomes aware of it. There’s no indication of anything wrong, the deploy simply stalls but seems as if it’s still processing fine. It would be nice if Octopus could send email alerts if any task runs over a configurable amount of time, not just when a deploy fails.
Your wish led to some internal discussion, and we agree, this would certainly be useful.
If we were to implement this, one implementation we have previously discussed would be a timeout value which could be set on each step, and the step would then fail if the timeout was exceeded. This would allow to configure a following step (in your case a Send an Email step, but it could be anything) which would only execute on failure of the previous step.
Would this suit your scenario? Do you consider the step failed if it exceeds the timeout?
And thanks for the feedback. It’s always very welcome
It would be useful to set a timeout on each step, definitely. I also think it would be useful to set a timeout for the total duration of the deployment plan, perhaps something configurable in a deployment project’s settings page, so it catches a hung deploy on any step.