Recently and starting to be consistently, the app pool is not coming back up after a deploy. I think i have traced it to this:
A process serving application pool ‘poolname’ exceeded time limits during shut down. The default for shutdown timeout is 90 seconds.
I think the deployment is occurring faster than the app-pool shutdown. So the corresponding startup is failing (though to my surprise I’m struggling to find that startup command in the log).
So I have a couple of questions here:
- Is Octopus responsible for restarting the app-pool, and if so when/where does it do it. If not is octopus really just doing a recycle and counting on iis to bring the pool back up?
- Is the shutdown really necessary? While i was never in the habbit of deleting whole directories like octopus is, I also never bothered shutting down a pool to deploy new web files.
- Is there something we can do to have Octopus wait for pool status changes? For example the deployment finishes in 38 seconds, but the pool shutdown doesn’t occur for 90 seconds. It’s pretty clear that Octopus continues along irrespective of pool status.
We are looking what is holding the pool from shutdown separately. But we’d like to stop babysitting a deploy to make sure the pool restarts.