My company uses octopus deploy in aws with autoscalling ec2 instances. When an instance comes up a program on the instance runs to register it as a tentacle with appropriate roles and envs.
Multiple times a week we go from our blue env to our green env, which means we spin up another 60 or so instances all at once across multiple octopus roles, and then some combination of about 100 projects ends up getting deployed to each of them through an autodeploy trigger.
For months, this has mostly worked without a problem, but sometime after a recent octopus upgrade (we are currently on 2018.6.2, but I am not sure the exact version that caused the problem, we stay up to date for the most part), the autodeploy triggers seem to be ignored. I will spin up 5 new boxes, they all register themselves automatically as tentacles, but the autodeploy triggers only kicks off on 2 of the new boxes. The other 3 were ignored by octopus. There is nothing in the octopus logs of relevance that I can see. It is acting as if the autodeploy triggers don’t exist.
I don’t really know where to start with regards to collecting diagnostic information to help troubleshoot this, but I am happy to co-operate and send in anything that would be helpful.