On Octopus 3.4.4 we’re encountering an issue currently where the server restarts randomly at intermittent periods. Investigation initially led me to believe that the issue occurred only when running a deployment but looking at our logs show it also happens at random periods overnight. What’s strange is that there seems to be no errors or exceptions thrown before the “Stopping the windows service” message. I also have noticed that the logs report every instance before a restart occurs that the database is upgrading.
The instance of Octopus we’re running has no tentacles and performs all deployments from the Octopus Server itself.
I’ve attached a sample OctopusServer log.
If there is any more info you need please let me know!
OctopusServer.txt (206 KB)
Thanks for getting in touch! The Octopus service itself will not shutdown or restart unless it is told to. Based off your logs giving the following message:
Stopping the Windows Service shows me that there is a process or activity that is telling the service to stop.
Generally when I have seen this from customers it will come from a third party tool or change management process such as DSC or Chef or anything that checks the state. There could also be a scheduled task or activity that is asking the service to stop and restart. These in the past have been setup to release handles for the NuGet handle leak that occurred in versions of Octopus less than 3.4.
I would suggest that you look at scheduled tasks, or external tools that are monitoring the Octopus Server service/server. Unfortunately we do not log what asked Octopus to shutdown only that it was a controled activity. If it was a crash you would not have this message showing up before the service started again.
The database upgrading that you see is on each start of the service our upgrade scripts check if they need to be run or not.
Let me know if you find anything hopefully I have given enough hints to find the source. Happy hunting.