As you can see in the attached image we are having some issues with memory usage from the octopus deploy server.
We had been running 2.5.7 until recently but today we finally upgraded to 2.6.5, and was somewhat hoping this memory usage would go down. But seeing as it haven’t I ended up posting here.
What is it exactly that can take up this much memory? We have TeamCity etc running on this machine as well so we can’t really dedicate that much ram to just one process, which is just sitting idle 90% of the day
We do have quite a few old releases ~6000 or so (we haven’t dared set a cleanup policy yet since we have the live environment on the same octopus instance and haven’t gotten around to playing with lifecycles yet), so perhaps that is a factor or is it something else entirely?
Thanks for getting in touch! Octopus 2.x has Raven as an embedded database. Embedded databases will use all the memory they can when they need it but then don’t release the memory after use. The amount of data is a factor for sure, because again it’s an embedded database. You will find that in 3.0 Octopus itself won’t use so much memory. You will also have the ability to have SQL externally or even on another server. Right now there is not a lot you can do to release this memory. Restarting the service would take it back down, but Raven would eat it back up again pretty fast.
3.0 is literally right around the corner. Sorry there is nothing more immediate you can do.
Thank you for letting me know. Then I know. We did fix with the retention policy so we only have about 1k old releases now and then rebooted the machine. Now it has been at a pretty solid ~500mb memory usage. So now I am a bit more satisfied