High CPU Utilisation Octopus 3.3.11

Legacy Uncategorised
1

Hi,

We are seeing very high CPU utilisation periodically on our New Octopus deploy server. This usually occurs when the server is idle i.e. not during a deploy and is resolved by restarting the service. However, restarting the service each time we see the utilisation spike is just masking whatever the underlying problem is.

Whenever we’ve noticed the issue, we’ve tried to troubleshoot using the usual tools: Procmon, ProcExp. Perfmon, ETW logs, Server logs but have been unable to find what exactly is happening within the Ocotpus.Server.Exe process to cause the excessive CPU.

The machine we are hosting the service on is a VMWare VM with 2vCPU (Xeon E5-2697 @ 2.6GHz) and 4GB memory. We’re hosting about 50 deployment projects on the machine and the underlying ESX infrastructure is not overly loaded (we’re not seeing excessive cpu ready count)

Could you help me with some strategies and or tips to try to work out what is happening within the service when we see the utilisation spike?

Thanks,

2

Hi,

Thanks for reaching out. in 3.3.15 we added a fix that might help with the high CPU usage, along with an improvement for the logging in these scenarios.

Could you please update to the latest version (3.3.17 at the moment) and let us know if that helps? If you find it doesn’t, please send us your Octopus server logs (C:\Octopus\logs on a default install).

Thanks,
Dalmiro.

3

Hi, I’ve upgraded my server to 3.3.18 and have set up an alert to tell me if / when we get excessive CPU.

I’ll keep it running for a few days and let you know if the problem persists.