Hi,
our dedicated Continua CI server is hosted on VMware ESXi seems to act very strangely. Although nobody used the CI system today and we did no checkin to any registered Subversion respository the CI system regularly utilizes 100% of the CPU core of the virtual machine. I attached a screenshot of the performance monitoring of VMware vSphere web client.
Our environment:
Hypervisor: VMware ESXi 5.5
Operating System: Windows Server 2012 R2 Standard
Continua CI: 1.5.1.49
Continua CI Agents: local agent, two online remote agents, one offline remote agent
Do you have any idea what might cause these CPU peaks?
Regards
Kay Zumbusch
Hi Kay,
Is the server configured with LDAP (or Mixed) authentication mode? Versions of Continua prior to 1.6 ran an LDAP synchronisation task every 30 minutes. This was processor intensive, especially with large domains and has now been rewritten to improve performance. If the synchronisation takes around 10 minutes to run, this would be consistent with the CPU peaks you are seeing. I’m quite sure that we don’t have any other processes which run at this interval.
I note that you are running a old 1.5 version of Continua from December last year. We recommend that you upgrade as you’ll find numerous improvements in performance in updates released since then, including the update to LDAP synchronisation.
If you still see these high CPU peaks after upgrading, first check the Resource Monitor at the peak times to check that it is indeed the Continua service which is causing this. Then enable debug logging, and send the log to support@finalbuilder.com after the server has been running through one of these CPU peaks. Also include details of the timings of these peaks.
Hi Dave,
we indeed connect to our Active Directory using LDAP although our AD is not that large. Thanks for the hint. The update to the most recent version of 1.6 is already planned. I’ll get back to you if the update does not help.
Regards
Kay Zumbusch