23 Sep 2024 03:42 PM - last edited on 24 Sep 2024 12:25 PM by Michal_Gebacki
Hi Folks,
Have you ever observed systemd cpu usage increase after OA install?
I have already raised a support ticket but maybe you had experience on it.
Environement: RHEL 7.9 with IBM Integration Bus.
After OA install systemd cpu consumption from AVG 0,1 % increase to AVG 10% (fluctuate between 0-30%).
Could you please advice? Which OA featrue should be switched off? 😉
Thanks in advance for your advices.
Best regards,
Mizső
23 Sep 2024 07:38 PM
Systemd is a native app, Dynatrace does not inject and does not make any modifications.
Do you see this in processes in Dynatrace or on the host using top / htop tools? Which of the systemd processes?
Does the customer have Nettracer or eBPF Service Discovery features enabled?
23 Sep 2024 08:35 PM - edited 23 Sep 2024 08:59 PM
Hi Julius,
I can see the systemd process on the host view at the process section.
Nettracer is not enabled, but eBPF Service Discovery feature is enabled.
Br, János
24 Sep 2024 07:09 AM
Ah, ok - maybe some other process falls into the the systemd group?
Can you run a process instance snapshot to confirm? Or collect a support archive and check the processesSnapshot.json for details of CPU consumption.
25 Sep 2024 03:37 PM
Hi Julius,
Thanks for the tips. I am gonig to check these ones at the client tomorrow. 😉
Best regards,
János
25 Sep 2024 02:46 PM
Hi @Mizső
For Oneagent service to take effect, a pending process restart would be the reason for the CPU usage spikes.
Regards,
peter.
25 Sep 2024 03:38 PM
Thanks for the tip. I will check it.
Best regards,
Mizső