Hi,
since updating to SFOS 17.1.2 MR-2 a customers XG HA-cluster in A/P mode is ~90% cpu load.
Its a small XG125 setup.
Anybody seen this before?:
This thread was automatically locked due to age.
confirming same high cpu problem here on XG330 rev2 HA pairs after updating from 17.0.6 MR-6 to 17.1.2 MR-2. "gethainfo" and "worker" processes seem to use the most. based on some specific issues logged during the sfos updates in our environment that might not apply to anyone else's, sophos support suggested rolling back or checking results after disabling HA and recreating it as part of troubleshooting . 17.1.2 MR-2 fixes NC-30288 and NC-30520 for us so rolling back will be unwanted. I haven't had an opportunity to try HA recreation after hours yet so I'm interested in what others have accomplished.
confirming same high cpu problem here on XG330 rev2 HA pairs after updating from 17.0.6 MR-6 to 17.1.2 MR-2. "gethainfo" and "worker" processes seem to use the most. based on some specific issues logged during the sfos updates in our environment that might not apply to anyone else's, sophos support suggested rolling back or checking results after disabling HA and recreating it as part of troubleshooting . 17.1.2 MR-2 fixes NC-30288 and NC-30520 for us so rolling back will be unwanted. I haven't had an opportunity to try HA recreation after hours yet so I'm interested in what others have accomplished.
Hi onward
Apologies for this continued inconvenience, this high CPU issue caused by gethainfo process could be related to NC-36927. Have you already raised a support case for further investigation? If so, please PM me with your support case ID for follow up.
To our community,
If you are also experiencing similar symptoms, please raise a support case and PM me with your ID for follow up.
Regards,
Hi,
So, for my case, my engineer assigned to asked for help by GES SOPHOS. My case is 8343418.
@FLO, can you give us the details of the bug NC-36927
Valentin
Hi Valentin,
FloSupport said:this high CPU issue caused by gethainfo process could be related to NC-36927.
This is the available info for this ID, more investigation by our GES engineers would be required to confirm.