[pve-devel] [RFC kernel] revert problematic TSC multiplier commit

[Eneko Lacunza]

Hi,

El 19/9/22 a las 10:36, Eneko Lacunza escribió:
> El 15/9/22 a las 12:44, Thomas Lamprecht escribió:
>> Am 14/09/2022 um 11:38 schrieb Eneko Lacunza via pve-devel:
>>> I just made a kick test and 5.19.7-1 kernel continues to have issues for us (tried live migration of 5 VMs from 5.13 to 5.19, 4 VMs froze using 100% CPU).
>> Well, that is somewhat expected..
>> How 5.19 to 5.19 is working out would be more interesting, if you got a test host/vm combo to try.
> I have time today for this.
>
> I'll test:
> - Starting VMs in a 5.19 kernel node, to see if that affects other VMs 
> in the same node.
> - Live migration of VMs between 5.19 kernel nodes.
>
> I'll use nodes with Ryzen 1700 and Ryzen 3700X CPUs.
>
> Any other thing you'd like me to try?
>

Live migration from 5.13 to 5.19 fails about half of the time in my tests.

Starting VMs in a 5.19 node, one by one, works OK.

Live migration between 5.19 nodes (Ryzen 1700 and Ryzen 3700X) works OK.

I bulk-live-migrated 12 VMs for those tests. That would have failed with 
5.15 .

I saw that a new 5.15 kernel (5.15.53) is available. I can repeat these 
tests with that kernel if that helps.

Now a pair of questions:

- Should I expect a new 5.19 kernel that fixes 5.13 to 5.19 live migration?
- Do you recommend to stay with 5.13 kernel or to upgrade to 5.19?

Thanks

Eneko Lacunza
Zuzendari teknikoa | Director técnico
Binovo IT Human Project

Tel. +34 943 569 206 |https://www.binovo.es
Astigarragako Bidea, 2 - 2º izda. Oficina 10-11, 20180 Oiartzun

https://www.youtube.com/user/CANALBINOVO
https://www.linkedin.com/company/37269706/