I just saw an update from Proxmox on their host backup last week, which has its first patches / first implementation. It would be great if you could combine your approach into the official one.
I ran into the same warning after upgrading from PVE 8 to PVE 9 with Ceph Squid.
It looks like /etc/init.d/ceph is simply an obsolete conffile left behind by ceph-base:
# dpkg -S /etc/init.d/ceph
ceph-base: /etc/init.d/ceph
# dpkg-query -W...
I can confirm EnablePerCpuClockTickScheduling=2 fixes / works around the issue. CPU load returns to 23H2 levels and CPU cores assigned to the VM are able to enter deeper C states as expected. Thank you!
I've attached this info to the QEMU ticket...
Danke für die News, aber wie immer hat RedHat kein Changelog veröffentlicht.
Daher habe ich die KI mal gebeten die Commits seit 285 zu vergleichen und eine Zusammenfassung zu schreiben. Wie immer ist das KI Ergebnis nicht unbedingt 100% aber...
No, it's off. 24H2 indeed introduced per core clock ticks, and apparently that is what kills residency in VMs. I don't really understand why, but that is one difference in the KiSetTime/KiGetTime functions in 24H2 compared to Server 2022, for...
hi,
at the moment there is no option to weigh CPU load balancing higher than memory usage balancing.
feel free to file a feature request on bugzilla [0]
[0] https://bugzilla.proxmox.com/
Ich hab’s bei mir so gemacht, dass ich zunächst Wireguard konfiguriert und die Verbindung getestet habe, und anschließend ›pveproxy‹ und den sshd an dessen Interface gehängt habe. Zuerst den sshd und das getestet, dann ›pveproxy‹ – da sehe ich...
@_gabriel exactly as @RoCE-geek said.
Hardware: 2x Xeon 6230R and both VMs had this config.
agent: 1
bios: ovmf
boot: order=scsi0;ide2;ide0;net0
cores: 52
cpu...
Danke dir!
Der erste Befehl zeigt und es geht dabei um die beiden Lexar nVMEs mit je 2TB Speicher
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS FSTYPE LABEL MODEL
sda 8:0...
Das sollte grundsätzlich klappen. Hardware-RAID hast du im Ugreen aber praktisch nicht, sowas gibts für NVMe in den Mini-Kisten nie. Zu 99% ist das ein ZFS-Mirror, den der PVE-Installer oder du selbst angelegt hat. Check mal zpool status, und...
yes, there is currently a bug with the permission on this api call, see the first version of the patch i sent:
https://lore.proxmox.com/pdm-devel/20260708133630.1807997-3-d.csapak@proxmox.com/
i did not have time yet to send a v2, but will do soon
hi,
at the moment there is no option to weigh CPU load balancing higher than memory usage balancing.
feel free to file a feature request on bugzilla [0]
[0] https://bugzilla.proxmox.com/
Hi!
Aktuell läuft Proxmox (aktuellste Version) auf einem Ugreen NAS SXP 4800 Pro, ohne das ehemalige OS (wurde entfernt), direkt installiert.
Ich würde gerne diese Installation auf einen Minisforum MS-01 umziehen.
Beide haben einen Intel...
I investigated a bit further, and found out that there seem to be only three cpufreq modules compiled for proxmox's kernel (and the kernel config isn't available):
root@pev:~# ls /usr/lib/modules/$(uname -r)/kernel/drivers/cpufreq/...
No, it's off. 24H2 indeed introduced per core clock ticks, and apparently that is what kills residency in VMs. I don't really understand why, but that is one difference in the KiSetTime/KiGetTime functions in 24H2 compared to Server 2022, for...
I can confirm EnablePerCpuClockTickScheduling=2 fixes / works around the issue. CPU load returns to 23H2 levels and CPU cores assigned to the VM are able to enter deeper C states as expected. Thank you!
I've attached this info to the QEMU ticket...