@fiona
Thank you for your reply.
I understand that it will be overwritten.
Is it correct to understand that on computers equipped with Intel CPUs where virtual machines configured with the Hyper-V role will not start without `level=30`, this...
Happy to hear about this! There is always a lot of good feedback from this gathering!
Sincerely,
Blockbridge Team.
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Can you share this from the node and every guest that uses ZFS?
zpool status -vtP
zpool get autotrim
I have some notes here on how to investiagte IO Delay that you might find useful.
Host :
root@pve:~# zpool status -vtP
pool: share1ntelzfs
state: ONLINE
scan: scrub repaired 0B in 00:17:33 with 0 errors on Sun May 10 00:41:34 2026
config:
NAME STATE...
Sorry for negrobumping, but: I have (I guess) a similar setup with a Supermicro AVAGO 3108 MegaRAID BIOS.
Which firmware version of the 3108 controller are you using?
(Mine is FW 4.670.00-0021)
Grateful for feedback.
The IO delay is not that high that I would be worried. Maybe the VMs were actually doing some useful idle/background IO like trimming or scrubbing/fsck.
EDIT: Maybe find out what your VMs that you maintain are doing before changing your Proxmox...
ok versuche ich :) danke dir ich parke das dann hier solange. ich werde einen monitor holen und anklemmen und warte bis der fehler kommt wieder (oder obs weg bleibt)
EDIT: seems like the i386 packages are now also present with the correct version in the standard Debian repository, so the workaround should not be necessary anymore.
DISCLAIMER: Do this at your own risk!
DISCLAIMER: I'm not writing this in my...
da ist kein Bildschirm Dran ich starte dann halt einfach neu. - hier im Forum und bei Reddit kam halt dieser offload fix zusprache und dann soll das weggewesen sein
The reason is that your manual -cpu host argument will override the other one completely, and thus turn the cet flag(s) back on, so you have to disable them again.
The following command gets the number of the relevant devices (replace 100 with...
I have been at it since midnight, lots of restarts both on the host as well as the VMs, all 3 of them are on ZFS storage which could explain the discard race perhaps ? The host was left on overnight so about 7-8 hours and the delay was nailed at 7%.