Thanks @fiona, this is indeed great to hear. I've read through that other thread and whilst it does appear to resolve at least one issue for some, I'm slightly suspicious that it hasn't addressed all the cases. Thankfully, I haven't seen the issue occurring in a few weeks but when it was at its worst, the bug would be triggered within about 5 mins of booting the VM. That's nothing like the hours required with a test harness to provoke the bug into occurring. Let's hope this is a big step in the right direction though.
Very much appreciate the update, too.
Very much appreciate the update, too.
Good to hear! Can you correlate that with a kernel or QEMU update you did or something else?
Yes, that does sound different from the other issue. If it ever happens again, please share the GDB output as suggested earlier in the thread. (If you are on Proxmox VE 8, it's
apt install pve-qemu-kvm-dbgsym gdb)I cannot, no. Zero changes made - no tinkering with ballooning, mitigations, nothing. There may have been a regular kernel update during that time (currently 6.2.16-4-bpo11-pve, so that's definitely been the case) but no proactive attempts at resolving it by hearsay methods in the forums.
Whilst that's massively unhelpful, I'm keeping my eye on things and if the issue does re-present itself, I'll get the debugging going.
AFAICS, that kernel was uploaded to the no-subscription repository on August 4th, so that might have been it. If you want to, you can check your system logs to see when you first booted into that kernel.