On Sat Aug 29, 2026 at 5:31 PM UTC, Dave Voutila wrote:
> The issue is due to a flaw in the interrupt delivery to vcpus and is
> present in both 7.8 and 7.9. The probability of hitting that race is
> probably higher as vmd has changed over the past year plus.
>
> It's architecture independent and impacts both Intel and AMD. My guess
> is this is more likely on machines with fewer, slower cores or when
> heavily over-subscribed with guests.

I understand. Another change with my AMD hosts is that they have SSDs.
Might be related. My VM, "potato", is on Odin which had only 7 other
VPSs on it, and very little activity. Prometheus, on the other hand, had
more usage including a Bitcoin node and a Monero node.

> If you could test and report back it would be very helpful.

I'm running this patch on Bragi now. I think I should have more VMs
running on Bragi, but so far I have a Bitcoin node running, doing a
full, pruned sync. I also have one running on Odin (where potato is) as
a control to see if it lasts a few days and dies again.

I'll report back, but so far the patch has been stable.

Thank you for your help!

-Slow Servers

Reply via email to