I’ve got an intermittent issue with my system massively slowing down or hanging when I start a qube - usually sys-usb, sometimes sys-net, occasionally something else.
It’s possible this only happens after an unclean shutdown. I can reproduce it around 50% of the time by holding down the power button.
I’m not sure what fixes it. Sometimes I can hard-shutdown again, then do a clean reboot without starting any qubes, and then it’ll immediately be okay, but that doesn’t always work. Occasionally I’ve seen the wait times go away by themselves after many minutes, but again, sometimes not.
My diagnostics are pretty much limited to running top/vmstat in dom0. I generally can’t get a terminal or console in the affected qubes, and I don’t know what I’m doing with monitoring CPUs at the Xen level - xentop doesn’t seem to be useful in this context and I’m not sure what else there is. I recall looking at long-running blkdiscards with a previous issue, but that could have been a previous installation and I don’t remember where I was looking. There’s also nothing I can identify as obviously relevant in dmesg/journalctl.
I wouldn’t be surprised if it was a hardware issue - I’m on a Framework 11 Intel - but other than this and running a bit hot with some applications, it seems to be doing okay.
sys-net and sys-usb are both disposable. I also created a separate disposable sys-usb2 dedicated to specific PCI devices and get the same problem with that. They’re both configured without strict PCI reset.
Investigating further needs dedicated testing time - since when the issue happens, I typically lose the following hour to recovery attempts and it’s my daily driver.
Appreciate being pointed to any known issues (I’ve searched here but might have missed something), theories and diagnostic ideas. Thanks.