[OmniOS-discuss] Ang: Re: r151014 KVM crash

[Dan McDonald]

> On Apr 10, 2015, at 3:06 AM, Johan Kragsterman <johan.kragsterman at capvert.se> wrote:
> 
> 
> I have now changed back again to r151014, and I seem to not be able to repete this behaviour. I have been building several chroot's and images, and doing updates in the way that earlier chrashed the KVM process, but I see no problem now, and no extensive logging. And therefore I don't have access to the consol messages I saw earlier, as well, unfortunatly...
> 
> I don't know wether this is good or bad, that it seem to work now. It would have been good to know what caused the problems, but it is nice that it seems to work without problems now...

I see two potential sources of a problem, neither of which I can easily test because nobody can easily reproduce the problem:

1.) The qemu/userland bits have missing pieces that should've been forward-ported.  NOTE: We aren't matching the joyent upstream's kvm-cmd because we don't have their Virtual Network Driver (VND) in illumos-omnios or upstream in illumos-gate yet.  I was helping Joyent to upstream it, but both Joyent and I got slammed with other things, so it didn't make it for r151014.  We have one of their security fixes, but we are missing a few commits from upstream.

2.) There's some kernel problem that perhaps upstream isn't seeing because their illumos has additional fixes.  I think this one is less likely, but the possibility exists.

Either one will require dedicated effort, which I don't have cycles for at the moment.

Thank you for the followup, Johan. My suspicion is it's a race, or other hard to synchronize and therefore reproduce problem.

Dan