[vfio-users] VFIO and random host crashes

Colin Godsey crgodsey at gmail.com
Thu May 19 15:24:21 UTC 2016


@Zycorax it definitely sounds like you have a similar issue to me.

That is interesting that you assume it’s disk access- that was definitely
my second big idea. I’ve recently caught it crash under trivial disk load
though, so I’m not sure.

So theres a few other things that lead me to think ‘power’- With a single
VM, it is very stable, but after a few hours I will start to get graphic
artifacts. With 2 VMs running, it will crash eventually, but interestingly
enough 1 VM will start to get artifacts sometime beforehand. Both VMs seem
to be susceptible to artifacts, but I only get crashes when both are under
load. Even if I have one just idling in a game pause menu, it’s stable.

I also recently got my voltage sensors working, and last crash I checked
and my 12v rail was down to about 11.5- which is getting close to the
danger zone. I’m less worried about the 11.5 I was seeing, and more worried
about the fluctuations i see there, plus how easily the whole system can
pull it from a healthy 12.02 to 11.5.

It also dawned on me that when running 2 GPUs like this, your PSU should at
least be sli/crossfire rated. Reading around, anyone with similar sounding
issues in an SLI setup had some kind of power issue (or motherboard,
actually…).

I ordered a round of new parts from Amazon, will swap things out and see if
thats it.

On Thu, May 19, 2016 at 2:42 AM Zycorax Tokoroa <zycorax at phoxden.xyz> wrote:

> I forgot to ask, were it the motherboard's fault, is there any known
> working model compatible with my processor?
>
> _______________________________________________
> vfio-users mailing list
> vfio-users at redhat.com
> https://www.redhat.com/mailman/listinfo/vfio-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20160519/5e072b22/attachment.htm>


More information about the vfio-users mailing list