Do you know if this setback would be solved by using LVM instead of a img file on some Filesystem?
On Oct 13, 2016 13:56, "Jiri 'Ghormoon' Novak" <ghorm...@gmail.com> wrote: > Hi, > > yes, also with SSD it is possible to hang it in my case, though it needs > quite a lot of workload to happen. also in my case it gets a bit worse than > on ext4 becuase of the CoW nature of ZFS, but it seems the root cause is > overloaded IO. > > IT would be nice to make the guest wait, not to pause the VM, that would > make more sense in this case. though it's true it might not be implemented > ... > > Gh. > > Andrei Grigore wrote: > > This was also my thought. It seems like upon some level of IO activity it > hangs for a bit. I am using a SSD already. > > I am not sure you can do that. > > On Oct 13, 2016 11:17, "Ghormoon" <ghorm...@gmail.com> wrote: > >> Hi, >> >> Shuttering/hangs usually happen to me if IO doesnt manage to cope with >> the load. >> When using zfs (zvol) this basically means that I can use this only on >> ssd drives. >> Can this be the same case? Any ideas how to make the guest wait for io, >> not qemu? >> >> Regards, >> Gh. >> >> On Oct 13, 2016 11:09, "Andrei Grigore" <andrei....@gmail.com> wrote: >> >> Hi all, >> >> What can cause random guest Hang-ups? I mean totally very short freezes. >> It can't be the input pheriphrals, since I am using Synergy. Can it be CPU >> related? >> >> I have a FX-8230 with 8 cores and I am using only 6 Cores for the Guest >> with the following pinning: >> >> Core 1 - 2-3 >> Core 2 - 4-5 >> Core 3 - 6-7 >> >> Tried with both .raw and .qcow2 on ext4 Storage. >> >> Thanks, >> Andrei. >> >> _______________________________________________ >> vfio-users mailing list >> vfio-users@redhat.com >> https://www.redhat.com/mailman/listinfo/vfio-users >> >> >> >
_______________________________________________ vfio-users mailing list vfio-users@redhat.com https://www.redhat.com/mailman/listinfo/vfio-users
