On Mon, Jan 21, 2019 at 6:10 PM Gianluca Cecchi <[email protected]> wrote:
> Hello, > it happened two times, always with VMs composed by more than 1 disk. > Now I have a VM with 9 disks and a total of about 370Gb: > 1 x 90Gb > 5 x 50Gb > 3 x 10Gb > > I export vm to an export domain and it started 2 hours and 10 minutes ago > at 15:46. > At beginning the write rate on export domain was 120MB/s (in line with I/O > capabilities of storage subsystem). > It seems 5 disks completed ok, while 4 don't complete, even if it seems to > be some reading activity: > > with command > iotop -d 3 -k -o -P > I get this on hypervisor where qemu-img convert command is executing > > Total DISK READ : 5712.75 K/s | Total DISK WRITE : 746.58 K/s > Actual DISK READ: 6537.11 K/s | Actual DISK WRITE: 6.89 K/s > PID PRIO USER DISK READ> DISK WRITE SWAPIN IO COMMAND > > 21238 idle vdsm 1344.18 K/s 183.12 K/s 0.00 % 0.15 % qemu-img > convert -p -t non~1d23-4829-8350-f81fa16ea8b0 > 21454 idle vdsm 1344.18 K/s 183.12 K/s 0.00 % 0.06 % qemu-img > convert -p -t non~443d-4563-8879-3dbd711b6936 > 21455 idle vdsm 1344.18 K/s 189.68 K/s 0.00 % 0.08 % qemu-img > convert -p -t non~b67d-4acd-9bb5-485bc3ffdf2e > 21548 idle vdsm 1344.18 K/s 190.67 K/s 0.00 % 0.06 % qemu-img > convert -p -t non~7fc > > It seems it was only a matter of waiting. Completed after 4 hours and 47 minutes: Vm raclab1 was exported successfully to xfer 21/01/1920:33:22 Probably depending on I/O priority set by default as "idle" for the "qemu-img convert" command Any fair/supported way to manually change I/O priority, just in case? Gianluca
_______________________________________________ Users mailing list -- [email protected] To unsubscribe send an email to [email protected] Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/[email protected]/message/RCPFHYDUVJQHKE4FYR4PUIJID7Y47WG7/
