Hi,
I have done some benchmark with librbd octopus,
results are really great with new the write-around cache policy. (Simply need
to enable cache=writeback in qemu)
Here some iops result with 1vm - 1disk - 4k block iodepth=64, librbd, no
iothread.
nautilus-cache=none nautilus-cache=writeback
octopus-cache=none octopus-cache=writeback
randread 4k 62.1k 25.2k
61.1k 60.8k
randwrite 4k 27.7k 19.5k
34.5k 53.0k
seqwrite 4k 7850 37.5k
24.9k 82.6k
We don't have anymore read degradation with cache=writeback, and write
performance is around 2,5x faster !
I think that when octopus will be ready for proxmox6, we could advise user to
change the disk cache to writeback.
(and maybe select it as default in gui when an rbd storage is used for a new vm
isk)
_______________________________________________
pve-devel mailing list
pve-devel@pve.proxmox.com
https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
