[sheepdog-users] [sheepdog] Qemu Live Migration and Cache

Liu Yuan namei.unix at gmail.com
Thu Jun 14 04:50:35 CEST 2012


Hi Bastian

On 06/14/2012 12:10 AM, Bastian Scholz wrote:
> Is there a kind of "standard" test for checking this kind of
> cache consistency?
> 

Thanks for your test. Unfortunately, no. If you come to any conclusion,
feel free to write anything useful to our github wiki.

> I try it with a virtual machine, running Flexible I/O tester
> fio (with randrw, verify) inside and a script, migrating this
> virtual machine from host A to B and back in a loop...
> 
> I have made only a few short checks at the moment, but fio
> show no verify failure, whatever cache strategy is used.
> 

Currently only kind of latest QEMU version can take advantage of object
cache. For older QEMU, 'cache=XXX' means nothing. Check /store/cache/
directory to see if object cache is really enabled or not.

> If I increase fio blocksize, the migration takes longer, I
> guess, because kvm/qemu needs more time for cache flush. If
> blocksize is too big, migration stops at all as long as fio
> runs, after finishing, migration works as expected again.

I guess QEMU is blocking for syncing dirty data and this behavior looks
right to me.

Thanks,
Yuan



More information about the sheepdog-users mailing list