Re: [Qemu-devel] CoW image commit+shrink(= make_empty) support

[Taisuke Yamada]

>> This split-image configration is done to keep VM I/Os on a SSD
>
> This is an interesting use case that I wasn't aware of yet. So you're
> not really interested in a snapshot here, but what you're trying to do
> is using the SSD as some sort of a cache, right?

Right. By splitting, not only I can have "SSD+HDD" configuration,
but I can also go with "local SSD cache + remote storage" configuration.
My aim is to reduce an I/O load from slow or expensive components like
HDD or network.

> The problem that a live block operation could solve sounds unrelated to
> me: While you perform your 'commit' monitor command, the VM doesn't run.
> Some kind of live commit would surely be helpful there (that's what Jeff
> Cody is working on, iirc).

It seems Jeff's work is very close to what I'm looking for.
I'm not yet sure with detail of live-commit, but transaction API does help.
I'll definitely check out wiki doc once Jeff updates it.

> One other thing to consider that complicates everything is that
> committing to the backing file when bdrv_make_empty is implemented
> obviously also means that reads go to the slow disk now. So I guess you
> really want to commit only part of the image on the SSD...

Yes. Althrough my primary focus is on write IOs, combined with qemu's
IO tracing capability, partial committing would be optimal as I can then
always keep frequently read sectors in cache. But that would be complex indeed.

Best Regards,