[sheepdog] [PATCH] sheep: fix oid scheduling in recovery

MORITA Kazutaka morita.kazutaka at lab.ntt.co.jp
Tue Jun 12 08:19:26 CEST 2012


At Fri, 8 Jun 2012 14:24:44 -0400,
Christoph Hellwig wrote:
> 
> On Tue, Jun 05, 2012 at 11:12:30PM +0900, MORITA Kazutaka wrote:
> 
> > I'm not against going back to aio callback codes, but I'd rather see a
> > block driver with a sheepdog client library so that we can more easily
> > develop sheepdog client codes.
> 
> Given that the coroutines are a qemu-local concept I think moving to
> a normal callback based event driven model is a prerequisite for that,
> otherwise the code won't be very useful outside of qemu.

Sheepdog tree already has codes to process sheepdog request
asynchronously (lib/net.c, lib/event.c, etc), so it looks easier to me
to implement an AIO client library in Sheepdog project rather than
recreate an async block driver in qemu.  But I'm not against moving to
the async driver as the first step.

Thanks,

Kazutaka



More information about the sheepdog mailing list