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

Christoph Hellwig hch at infradead.org
Tue Jun 5 13:07:06 CEST 2012


On Mon, Jun 04, 2012 at 07:00:17PM +0800, Liu Yuan wrote:
> > I think the right fix is to simply give each recover_object_work() call
> > it's own work_struct in a structure also containing the oid.  While
> > this means a memory allocation per object to be recovered it also means
> > complete independence between recovery operations, including kicking off
> > onces that have I/O pending ASAP and allowing multiple recoveries in
> > parallel.  I'm about to leave for a long haul flight and will try to
> > implement this solution while I'm on the plane.
> > 
> 
> 
> Okay, I'll apply first 3 patches and leave this one

I didn't manage to finish my implementation in time.  I'd suggest to
put your change 3 in for now, but I plan to get back to this.




More information about the sheepdog mailing list