[sheepdog] [PATCH 3/3] sheep: split notification messages into two queues

[MORITA Kazutaka]

At Thu, 30 Aug 2012 11:40:26 +0800,
Liu Yuan wrote:
> 
> On 08/30/2012 11:20 AM, MORITA Kazutaka wrote:
> > With your approach, we need to let the cluster driver know which
> > messages must be ordered with confchg events.  Adding all notified
> > messages to confchg (nonblocking) list looks simpler and easier to
> > understand to me.  The reason we introduced two queues is that
> > blocking messages which have process_work() block all cluster events,
> > so I think this patch also works correctly.
> 
> I haven't seen the patch yet, but you mentioned that even with this
> patch, we can't pass 037 test sometimes. no? It is another problem, not
> related to this event ordering that this patch trying to solve?

Yes, I think there is another problem.  I couldn't pass the test even
if I reverted "corosync: fix cluster hang by cluster requests blocking
confchg", which introduced two queues before.

Thanks,

Kazutaka