On 05/18/2012 07:23 PM, Christoph Hellwig wrote: > On Fri, May 18, 2012 at 07:10:51PM +0800, Liu Yuan wrote: >> Dropped for now, tests show that we need more work to allow concurrent >> access to the same object both by gateway requests and recovery requests. > > We already update the epoch and node list from the maseter thread while > I/O is still in progress, so we'll need to fix this quickly. > > Do you already have a plan of attack for fixing this? > Yes, I have some unsent patches sit on my tree. For Gateway requests only, it runs without problem, but there is some corner case when recovery involved in, especially multiple node confchg events that my patch set doesn't handle very well. So I only post the series that doesn't cause trouble yet to get reviews. I think these patches can get in first. When I pass my tests, I'll post left ones. Thanks, Yuan |