[Bugs] [Bug 1449329] When either killing or restarting a brick with performance.stat-prefetch on , stat sometimes returns a bad st_size value.

bugzilla at redhat.com bugzilla at redhat.com
Thu May 11 08:52:33 UTC 2017


https://bugzilla.redhat.com/show_bug.cgi?id=1449329

Ravishankar N <ravishankar at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|MODIFIED                    |POST



--- Comment #5 from Ravishankar N <ravishankar at redhat.com> ---
(In reply to miklos.fokin from comment #4)
> (In reply to Ravishankar N from comment #2)
> > Miklos, you might want to give the patch a test. Or you could wait until it
> > gets reviewed first. I wasn't able to hit the issue any more on 3.10 with
> > the patch.
> 
> Hello, sorry for the late response.
> I tried the test with the patch, the results seem to be the same for me as
> with the original arbiter check that I implemented.
> The faulty behavior when killing the brick stopped, so only the problem when
> restarting the brick remains, when it returns a bit less than what it should.
> The reproduction time varied widely though, I tried it 3 times and 2 times
> the reproduction took around 5-10 minutes, while the last one was around 30.

Hi Miklos, the fix was supposed to solve even the lesser stat size problem. I
wasn't able to hit it with the fix on a 30 minute run when I tested the patch
but if you are able to hit it 3/3 times, looks like there is something more to
be fixed.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=fuTRqqLPPR&a=cc_unsubscribe


More information about the Bugs mailing list