[Gluster-devel] intermittent test failures when own-thread is set

[Jeff Darcy]

On 06/05/2013 05:42 AM, John Smith wrote:
> Thanks to the recent commit 5c1710e, the 'prove' tests no longer hang
> for me when setting 'option transport.socket.own-thread on' in the
> .vol files.
> 
> However, now I am getting (intermittent) failues of these tests when
> own-thread is set :
> 
> ./tests/bugs/bug-857330/normal.t
> ./tests/bugs/bug-884455.t
> ./tests/performance/open-behind.t

I ran all three of these multiple times, and the only failure I could
reproduce was the open-behind one.  AFAICT that test is totally bogus,
not even coming close to testing what it claims.  I'll be submitting a
patch for it, which both tests the right thing and passes with
own-thread set.  As for the others, please update

	https://bugzilla.redhat.com/show_bug.cgi?id=970194

with the new info.  I'm not saying these failures aren't real, but I
need to work on some other things for a while and if the information is
only on the mailing list then it's practically guaranteed that I'll be
unable to pick up the thread later.