samba3 raw.mux smtorture test is flaky

[Andrew Bartlett]

On Fri, 2011-02-25 at 09:33 +1100, Andrew Bartlett wrote:
> On Thu, 2011-02-24 at 11:40 -0800, Jeremy Allison wrote:
> > On Tue, Feb 22, 2011 at 03:24:54PM +1100, Andrew Bartlett wrote:
> > > Jeremy,
> > > 
> > > I've pushed an update to the s3 selftest system, and it is now much more
> > > reliable.  During one of my tests, I found that raw.mux isn't reliable,
> > > and failed for me with the output I include below.
> > > 
> > > It does not always fail, but I'm sure you want to move this back into
> > > the selftest as soon as possible.  Is there any chance you can look over
> > > this?
> > 
> > Ok, I'm taking a look at this and I can't get it to fail (10-20 repititions).
> > Any hints as to why it's unreliable for you ?
> 
> Not really.  I think I pasted you the full output I got - I'm pretty
> sure I did get it more than once, but not every time.  We can put it
> back into the list and wait for it to show up again if you like. 
> 
> > I'm getting repeatable failures on samba3.smbtorture_s3.plain(dc).TORTURE.smbtorture
> > on my laptop, which don't show up in autobuild :-(. So our tests still aren't
> > where we need them to be yet :-).
> 
> Any details?  Stuff in the environment I might be able to help with, but
> in retrospect, I actually phrased my 'is now reliable' poorly - it's
> actually more reliably showing the unreliable nature of the tests
> (because it now actually finds failures). 

I think that the test is a little flaky - it just failed on samba4 in
tridge's flaky detection. 
http://git.samba.org/autobuild.flakey/2011-03-02-2224/samba4.stdout

Andrew Bartlett 
-- 
Andrew Bartlett                                http://samba.org/~abartlet/
Authentication Developer, Samba Team           http://samba.org
Samba Developer, Cisco Inc.