Why do my autobuilds keep stopping with no messages returned ?

[Jeremy Allison]

On Sat, Jun 17, 2017 at 05:21:52PM +1200, Andrew Bartlett wrote:
> On Sat, 2017-06-17 at 04:31 +0000, Jeremy Allison wrote:
> > 
> > 
> > Nope. And it stopped in *exactly* the same place as last
> > time. No error messages at all, no email saying the
> > build failed.
> > 
> > Any clue as to what might be wrong here ? All local
> > make tests pass correctly.
> > 
> > SOCKET_WRAPPER_DIR=/memdisk/jra/a/b478575/samba/bin/ab/w
> > DNS: Faking nameserver
> > [1(0)/2129 at 0s] samba.tests.source
> > [2(6)/2129 at 0s] samba.tests.docs
> > [3(8)/2129 at 2m26s] samba.tests.blackbox.ndrdump
> > [4(13)/2129 at 2m26s] ldb.python
> > [5(170)/2129 at 2m27s] samba.tests.credentials
> > [6(206)/2129 at 2m27s] samba.tests.credentials.python3
> > [7(242)/2129 at 2m27s] samba.tests.registry
> > [8(246)/2129 at 2m27s] samba.tests.auth
> > [9(251)/2129 at 2m28s] samba.tests.auth.python3
> > [10(256)/2129 at 2m28s] samba.tests.get_opt
> > [11(260)/2129 at 2m28s] samba.tests.get_opt.python3
> > [12(264)/2129 at 2m28s] samba.tests.security
> > [13(285)/2129 at 2m29s] samba.tests.dcerpc.misc
> > [14(292)/2129 at 2m29s] samba.tests.dcerpc.misc.python3
> > [15(299)/2129 at 2m30s] samba.tests.dcerpc.integer
> > [16(334)/2129 at 2m32s] samba.tests.param
> > [17(347)/2129 at 2m33s] samba.tests.param.python3
> > [18(360)/2129 at 2m34s] samba.tests.upgrade
> > [19(362)/2129 at 2m34s] samba.tests.core
> > [20(371)/2129 at 2m35s] samba.tests.core.python3
> > [21(380)/2129 at 2m35s] samba.tests.provision
> > [22(391)/2129 at 2m35s] samba.tests.samba3
> > [23(412)/2129 at 2m38s] samba.tests.strings
> > [24(414)/2129 at 2m38s] samba.tests.netcmd
> > [25(419)/2129 at 2m39s] samba.tests.dcerpc.rpc_talloc
> > [26(420)/2129 at 2m39s] samba.tests.dcerpc.array
> > [27(425)/2129 at 2m39s] samba.tests.dcerpc.string
> > [28(431)/2129 at 2m40s] samba.tests.hostconfig
> > [29(436)/2129 at 2m40s] samba.tests.messaging(ad_dc_ntvfs:local)
> > [30(444)/2129 at 2m57s] samba.tests.samba3sam
> > [31(454)/2129 at 2m58s] wafsamba.tests.test_suite
> > [32(474)/2129 at 2m58s] samba4.blackbox.demote-saveddb
> > [33(479)/2129 at 3m3s] samba4.blackbox.dbcheck.alpha13
> > [34(505)/2129 at 3m53s] samba4.blackbox.dbcheck.release-4-0-0
> > [35(531)/2129 at 4m43s] samba4.blackbox.dbcheck.release-4-1-0rc3
> 
> Your current build has got further:
> [365(1990)/2129 at 23m43s] samba3.base.delete(nt4_dc)
> [366(2022)/2129 at 23m45s] samba3.base.delete(ad_dc)
> [367(2054)/2129 at 23m59s] samba3.base.deny1(nt4_dc)
> [368(2055)/2129 at 24m52s] samba3.base.deny1(ad_dc)
> [369(2056)/2129 at 25m44s] samba3.base.deny2(nt4_dc)

Yeah, that's a bisected build to try and discover
if there's a patch autobuild doesn't like.

> I'm not really sure what is going on.  Did you know you can run
> autobuild locally?  Use:
> 
>  ./script/autobuild.py --testbase=/tmp
> 
> A --tail might be helpful to follow along.

Thanks for the help. If this build gets through
and the second patchset doesn't I'll use that
to investigate further (the full patchset passes
a complete make test though).

> I've grabbed your tree and am running it on the Catalyst cloud also, in
> case that gives more clues.

Thanks, much appreciated !

Jeremy.