Bug#830452: lbcd: FTBFS: dh_auto_test: make -j1 check VERBOSE=1 returned exit code 2

[Russ Allbery]

Lucas Nussbaum  writes:

> I found the source (or more exactly: Antonio Terceiro and Eric Wong did
> in #830353).

> It seems that the Debian EC2 image uses non-default TCP buffer sizes.
> (see https://lists.debian.org/20160719211715.ga8...@xanadu.blop.info )
> When setting them back to the default value, both lbcd and remctl build
> fine.

Ah, thank you!  I'll see if I can adjust the test to be less fragile.
Usually some minor tuning will resolve issues like this.  (It's
irritatingly difficult to properly test network code because the OS loves
hiding things like this.)

-- 
Russ Allbery (r...@debian.org)   

Bug#830452: lbcd: FTBFS: dh_auto_test: make -j1 check VERBOSE=1 returned exit code 2

[Lucas Nussbaum]

On 14/07/16 at 19:04 -0700, Russ Allbery wrote:
> Lucas Nussbaum  writes:
> 
> > Source: lbcd
> > Version: 3.5.2-1
> > Severity: serious
> > Tags: stretch sid
> > User: debian...@lists.debian.org
> > Usertags: qa-ftbfs-20160707 qa-ftbfs
> > Justification: FTBFS on amd64
> 
> > During a rebuild of all packages in sid, your package failed to build on
> > amd64.
> 
> >> util/network/client.FAILED 21-22
> 
> Hi Lucas,
> 
> It looks like there's something about your testing environment that causes
> problems with network timeouts.  You reported two bugs like this against
> two of my packages that include the same network code and run the same
> test, and it failed the same way in both cases.  Both test an attempted
> network action against a local server spun up by the same test that
> doesn't respond.
> 
> The failure you saw here was not reproducible locally or in a Debian
> unstable chroot, so I prepared a new version of lbcd and uploaded it to
> see if the buildd environment was different.  But the build and test also
> succeeds on all the Debian buildds.
> 
> Given that, I'm closing this as a false positive that I can't reproduce,
> but if you have any additional information that indicates this is wrong,
> please do let me know.

Hi,

I found the source (or more exactly: Antonio Terceiro and Eric Wong did
in #830353).

It seems that the Debian EC2 image uses non-default TCP buffer sizes.
(see https://lists.debian.org/20160719211715.ga8...@xanadu.blop.info )
When setting them back to the default value, both lbcd and remctl build
fine.

Lucas

Bug#830452: lbcd: FTBFS: dh_auto_test: make -j1 check VERBOSE=1 returned exit code 2

[Lucas Nussbaum]

Source: lbcd
Version: 3.5.2-1
Severity: serious
Tags: stretch sid
User: debian...@lists.debian.org
Usertags: qa-ftbfs-20160707 qa-ftbfs
Justification: FTBFS on amd64

Hi,

During a rebuild of all packages in sid, your package failed to build on
amd64.

Relevant part (hopefully):
> make[2]: Entering directory '/«PKGBUILDDIR»'
> cd tests && ./runtests -l /«PKGBUILDDIR»/tests/TESTS
> 
> Running all tests listed in TESTS.  If any tests fail, run the failing
> test program with runtests -o to see more details.
> 
> docs/podskipped (POD syntax tests normally skipped)
> docs/pod-spelling...skipped (Spelling tests only run for author)
> perl/critic.skipped (Coding style tests only run for author)
> perl/minimum-versionskipped (Minimum version tests normally skipped)
> perl/strict.skipped (Strictness tests normally skipped)
> portable/asprintf...ok
> portable/daemon.ok
> portable/getaddrinfook (skipped 20 tests)
> portable/getnameinfook (skipped 7 tests)
> portable/inet_aton..ok
> portable/inet_ntoa..ok
> portable/inet_ntop..ok
> portable/snprintf...ok
> portable/strlcatok
> portable/strlcpyok
> portable/strndupok
> server/basicok
> server/errors...ok
> util/fdflag.ok
> util/messages...ok
> util/network/addr-ipv4..ok
> util/network/addr-ipv6..ok
> util/network/client.FAILED 21-22
> util/network/server.ok
> util/vector.ok
> util/xmallocskipped (xmalloc tests only run for author)
> util/xwrite.ok (skipped 2 tests)
> 
> Failed Set Fail/Total (%) Skip Stat  Failing Tests
> -- --    
> util/network/client   2/22 9%00  21-22
> 
> Failed 2/1725 tests, 99.88% okay, 35 tests skipped.
> Files=27,  Tests=1725,  6.01 seconds (0.08 usr + 0.00 sys = 0.08 CPU)
> make[2]: *** [check-local] Error 1
> Makefile:1959: recipe for target 'check-local' failed
> make[2]: Leaving directory '/«PKGBUILDDIR»'
> make[1]: *** [check-am] Error 2
> Makefile:1778: recipe for target 'check-am' failed
> make[1]: Leaving directory '/«PKGBUILDDIR»'
> dh_auto_test: make -j1 check VERBOSE=1 returned exit code 2

The full build log is available from:
   
http://people.debian.org/~lucas/logs/2016/07/07/lbcd_3.5.2-1_unstable_reb.normal.log

A list of current common problems and possible solutions is available at
http://wiki.debian.org/qa.debian.org/FTBFS . You're welcome to contribute!

About the archive rebuild: The rebuild was done on EC2 VM instances from
Amazon Web Services, using a clean, minimal and up-to-date chroot. Every
failed build was retried once to eliminate random failures.