Bug#747361: sdformat: FTBFS: most odd-numbered tests time out

[Aaron M. Ucko]

So I see.  At any rate, you can find the full logs at

https://buildd.debian.org/status/logs.php?pkg=sdformat&ver=2.0.0-2&suite=sid

Also, FWIW, one unusual aspect of the autobuilders' environment that
sometimes breaks tests is running with a "home directory" of
/sbuild-nonexistent; see, e.g.,

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=746624

Thanks,
Aaron

Jose Luis Rivero <jrivero at osrfoundation.org> writes:

> I've never seen this before in sdformat upstream project or our own
> building farms. The fact that only odd tests are failing is because
> the even ones are just checkers of the output of the test run
> before. So the conclusion is: all of them are returning a timeout.
>
> May I have access to the building log? What command is exactly being
> run during the test phase?
>
> Thanks!
>
> On 05/07/2014 10:17 PM, Aaron M. Ucko wrote:
>> Source: sdformat
>> Version: 2.0.0-1
>> Severity: serious
>> Justification: fails to build from source
>>
>> Automated builds of sdformat have all been encountering test timeouts,
>> in a distinctive pattern:
>>
>>    The following tests FAILED:
>>    	  1 - INTEGRATION_audio (Timeout)
>>    	  3 - INTEGRATION_cfm_damping_implicit_spring_damper (Timeout)
>>    	  5 - INTEGRATION_fixed_joint_reduction (Timeout)
>>    	  7 - INTEGRATION_joint_axis_frame (Timeout)
>>    	  9 - INTEGRATION_provide_feedback (Timeout)
>>    	  11 - PERFORMANCE_parser_urdf (Timeout)
>>    	  13 - UNIT_SDF_TEST (Timeout)
>>    	  15 - UNIT_Console_TEST (Timeout)
>>    	  19 - UNIT_parser_urdf_TEST (Timeout)
>>
>> It's not clear why, but I do see that the tests were always running in
>> parallel; could they somehow interfere with each other?  At any rate,
>> please do take a look.
>>
>> Thanks!