tomcat-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Thomas <ma...@apache.org>
Subject Re: buildbot failure in ASF Buildbot on tomcat-trunk
Date Tue, 13 Nov 2012 13:28:50 GMT
On 13/11/2012 11:31, Mark Thomas wrote:
> On 12/11/2012 05:25, Konstantin Kolinko wrote:
>> 2012/11/12  <buildbot@apache.org>:
>>> The Buildbot has detected a new failure on builder tomcat-trunk while building
ASF Buildbot.
>>> Full details are available at:
>>>  http://ci.apache.org/builders/tomcat-trunk/builds/3550
>>>
>>> Buildbot URL: http://ci.apache.org/
>>>
>>> Buildslave for this Build: bb-vm_ubuntu
>>>
>>> Build Reason: scheduler
>>> Build Source Stamp: [branch tomcat/trunk] 1408165
>>> Blamelist: markt
>>>
>>> BUILD FAILED: failed compile_1
>>>
>>
>> http://ci.apache.org/projects/tomcat/tomcat8/logs/1408165/TEST-org.apache.catalina.core.TestAsyncContextImpl.NIO.txt
>>
>> Testsuite: org.apache.catalina.core.TestAsyncContextImpl
>> Tests run: 41, Failures: 1, Errors: 0, Time elapsed: 79.256 sec
>>
>> Testcase: testTimeoutListenerCompleteNoDispatch took 5.125 sec
>> 	FAILED
>> Uri: /async, Status: 200, Time: 2921
>> junit.framework.AssertionFailedError: Uri: /async, Status: 200, Time: 2921
>> 	at org.apache.catalina.valves.TesterAccessLogValve.validateAccessLog(TesterAccessLogValve.java:81)
>> 	at org.apache.catalina.core.TestAsyncContextImpl.doTestTimeout(TestAsyncContextImpl.java:499)
>> 	at org.apache.catalina.core.TestAsyncContextImpl.testTimeoutListenerCompleteNoDispatch(TestAsyncContextImpl.java:387)
>>
>>
>> The code is
>>             alvGlobal.validateAccessLog(1, 200, TimeoutServlet.ASYNC_TIMEOUT,
>>                     TimeoutServlet.ASYNC_TIMEOUT + TIMEOUT_MARGIN +
>>                     REQUEST_TIME);
>>
>> TimeoutServlet.ASYNC_TIMEOUT is 3000, so it executed faster (2921)
>> than expected.
> 
> That is close enough that I wonder if we have a clock resolution issue
> here. The other possibility is differences in when the various timers
> are started. A short GC pause in the wrong place might be able to
> trigger this. I'll dig a little deeper.

Looks like it could be either. I'll up the error margin from 10ms to
100ms. That should fix most false failures without introducing any false
passes.

Mark


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


Mime
View raw message