PR to see if I can figure out the hanging tests issues on Travis.
I can't reproduce the same issue locally. However what I am seeing on my machine is that tests sometimes fail due to an java.lang.RuntimeException: java.net.BindException: Address already in use: bind exception.
After server.stop(), the INET_PORT remains in a TIME_WAIT state it seems .
Adding a short Thread.sleep() after stopping or a while(available(INET_PORT)) Thread.sleep(500); where available is something like this: http://svn.apache.org/viewvc/camel/trunk/components/camel-test/src/main/java/org/apache/camel/test/AvailablePortFinder.java?view=markup#l130 fixes the issue on my machine.
No sure idea if this is related to the hanging SSE tests though.
So it's not related to the SSE hang up.
Any idea how I can get Travis to show souts in its logs?
And/or, could a temporary hook to output a thread dump be added to the after_failure stage?
PR to see if I can figure out the hanging tests issues on Travis.
I can't reproduce the same issue locally. However what I am seeing on my machine is that tests sometimes fail due to an
java.lang.RuntimeException: java.net.BindException: Address already in use: bind
exception. Afterserver.stop()
, theINET_PORT
remains in aTIME_WAIT
state it seems . Adding a shortThread.sleep()
after stopping or awhile(available(INET_PORT)) Thread.sleep(500);
whereavailable
is something like this: http://svn.apache.org/viewvc/camel/trunk/components/camel-test/src/main/java/org/apache/camel/test/AvailablePortFinder.java?view=markup#l130 fixes the issue on my machine. No sure idea if this is related to the hanging SSE tests though.