On 13/02/2015 12:45, Rémy Maucherat wrote: > 2015-02-13 10:05 GMT+01:00 <ma...@apache.org>: > >> Author: markt >> Date: Fri Feb 13 09:05:28 2015 >> New Revision: 1659471 >> >> URL: http://svn.apache.org/r1659471 >> Log: >> Prevent multiple async timeouts on slow running / heavily loaded systems >> > > Unless it's a coincidence, this causes some testsuites regressions. Odd > though. > > Still no real idea on the NIO2 SSL failure, but my NPE fix now demonstrates > some double socket closing.
I'm fairly sure a good chunk of these (if not all) are fixed by: http://svn.apache.org/r1659486 With that patch applied the tests (appear to) pass. The NIO2 SSL issue is the one I'm not sure about either. When trying to track the root cause of the failures down the unit tests weren't triggering the failures reliably. I had to add a few break points and use a debugger to reliably trigger the error. I'm keeping an eye on the CI systems. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
