Do you know why it fails when jenkins run and not when we run it? Is it a an issue with the Apache machine running it? In general, I just think it would be good to have it more reliably running it, otherwise it defeats the purpose of having jenkins running it and sending these reports.
-Flavio On Mar 26, 2012, at 12:20 PM, Ivan Kelly wrote: > Simple solution is to > a) make each hedwig test fork a new jvm > and > b) increase the max heap > > The later asserts are probably due to the earlier direct memory issues. > they're in the same test. > > -Ivan > > On 25 Mar 2012, at 14:40, Flavio Junqueira wrote: > >> I was wondering about how to fix the direct buffer memory issue we are >> seeing on hedwig server tests running on jenkins: >> >> https://builds.apache.org/job/bookkeeper-trunk/426/#showFailuresLink >> >> Note that another issue popping up in some tests is an assertion failure and >> not an out of memory issue, like with this one: >> >> >> https://builds.apache.org/job/bookkeeper-trunk/org.apache.bookkeeper$hedwig-server/426/testReport/junit/org.apache.hedwig.server.integration/TestHedwigHub/testServerFailoverPublishOnly_1_/ >> >> -Flavio >> > flavio junqueira senior research scientist [email protected] direct +34 93-183-8828 avinguda diagonal 177, 8th floor, barcelona, 08018, es phone (408) 349 3300 fax (408) 349 3301
