It turned out to be a problem with the MemoryJournal that is used in the tests. I don't know why the test wasn't failing for me before. I'm certain I ran them for such an important change. The test that failed seems to test creation of initial index, which shouldn't depend on correctly processing journal entries of external changes. At least I don't see why it would have to. But it does. The Jenkins job seems to have been choking on some unrelated error and is now disabled entirely. See my earlier message today about that.
-- Unico On Tue, Aug 4, 2015 at 3:39 PM, Unico Hommes <[email protected]> wrote: > This issue was fixed, see for reference also [1]. I just forgot to > resolve it. I will look into the failing test. Looks somewhat > unrelated at first glance. > > 1. https://issues.onehippo.com/browse/REPO-1216 > > On Tue, Aug 4, 2015 at 12:38 PM, Alex Parvulescu > <[email protected]> wrote: >> Hi, >> >> Following my earlier message, it looks like JCR-3888 [0] is not marked as >> resolved yet, but I have no insight into what the status is. >> I would suggest reverting the commit (a fix should not introduce failing >> tests anyways) and cutting the release without it. >> >> best, >> alex >> >> >> [0] https://issues.apache.org/jira/browse/JCR-3888 >> >> >> >> On Tue, Aug 4, 2015 at 12:32 PM, Alex Parvulescu <[email protected]> >> wrote: >>> >>> Hi, >>> >>> With the help of git bisect I identified the following commit as the >>> culprit [0]. >>> >>> 4bc1dba12fd8ce8ec2b95e5da2e44387db21ba3e is the first bad commit >>> commit 4bc1dba12fd8ce8ec2b95e5da2e44387db21ba3e >>> Author: Unico Hommes <[email protected]> >>> Date: Wed Jun 24 10:21:04 2015 +0000 >>> verify double check works in for multiple entries errors >>> git-svn-id: >>> https://svn.apache.org/repos/asf/jackrabbit/trunk@1687223 >>> 13f79535-47bb-0310-9956-ffa450edef68 >>> >>> >>> and indeed reverting that one makes the tests pass. >>> >>> hope this helps, >>> alex >>> >>> [0] >>> https://github.com/apache/jackrabbit/commit/6cd4fce5d84ffb24c2ae919c50c8aa50b342ba37 >>> >>> >>> >>> On Tue, Aug 4, 2015 at 11:17 AM, Davide Giannella <[email protected]> >>> wrote: >>>> >>>> On 03/08/2015 14:34, Davide Giannella wrote: >>>> > Hello team, >>>> > >>>> > in order to unlock the release of Oak 1.3.4 we should release JR >>>> > 2.10.2. >>>> > >>>> > I'm planning to do it tomorrow (4/8) around 11AM CEST. >>>> > >>>> The release is currently blocked as it fails constantly on >>>> >>>> Failed tests: >>>> IndexingQueueTest>AbstractJCRTest.run:464->testInitialIndex:128 >>>> expected:<110> but was:<0> >>>> >>>> and sometimes on >>>> >>>> IndexingQueueTest>AbstractJCRTest.run:464->testQueue:71 null >>>> >>>> I'll have a look; but it anyone already has hints about the root cause >>>> it will speed-up things >>>> >>>> Cheers >>>> Davide >>>> >>>> >>> >>
