On Sat, Jun 30, 2012 at 11:19 AM, Jukka Zitting <[email protected]> wrote: > Hi, > > On Sat, Jun 30, 2012 at 4:15 AM, Travis-CI <[email protected]> > wrote: >> View the full build log and details: >> http://travis-ci.org/apache/jackrabbit-oak/builds/1741471 > > The build says: > > Failed tests: > readRevisions[2](org.apache.jackrabbit.mk.large.ManyRevisionsTest): > 00000000000000c8..00000000000000c9: [] expected:<2> but was:<0> > > Can this be a result of the background garbage collector cleaning up > the journal before ManyRevisionsTest has a chance to read them? If
yes, that's IMO the most likely cause. > yes, I think we should either fix the test case or change the behavior > of the garbage collector. In either case I think it would be a good > idea to better define and document how long a client can expect > revisions to remain available. agreed, i created an issue a while ago [0]. let's move forward there. cheers stefan [0] https://issues.apache.org/jira/browse/OAK-114 > > BR, > > Jukka Zitting
