hi,
potentially troublesome due to the extra contention and possible
conflicts on the commitinfo node.
Conflicts at that point cannot occur within a single cluster node since
we are already serialised here.
this may be correct today, but I'm working on the MongoMK to support
concurrent
Build Update for apache/jackrabbit-oak
-
Build: #2403
Status: Still Failing
Duration: 421 seconds
Commit: c69fc68ab058db2f40fec89093d4ac80fc71a692 (trunk)
Author: Thomas Mueller
Message: OAK-641: Improved benchmark tooling - new CreateManyNodesTest
The CommitInfoEditorProvider class was missing the licence header, should
be fine now.
alex
On Thu, Oct 17, 2013 at 10:29 AM, Travis CI ju...@apache.org wrote:
Build Update for apache/jackrabbit-oak
-
Build: #2403
Status: Still Failing
Duration: 421
Build Update for apache/jackrabbit-oak
-
Build: #2404
Status: Still Failing
Duration: 326 seconds
Commit: 10951318e47db32749a5e7c9d1789439497c (trunk)
Author: Alexandru Parvulescu
Message: OAK-1055: Occasional test failure in ObservationTest.observation
-
On 17.10.13 10:58 , Thomas Mueller wrote:
I suggest to change the mechanism. At least each cluster node should use a
different node name.
I'm on it.
Michael
now there really are failed tests :)
Failed tests:
equalsHashCode(org.apache.jackrabbit.oak.plugins.mongomk.RevisionTest)
On Thu, Oct 17, 2013 at 11:00 AM, Travis CI ju...@apache.org wrote:
Build Update for apache/jackrabbit-oak
-
Build: #2404
Status:
Build Update for apache/jackrabbit-oak
-
Build: #2405
Status: Fixed
Duration: 1497 seconds
Commit: 42837509d85b533c31683184b688d96842f273f5 (trunk)
Author: Michael Duerig
Message: OAK-144 Implement observation
Pass CommitInfo to merge methods instead of round
From the user point of view I think this makes sense and it'd be worth the
effort, also it looks like the global space approach would allow to do that
without loosing too much in terms of performance.
My 2 cents,
Tommaso
2013/10/17 Michael Dürig mdue...@apache.org
Hi,
Currently we can't
Hi,
On Thu, Oct 17, 2013 at 8:16 AM, Michael Dürig mdue...@apache.org wrote:
To avoid the second pass, we could also remember the paths of the moved
nodes in a global place (*).
Perhaps the new CommitInfo?
I assume we'd only need to track moves for local commits as that's
where permission
IMHO it's worth trying to get a solution for this.
if we do, i would suggest that we detect the move before
passing the information down to the commit hooks. in other
words, there should be a method like nodeMoved in order to prevent
that every move-sensitive hook has to write the same code in
Build Update for apache/jackrabbit-oak
-
Build: #2407
Status: Broken
Duration: 763 seconds
Commit: 57649cec4a056ac75707de1838047b0c9033e46c (trunk)
Author: Angela Schreiber
Message: OAK-527: benchmark tests (minor improvement)
git-svn-id:
Build Update for apache/jackrabbit-oak
-
Build: #2408
Status: Fixed
Duration: 1886 seconds
Commit: 10f2c60d0f376fe9d79553b61a231aca4c168ca5 (trunk)
Author: Jukka Zitting
Message: OAK-150: Basic JCR LockManager support
Avoid an NPE in getLock() in case one of
12 matches
Mail list logo