On the Jenkins side, it does look like most of the builds passed all
their tests but they are still marked unstable yellow. We have not been
blue since Dec 02 and that always makes me nervous. If there is some
infrastructure impediment can somebody please describe it so we can see
what can be done?
On 12/7/11 3:40 PM, Jeff Eastman wrote:
Well, that intermittent problem with TestDRM happened twice in a row
for me, hence my last posting. Last run got through it (in 54 minutes)
and did surface a couple more "clusters-0-final" tests which failed.
Working through those...
On 12/7/11 2:59 PM, Grant Ingersoll wrote:
On Dec 7, 2011, at 3:18 PM, Sean Owen wrote:
The build isn't broken
Agreed.
-- Jenkins is, I believe. At least ever error I've
looked at seems to be an error in Apache infra. The tests pass
locally for
me. usually. DistributedRowMatrix tests seem to have some rare
condition
that makes them fail spuriously because some output isn't cleaned up
somewhere.
I think these unstable warnings have to do w/ the various analysis
tools we are running, right?
https://builds.apache.org/job/Mahout-Quality/1234/ and
https://builds.apache.org/job/Mahout-Quality/1234/violations/ and
frankly, some of them just seem downright silly:
https://builds.apache.org/job/Mahout-Quality/1234/violations/file/x1/jenkins/jenkins-slave/workspace/Mahout-Quality/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java/
"'+' should be on a new line." -- That's marked with a "high"
severity? Really?
I run my own Jenkins locally and it builds Mahout once an hour, but
it doesn't have all of these "extra" and it runs clean pretty much
all the time unless there is some issue
checking out from SVN.
On Wed, Dec 7, 2011 at 8:15 PM, Jeff
Eastman<[email protected]>wrote:
There seem to be a lot of changes in Jenkins since the build last ran
successfully. For me it is failing in
DistributedRowMatrix.**testTranspose.
I have a patch I want to commit but need to run all the unit tests
before
doing so. Clearly, there have been a number of commits since the build
broke.
Can we establish a moratorium on more commits until the build is
fixed?
Having it broken at this point is not good.
On 12/7/11 9:25 AM, Apache Jenkins Server wrote:
See<https://builds.apache.org/**job/Mahout-Quality/changes<https://builds.apache.org/job/Mahout-Quality/changes>
--------------------------------------------
Grant Ingersoll
http://www.lucidimagination.com