Apart from false failures because of wrongly targetted branch I saw the job fail the build on master for patches that were to be applied on master. Haven't seen the job reporting in the last few days however.
Have you started on the fix for the branch mapping? If not - I could start taking a look sometime this week. On Mon, Aug 12, 2013 at 02:30:10PM +0200, Hugo Trippaers wrote: > Prassana, > > What is broken? Do you have some details (just got back from some > time off, so i might have missed a few mails) > > By the way on the subject on gerrit, we had this discussion before. > If anyone wants to refresh the arguments from last time: > http://markmail.org/message/inerurmjtc6v57ba > > Cheers, > > Hugo > > > On Aug 10, 2013, at 7:15 AM, Prasanna Santhanam <t...@apache.org> wrote: > > > The jenkins integration is broken atm. If someone has time to look > > into it please send a pull request with the fix here: > > https://github.com/CloudStack-extras/reviewboard-tools/issues/2 > > > > On Fri, Aug 09, 2013 at 10:35:01AM +0900, Dave Cahill wrote: > >> Hi, > >> > >> Alex - sounds good, let's raise the Gerrit topic in a new thread after 4.2. > >> > >> Daan - you're right that if Jenkins builds, that should prove the code > >> compiles cleanly. > >> Have we actually been gating commits on the catofong report though? I > >> looked at the > >> Jenkins build behind it [1], and it hasn't succeeded in 4 days, so I'm > >> guessing we're > >> not gating based on it. > >> > >> Aside from the Jenkins integration, one of the reasons I wanted Gerrit is > >> to put a step > >> between writing code and it hitting master. Ideally, it would be nice to > >> enforce that a > >> positive code review from someone other than the author is present before > >> pushing. > >> > >> Thanks, > >> Dave. > >> > >> [1] > >> http://jenkins.cloudstack.org/view/management/job/mgmt-build-reviewboard-requests/ > >> > >> On Fri, Aug 9, 2013 at 4:16 AM, Daan Hoogland > >> <daan.hoogl...@gmail.com>wrote: > >> > >>> Dave, wouldn't jenkins do a build like that? Meanng the positive cato > >>> report is a must to allow commits > >>> > >>> On Thu, Aug 8, 2013 at 4:54 PM, Alex Huang <alex.hu...@citrix.com> wrote: > >>>> Yup. That was the reason why it worked on my machine but doesn't work > >>> on jenkins. It had a copy of the jar in the local maven cache so it > >>> continued to build even though it's missing. Thanks for reverting. Will > >>> fix. > >>>> > >>>> I still don't understand why we have problems with Gerrit. It would > >>> have prevented exactly this type of problems. I know a lot of people are > >>> busy on 4.2. Will bring it up post 4.2 again. > >>>> > >>>> --Alex > >>>> > >>>> > >>>> > >>>> From: Dave Cahill [mailto:dcah...@midokura.com] > >>>> Sent: Thursday, August 8, 2013 3:45 AM > >>>> To: dev@cloudstack.apache.org; Dave Cahill; Alex Huang > >>>> Subject: Re: Master build failure > >>>> > >>>> Thanks Prasanna. > >>>> > >>>> One suggestion I have is that if anyone makes pom file changes (or any > >>> large changes at > >>>> all, to be honest), that they nuke their local Maven cache (rm -rf > >>> ~/.m2/repository) and do > >>>> a "clean install" before submitting a patch. This way, you don't end up > >>> thinking your work > >>>> builds when it just happens to build due to leftover, cached jars. > >>>> > >>>> Alex, do you have any ideas on build stability? There have been 2 build > >>> breakages in the > >>>> past few days (ref: "compile error on 4.2" thread), and looking at git > >>> log I don't see much > >>>> sign of review on the changes. We discussed Gerrit a while back, but I > >>> think the conclusion > >>>> was that it would slow us down? > >>>> > >>>> Thanks, > >>>> Dave. > >>>> > >>>> On Thu, Aug 8, 2013 at 6:13 PM, Prasanna Santhanam <t...@apache.org > >>> <mailto:t...@apache.org>> wrote: > >>>> Reverted the range 30c33415..f6a2c817bc. Master builds now. Others > >>>> who've checked in after these will still retain their commits but > >>>> please build before push. It's clear we are only worried about 4.2 now > >>>> but we are breaking master for those who are beginning to work ahead > >>>> for 4.3 > >>>> > >>>> On Thu, Aug 08, 2013 at 02:15:58PM +0530, Prasanna Santhanam wrote: > >>>>> Dave's been attempting to fix this and it looks to be intertwined in a > >>>>> few commits. Large swaths of code has moved around. I'll revert to > >>>>> last known state from Jenkins if we can't figure out a simple fix so > >>>>> Dave can proceed. > >>>>> > >>>>> On Thu, Aug 08, 2013 at 04:11:36PM +0900, Dave Cahill wrote: > >>>>>> Hi, > >>>>>> > >>>>>> As Prasanna noted in IRC, the cloudstack-master-maven Jenkins build is > >>>>>> failing with a similar error: > >>>>>> > >>>>>> https://builds.apache.org/job/cloudstack-master-maven/2645/console > >>>>>> > >>>>>> Failed to execute goal on project cloud-engine-components-api: Could > >>>>>> not resolve dependencies for project > >>>>>> org.apache.cloudstack:cloud-engine-components-api:jar:4.3.0-SNAPSHOT: > >>>>>> Could not find artifact > >>>>>> org.apache.cloudstack:cloud-framework-config:jar:4.3.0-SNAPSHOT in > >>>>>> Nexus (http://repository.apache.org/snapshots) > >>>>>> cause : Could not resolve dependencies for project > >>>>>> org.apache.cloudstack:cloud-engine-components-api:jar:4.3.0-SNAPSHOT: > >>>>>> Could not find artifact > >>>>>> org.apache.cloudstack:cloud-framework-config:jar:4.3.0-SNAPSHOT in > >>>>>> Nexus (http://repository.apache.org/snapshots) > >>>>>> > >>>>>> > >>>>>> Thanks, > >>>>>> > >>>>>> Dave. > >>>>>> > >>>>>> > >>>>>> On Thu, Aug 8, 2013 at 3:53 PM, Dave Cahill <dcah...@midokura.com > >>> <mailto:dcah...@midokura.com>> wrote: > >>>>>> > >>>>>>> Hi, > >>>>>>> > >>>>>>> I'm seeing a build failure with master - I've checked the usual > >>> gotchas, > >>>>>>> but > >>>>>>> this seems to be something else. Possibly related to the 4.2.0 to > >>> 4.3.0 > >>>>>>> update? > >>>>>>> > >>>>>>> The missing dependency is cloud-engine-compute. It's listed as a > >>>>>>> dependency in > >>>>>>> engine/service/pom.xml with no "type" specified, which AFAIK > >>> defaults to > >>>>>>> "jar". > >>>>>>> So my understanding is that it's trying to download version > >>> 4.3.0-SNAPSHOT > >>>>>>> of the cloud-engine-compute jar from repository.apache.org< > >>> http://repository.apache.org>. How does the > >>>>>>> jar get > >>>>>>> uploaded there in the first place? Maybe that's a step we missed > >>> when > >>>>>>> witching > >>>>>>> from 4.2.0 to 4.3.0? > >>>>>>> > >>>>>>> Here's what I ran: > >>>>>>> > >>>>>>> # Nuke my local maven cache > >>>>>>> $ rm -rf ~/.m2/repository > >>>>>>> > >>>>>>> # Get latest master > >>>>>>> $ git pull > >>>>>>> > >>>>>>> # Clean any files hanging around my git working copy just in case > >>>>>>> $ git clean -dxf > >>>>>>> > >>>>>>> # Build > >>>>>>> $ mvn clean install -P developer,systemvm -DskipTests > >>>>>>> > >>>>>>> Here's the error I got: > >>>>>>> > >>>>>>> [ERROR] Failed to execute goal on project cloud-engine-service: > >>> Could not > >>>>>>> resolve dependencies for project > >>>>>>> org.apache.cloudstack:cloud-engine-service:war:4.3.0-SNAPSHOT: > >>> Failure to > >>>>>>> find org.apache.cloudstack:cloud-engine-compute:jar:4.3.0-SNAPSHOT > >>> in > >>>>>>> http://repository.apache.org/snapshots was cached in the local > >>>>>>> repository, resolution will not be reattempted until the update > >>> interval of > >>>>>>> apache.snapshots has elapsed or updates are forced -> [Help 1] > >>>>>>> > >>>>>>> Thanks, > >>>>>>> Dave. > >>>>>>> > >>>>>>> > >>>>>>> > >>>>> > >>>>> -- > >>>>> Prasanna., > >>>>> > >>>>> ------------------------ > >>>>> Powered by BigRock.com > >>>> > >>>> -- > >>>> Prasanna., > >>>> > >>>> ------------------------ > >>>> Powered by BigRock.com > >>>> > >>> > > > > -- > > Prasanna., > > > > ------------------------ > > Powered by BigRock.com > > -- Prasanna., ------------------------ Powered by BigRock.com
