Oh I see, thanks for clarifying. Wondering if that's global jenkins configuration or something we changed? Is that the behavior we want? I think sometimes one might just want to kick hadoop-qa if failure seems spurious or sporadical.
-Mikhail On Wed, Apr 8, 2015 at 9:56 PM, Srikanth Srungarapu <[email protected]> wrote: > Yeah, Sean is right. It aborts if the attachment is already tested once. > Sample error message. > >> Attachment 12723874 is already tested for HBASE-13423 >> Build step 'Execute shell' marked build as failure >> Archiving artifacts >> Sending artifact delta relative to PreCommit-HBASE-Build #13621 >> Archived 28 artifacts >> Archive block size is 32768 >> Received 9 blocks and 207014 bytes >> Compression is 58.8% >> Took 3.5 sec >> Recording test results >> [description-setter] Could not determine description. >> Finished: FAILURE >> > > On Wed, Apr 8, 2015 at 9:37 PM, Sean Busbey <[email protected]> wrote: > >> You can check the Jenkins log to be sure, but the patch tester has a check >> now that short-circuit aborts test runs when it has already seen a specific >> JIRA attachment. >> >> -- >> Sean >> On Apr 8, 2015 11:32 PM, "Mikhail Antonov" <[email protected]> wrote: >> >> > Hm, I see, that's interesting. Thanks Ted. >> > >> > Though I think I observed this behavior with patches not being picked >> > up during several last weeks at least. >> > >> > -Mikhail >> > >> > On Wed, Apr 8, 2015 at 9:27 PM, Ted Yu <[email protected]> wrote: >> > > Today git repo temporarily was inaccessible. >> > > e.g. from >> > https://builds.apache.org/job/PreCommit-HBASE-Build/13626/console >> > > >> > > FATAL: Failed to fetch from >> > > >> > >> https://git-wip-us.apache.org/repos/asf/hbase.githudson.plugins.git.GitException >> > : >> > > Failed to fetch from >> > > https://git-wip-us.apache.org/repos/asf/hbase.git at >> > > hudson.plugins.git.GitSCM.fetchFrom(GitSCM.java:647) >> > > at hudson.plugins.git.GitSCM.retrieveChanges(GitSCM.java:889) >> > > >> > > >> > > FYI >> > > >> > > >> > > On Wed, Apr 8, 2015 at 9:22 PM, Mikhail Antonov <[email protected]> >> > > wrote: >> > > >> > >> Hey guys, >> > >> >> > >> I think hadoop-qa used to pick up and run the last attached patch when >> > >> you click cancel/submit patch. Lately I've noticed that (sometimes?) >> > >> doesn't work - you'd need to re-attach the same file first and then >> > >> click cancel/submit. >> > >> >> > >> Anyone else observed that? Is it some kind of misconfiguration or >> > >> something else? >> > >> >> > >> -Mikhail >> > >> >> > >> > >> > >> > -- >> > Thanks, >> > Michael Antonov >> > >> -- Thanks, Michael Antonov
