findbugs errors -- any way to prevent mad bot from rampant -1s?
Hi, Is there any way to prevent this [1] from happening? Some patch got merged that findbugs complains about. Basically no patches can be merged until that is fixed and everybody rebases everything (because the bot -1's everything). That doesn't seem right. [1] http://i.imgur.com/NNHx1bq.png -- Greg Sheremeta Red Hat, Inc. Sr. Software Engineer, RHEV Cell: 919-807-1086 gshereme@redhat.com
----- Original Message -----
From: "Greg Sheremeta" <gshereme@redhat.com> To: infra@ovirt.org, devel@ovirt.org Cc: "Eyal Edri" <eedri@redhat.com> Sent: Tuesday, December 9, 2014 1:52:01 AM Subject: findbugs errors -- any way to prevent mad bot from rampant -1s?
Hi,
Is there any way to prevent this [1] from happening?
Some patch got merged that findbugs complains about. Basically no patches can be merged until that is fixed and everybody rebases everything (because the bot -1's everything). That doesn't seem right.
What really doesn't seem is that: a. a maintainer merged a patch that doesn't pass findbugs. b. it's been a couple of days and this hasn't been fixed yet. Having said that, I posted a fix[1] for the last issue, pending front-end maintainers' review [1] http://gerrit.ovirt.org/#/c/36002/
[1] http://i.imgur.com/NNHx1bq.png
-- Greg Sheremeta Red Hat, Inc. Sr. Software Engineer, RHEV Cell: 919-807-1086 gshereme@redhat.com _______________________________________________ Infra mailing list Infra@ovirt.org http://lists.ovirt.org/mailman/listinfo/infra
--3a/Z8KDuKqDOIvAo Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 12/08, Greg Sheremeta wrote:
Hi, =20 Is there any way to prevent this [1] from happening? =20 Some patch got merged that findbugs complains about. Basically no patches can be merged until that is fixed and everybody rebases everything (because the bot -1's everything). That doesn't seem right.
What is not right is merging a patch that breaks findbugs. How do you suggest avoiding that? Ignoring the findbugs result always? Forcing to pass it always? (that is, not allowing maintainers to remove the jenkins reviewer). If we only test you code (check out instead of cherry-pick) then you will not have that issue, but you'll have to test anyhow the cherry-pick before merging and we have the same issue + a lot of tests on old code that when rebasing become opsolete. You could try to cross compare jobs running on merged code with the jobs runing on non-merged code and disable voting if merged code is breaking, but you will not be able to merge anyhow as you must make sure the tests pass before merging, so you need a stable repo. So in summary, yes there's something we can do, fix the offeding patch (or revert it) and force any patchset to pass the tests before merging.
=20 [1] http://i.imgur.com/NNHx1bq.png =20 --=20 Greg Sheremeta Red Hat, Inc. Sr. Software Engineer, RHEV Cell: 919-807-1086 gshereme@redhat.com _______________________________________________ Infra mailing list Infra@ovirt.org http://lists.ovirt.org/mailman/listinfo/infra
--=20 David Caro Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605 --3a/Z8KDuKqDOIvAo Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUhrxHAAoJEEBxx+HSYmnDfQwH/jqjVgTCK6iow63c+kjD8d2C RZtN+ZToRT6rlPYSJ7poLGgEvlyQMm0zQXgYzFr/XtJUbhCJC3FHubACJcKLgXwM Vr+h1/ZMjFaXwl1xvOpMPIFAdEN5YtqOC4olHjgSLwxarsYXiNzQGce3DjtOR4/z IUPoBdmNmwxpOWpSEpwAL/C2778rEh9S9ID43fKOM0WjZYCBnQa75wFMCRs04z4X mdegJJCbX4cuqntpuxzD6yTJ9akCKuFbkOm8SSnzzLKXE7DGrOgSER7LCtWSQyoS Tl4Qc5RND0YtJf5HnMUQk26RsIGPffAWvoH4mYRUDg0nVsPLkUSyzG4Q8TdMhXQ= =Kf7F -----END PGP SIGNATURE----- --3a/Z8KDuKqDOIvAo--
----- Original Message -----
From: "David Caro" <dcaroest@redhat.com> To: "Greg Sheremeta" <gshereme@redhat.com> Cc: "Eyal Edri" <eedri@redhat.com>, devel@ovirt.org, infra@ovirt.org Sent: Tuesday, December 9, 2014 11:09:27 AM Subject: Re: findbugs errors -- any way to prevent mad bot from rampant -1s?
On 12/08, Greg Sheremeta wrote:
Hi,
Is there any way to prevent this [1] from happening?
Some patch got merged that findbugs complains about. Basically no patches can be merged until that is fixed and everybody rebases everything (because the bot -1's everything). That doesn't seem right.
What is not right is merging a patch that breaks findbugs.
How do you suggest avoiding that? Ignoring the findbugs result always? Forcing to pass it always? (that is, not allowing maintainers to remove the jenkins reviewer). If we only test you code (check out instead of cherry-pick) then you will not have that issue, but you'll have to test anyhow the cherry-pick before merging and we have the same issue + a lot of tests on old code that when rebasing become opsolete.
You could try to cross compare jobs running on merged code with the jobs runing on non-merged code and disable voting if merged code is breaking, but you will not be able to merge anyhow as you must make sure the tests pass before merging, so you need a stable repo.
So in summary, yes there's something we can do, fix the offeding patch (or revert it) and force any patchset to pass the tests before merging.
totally agree. here's the offending patch that was merged both to master and to 3.5 stable branch: http://gerrit.ovirt.org/#/c/35931/ you can see clearly jenkins failed on it: Patch Set 3: Build Unstable http://jenkins.ovirt.org/job/ovirt-engine_master_unit-tests_gerrit/22350/ : SUCCESS http://jenkins.ovirt.org/job/ovirt-engine_master_find-bugs_gerrit/21862/ : The patch does not pass the findbugs http://jenkins.ovirt.org/job/ovirt_engine_master_compile_checkstyle_gerrit/3... : SUCCESS so please either revert it or fix it to unblock all other patches. e.
[1] http://i.imgur.com/NNHx1bq.png
-- Greg Sheremeta Red Hat, Inc. Sr. Software Engineer, RHEV Cell: 919-807-1086 gshereme@redhat.com _______________________________________________ Infra mailing list Infra@ovirt.org http://lists.ovirt.org/mailman/listinfo/infra
-- David Caro
Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605
_______________________________________________ Infra mailing list Infra@ovirt.org http://lists.ovirt.org/mailman/listinfo/infra
participants (4)
-
Allon Mureinik -
David Caro -
Eyal Edri -
Greg Sheremeta