On Thu, Nov 2, 2017 at 12:22 PM, Eyal Edri <eedri@redhat.com> wrote:
>
>
>
> On Thu, Nov 2, 2017 at 12:12 PM, Yedidyah Bar David <didi@redhat.com> wrote:
>>
>> On Thu, Nov 2, 2017 at 9:56 AM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
>> >
>> >
>> >
>> > 2017-11-02 8:46 GMT+01:00 Eyal Edri <eedri@redhat.com>:
>> >>
>> >>
>> >>
>> >> On Thu, Nov 2, 2017 at 9:24 AM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
>> >>>
>> >>>
>> >>>
>> >>> 2017-11-01 17:16 GMT+01:00 Dafna Ron <dron@redhat.com>:
>> >>>>
>> >>>> Hi,
>> >>>>
>> >>>> we have failures in 4.1 imagebased.
>> >>>>
>> >>>> I think it's also related to the same package from 4.2 that is downloaded in the 4.1 repo.
>> >>>>
>> >>>> 2017-11-01 13:17:53,817::ssh.py::ssh::96::lago.ssh::DEBUG::Command 10946280 on lago-basic-suite-4-1-host-0  errors:
>> >>>>  Error: Package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch (alocalsync)
>> >>>>            Requires: vdsm-client >= 4.20.0
>> >>>>            Available: vdsm-client-4.19.35-2.gitc1d5a55.el7.centos.noarch (alocalsync)
>> >>>>                vdsm-client = 4.19.35-2.gitc1d5a55.el7.centos
>> >>>> Error: Package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch (alocalsync)
>> >>>>            Requires: vdsm >= 4.20.0
>> >>>>            Available: vdsm-4.19.35-2.gitc1d5a55.el7.centos.x86_64 (alocalsync)
>> >>>>                vdsm = 4.19.35-2.gitc1d5a55.el7.centos
>>
>> Where exactly is this failure? Please add links in such reports. Thanks.
>>
>> >>>>
>> >>>>
>> >>>> From what I can see, the package is coming from the centos mirros:
>> >>>>
>> >>>> yum -y install ovirt-host
>> >>>> rm -rf /dev/shm/*.rpm /dev/shm/yum
>> >>>> "
>> >>>> 2017-11-01 13:17:53,817::ssh.py::ssh::81::lago.ssh::DEBUG::Command 10946280 on lago-basic-suite-4-1-host-0 returned with 0
>> >>>> 2017-11-01 13:17:53,817::ssh.py::ssh::89::lago.ssh::DEBUG::Command 10946280 on lago-basic-suite-4-1-host-0 output:
>> >>>>  Loaded plugins: fastestmirror
>> >>>> Loading mirror speeds from cached hostfile
>> >>>>  * base: mirror.keystealth.org
>> >>>>  * extras: mirror.lax.hugeserver.com
>> >>>>  * updates: dallas.tx.mirror.xygenhosting.com
>> >>>> Resolving Dependencies
>> >>>> --> Running transaction check
>> >>>> ---> Package ovirt-host.noarch 0:4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos will be installed
>> >>>> --> Processing Dependency: vdsm >= 4.20.0 for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: vdsm-client >= 4.20.0 for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: collectd for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: collectd-disk for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: collectd-netlink for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: collectd-virt for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: collectd-write_http for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: fluentd for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: ovirt-vmconsole for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: ovirt-vmconsole-host for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: rubygem-fluent-plugin-collectd-nest for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: rubygem-fluent-plugin-rewrite-tag-filter for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: rubygem-fluent-plugin-secure-forward for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: rubygem-fluent-plugin-viaq_data_model for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.noarch
>> >>>> --> Processing Dependency: socat for package: ovirt-host-4.2.0-0.0.master.20171019135233.git1921fc6.el7.centos.no
>> >>>>
>> >>>>
>> >>>> Sandro, can you please help to remove this package?
>> >>>
>> >>>
>> >>> Removed from tested repo. It should have been ovirt-host-4.1.0 not 4.2.0. Not sure how it got there passing CI.
>> >>
>> >>
>> >> It got there before the test that was trying to install it was added to OST, so nothing prevented it from being pushed to 4.2.
>>
>> Not sure I fully understand. What test does install it now?
>> All I see is [1], which is a performance cheat/hack, not a proper test,
>> and was merged long ago (1.5 weeks before I started breaking ovirt-host-4.1).
>>
>> In principle, only the engine should instruct to install ovirt-host
>> during *testing*, and only a 4.2 engine. Do we have tests that add
>> a 4.1 host to a 4.2 engine?
>
>
> This was another bug in OST, the installation of ovirt-host should only have be done on 4.2, its also fixed now so future runs should install ovirt-host only on 4.2.

If in "another bug" you mean "the reason for [1]" (it has no bug-url),

then I do not think it's a bug, at least not now (after 4.1 repo is fixed).
Installing ovirt-host-4.1 for a 4.1 host should have exactly the same
principle effect as installing ovirt-host-4.2 for a 4.2 host. The only
difference is that only 4.2 engines /rely/ on this behavior, but if
the patch before [1] was only added as a performance hack, there is no
problem in using it also for 4.1.

[1] https://gerrit.ovirt.org/83448

>
>>
>>
>> [1] https://gerrit.ovirt.org/82461
>>
>> >
>> >
>> > ok, this explain
>> >
>> >>
>> >> If you know that a certain pkg got built with the wrong version, please let us know so we can act on it and perhaps avoid such failures.
>>
>> Sorry, that's my fault, and I didn't think about it.
>> As Sandro wrote in private, I started the work on 4.1
>> by cherry-picking patches for master that I decided we
>> need in 4.1, and only later changed the version to 4.1.
>> In retrospective, I should probably have done this adaptation
>> as the first patch in the series.
>>
>> >
>> >
>> > I've been made aware this morning
>> >
>> >>
>> >>
>> >> The 4.1 build was changed in [1] but before that it was building 4.2 version.
>> >
>> >
>> > Noted that while investigating this morning
>> >
>> >>
>> >>
>> >>
>> >> [1] https://gerrit.ovirt.org/#/c/82875/
>> >>
>> >>>
>> >>>
>> >>>
>> >>>>
>> >>>>
>> >>>> Thanks,
>> >>>> Dafna
>> >>>>
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>>
>> >>> SANDRO BONAZZOLA
>> >>>
>> >>> ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D
>> >>>
>> >>> Red Hat EMEA
>> >>>
>> >>> TRIED. TESTED. TRUSTED.
>> >>>
>> >>>
>> >>> _______________________________________________
>> >>> Devel mailing list
>> >>> Devel@ovirt.org
>> >>> http://lists.ovirt.org/mailman/listinfo/devel
>> >>
>> >>
>> >>
>> >>
>> >> --
>> >>
>> >> Eyal edri
>> >>
>> >>
>> >> MANAGER
>> >>
>> >> RHV DevOps
>> >>
>> >> EMEA VIRTUALIZATION R&D
>> >>
>> >>
>> >> Red Hat EMEA
>> >>
>> >> TRIED. TESTED. TRUSTED.
>> >> phone: +972-9-7692018
>> >> irc: eedri (on #tlv #rhev-dev #rhev-integ)
>> >
>> >
>> >
>> >
>> > --
>> >
>> > SANDRO BONAZZOLA
>> >
>> > ASSOCIATE MANAGER, SOFTWARE ENGINEERING, EMEA ENG VIRTUALIZATION R&D
>> >
>> > Red Hat EMEA
>> >
>> > TRIED. TESTED. TRUSTED.
>> >
>>
>>
>>
>> --
>> Didi
>
>
>
>
> --
>
> Eyal edri
>
>
> MANAGER
>
> RHV DevOps
>
> EMEA VIRTUALIZATION R&D
>
>
> Red Hat EMEA
>
> TRIED. TESTED. TRUSTED.
> phone: +972-9-7692018
> irc: eedri (on #tlv #rhev-dev #rhev-integ)




--

Didi