canonical-ci-engineering team mailing list archive

[Larry Works <larry.works@xxxxxxxxxxxxx>]

No, you can put it back into operation if you want. I put in to shutdown
mode immediately after starting it so I could make sure all of the nodes
were started.

One note to put out there; the CDU that all of the kernel SRU/smoke test
systems (alkaid, phact, rukbah, tarf, onza, onibi, zmeu and zuijin) are
connected to for power is not our normal ServerTech CDU. The only
outlets above the rack those systems are hosted in were 30amp outlets
which we cannot use with out regular CDUs. Instead that rack is
supported by an APC PDU. We did not get to configure those two PDUs as
they take a different type (6 pin, probably rj-25) of connector that the
normal 8 pin rj-45 that our normal console cables have. Once we track
down one of those we will configure those PDU so that they are network
accessible. It will possible also involve modifying the jenkins jobs or
cobbler profiles to use the correct connection method.

The Dell stack, HP stack and Lenovo stacks used by various factions
(server team, cloud, MaaS) should be online for those folks to start
back into their testing. I still need to turn the kernel team systems
back on and make sure their jenkins instance is starting up and
reachable over the network.

I will work on getting the other jenkins instances (s-jenkins,
q-jenkins, m-jenkins and dev-jenkins) up and running tomorrow afternoon
but probably won't have them all up until Friday. Also note, we have two
different cobbler instances from the original labs; one from jiufeng
(replaced by tachash) and the other from magners-orchestra (to be
replaced by jatayu). I will consolidate that all into a single cobbler
instance in the very near future (migrating all of the systems and
profiles) so we can just use the DNS alias cobbler.ubuntu-ci.

I also need to consolidate  the two nagios and munin servers into a
single server, move m-jenkins to a different server, move dev-jenkins to
a physical server and Rick and I need to fix naartjie. Phones and
tablets will be moved tomorrow as should the four Acer Veritons used for
bootspeed testing.

But right now I need to get a little sleep. I have about three hours
before I need to get up and get ready to catch my flight back home.

~w
 
On 11/14/2013 12:01 AM, Paul Larson wrote:
> Fixed.
> There was a missing "/" at the end of the HUDSON_MASTER line in
> auto-upgrade-tester-jenkins-slave. Starts up fine now. I left jenkins
> in "shutting down" mode, since I assume you have it that way for a reason.
>
>
> On Wed, Nov 13, 2013 at 6:01 PM, Larry Works
> <larry.works@xxxxxxxxxxxxx <mailto:larry.works@xxxxxxxxxxxxx>> wrote:
>
>     All,
>
>     The jenkins instance that used to reside at http://10.98.0.1:8080 (or
>     http://10.189.74.2:8080), d-jenkins.ubuntu-ci, is now back online and
>     ready for the tires to be kicked. It can be reached at it's new
>     address
>     of http://10.98.3.6:8080 or http://d-jenkins.ubuntu-ci:8080 if you are
>     using IS' internal DNS.
>
>     There is only one hang up atm, the node aldebaran-upgrade doesn't seem
>     to want to connect. If someone who has access to aldebaran
>     wouldn't mind
>     poking around at that it would be greatly appreciated. I still have a
>     few more systems to connect before the day is gone and I'm running
>     very
>     short on time.
>
>     ~w
>
>     --
>     Mailing list: https://launchpad.net/~canonical-ci-engineering
>     <https://launchpad.net/%7Ecanonical-ci-engineering>
>     Post to     : canonical-ci-engineering@xxxxxxxxxxxxxxxxxxx
>     <mailto:canonical-ci-engineering@xxxxxxxxxxxxxxxxxxx>
>     Unsubscribe : https://launchpad.net/~canonical-ci-engineering
>     <https://launchpad.net/%7Ecanonical-ci-engineering>
>     More help   : https://help.launchpad.net/ListHelp
>
>