canonical-ci-engineering team mailing list archive

[Vincent Ladeuil <vila+ci@xxxxxxxxxxxxx>]

>>>>> Evan Dandrea <evan.dandrea@xxxxxxxxxxxxx> writes:
    > FYI
    > Thanks for putting this together, Larry.


    > ---------- Forwarded message ----------
    > From: Larry Works <larry.works@xxxxxxxxxxxxx>
    > Date: 22 November 2013 06:37
    > Subject: Lab update for 21 November
    > To: Evan Dandrea <evan.dandrea@xxxxxxxxxxxxx>


    > Evan,

    > Here is what we managed to accomplish today:

    >   - The DMZ network is entirely up and accessible through the VPN. Both
    > drude and rabisu are in DNS. The alias dmz-jenkins has also been added
    > to point people at the jenkins server there. All of the jenkins slave
    > nodes are connected and tests were running.

Confirmed with jibel, vanguard ticket created to get accounts created
there for whoever need them (ev created one for me on drude) as I don't
know who need them.

That can probably be delayed until we deploy wossname which give us (ci)
accounts where we need ?

    >   - The slowness of the outbound connections from within the lab has
    > been resolved. Rick applied a configuration change to the existing set
    > up (as in we did not put a server in place to act as a router) and the
    > download speeds increased from ~250KB/s to between 10+MB/s on the low
    > side and 37+MB/s on the high side.

\o/ 

    >   - The Acer Veritons (bootspeed systems) are connected to network, CDU
    > and KVM. We're not seeing anything from them on the KVM except a cursor
    > across a black screen; don't know if it has something to do with their
    > configuration of if they need to connect to a server to pull something
    > additional. Paul and/or Max can poke at them whenever they get a chance
    > to see what else we may need to do to get them back into the test pool;
    > all four are reachable over the network.

May be not related but I looked at a verizon node on q-jenkins and
noticed a /var/lib/jenkins for the remote FS root.

I didn't act on it as I was working on something else and had no idea
how to verify a fix but someone may be interested.

    >   - Nagios is up and running within the lab. It can be reached at
    > http://nagios.ubuntu-ci/nagios3. There are quite a few things showing as
    > either critical or warning but most of those are probably configuration
    > related; I added several more tests for all of the systems and may have
    > a parameter or two slightly off. I will poke at those some more tomorrow.

I couldn't find anything meaningful there probably because of the 'It
appears as though you do not have permission to view process
information...' messages I got here and there (connected with U/u)

Can I haz a demo and a crash course on how to add nagios alerts (I won't
mind a playbook instead of a demo ;).

     Vincent (so glad we got so much fixed in the last hours/days)