472913 - "MacOSX 10.4 comm-central-calendar ltn nightly" not building

Status: RESOLVED FIXED

(Mozilla Messaging Graveyard :: Server Operations, defect)

Description

[Mark Banner (:standard8)]

On its build of Jan 08 07:24, "MacOSX 10.4 comm-central-calendar ltn nightly" lost its connection to the slave, for some reason it hasn't built since (though the dep builder has).

Comment 1

[Philippe M. Chiasson (:gozer)]

restarted buildbot slave on that host, should probably fix it. The buildbot master restarted a few times recently, and sometimes, as a result, clients get confused.

Comment 2

[Philippe M. Chiasson (:gozer)]

In this case, the usual thing to do would have been to respin the nightly by telling the irc bot (sunbuild) to respin.

Note the dep/nightly builders are one and the same buildbot builder/host, just different configurations to build.

Comment 3

[Mark Banner (:standard8)]

(In reply to comment #2)
> In this case, the usual thing to do would have been to respin the nightly by
> telling the irc bot (sunbuild) to respin.

I didn't do that as the buildbot master said there was already a build pending, but it just wasn't kicking it off.

Comment 4

[Dan Mosedale (:dmosedale, :dmose)]

Sadly, no new builds have appeared since the 7th.  Furthermore, attempting to look at buildbot yields the following error: 

Proxy Error

The proxy server received an invalid response from an upstream server.
The proxy server could not handle the request GET /buildbot/production/waterfall.

Reason: Error reading from remote server

Apache/2.2.3 (CentOS) Server at build.mozillamessaging.com Port 80

The bot that I see in #calendar seems to be called calbuild rather than sunbuild, and telling it "respin" in a /msg doesn't seem to cause anything to happen.

Suggestions?

Comment 5

[Mark Banner (:standard8)]

It appears all the Sunbird mac machines are broken. I've poked them a couple of times with clobbers, but I've got no idea what is happening. cc'ing ause as he may know/have access to them.

Comment 6

[ause]

last command shown in the logs (hdutil) seems to run successful. the expected output (hybrid.dmg) exists.
the twisted.log shows

-------
...
2009/01/13 03:12 PST [-] command timed out: 1200 seconds without output, killing pid 5609
2009/01/13 03:12 PST [-] trying os.kill(-pid, 9)
2009/01/13 03:12 PST [-]  signal 9 sent successfully
2009/01/13 03:12 PST [-] Traceback (most recent call last):
...
-------

so it might be a simple timeout issue that probably have to be handled on the bot master.

Comment 7

[ause]

found a hanging launchd eating up cpu time. rebooting the machine and starting the bot from the gui gave me gree builds again.

Comment 8

[ause]

all builds green