721648 - tegra-177 needs a new SD card or some quality time at the far end of a sledgehammer

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P3)

Description

[Phil Ringnalda (:philor)]

The SD card is jmaher's idea, from when he noticed it.

The sledgehammer is my idea, from when I added rather more history to https://build.mozilla.org/buildapi/recent/tegra-177 than you would like, and saw that it's been doing nothing but retry and exception for more than two months and 1088 jobs. It's welcome to all those retry ones, but that includes ~450 exceptions, of which I probably starred 300 :(

Comment 1

[[e-mail] (:bhearsum)]

I disabled this slave. Bear, what should we do with it?

Comment 2

[u429623]

open or add to current "needs TLC" bug, similar to Bug 719989, aiui

Comment 3

[Phil Ringnalda (:philor)]

Didn't successfully disable it, I'm afraid - 16 jobs so far today.

Comment 4

[Nick Thomas [:nthomas] (UTC+12)]

Ran the stop_cp script on this one. check.sh says that clientproxy and the buildbot slave are offline.

Comment 5

[u429623]

For some reason, the slave that had been running didn't stop, and was still accepting jobs. It finally stopped about 7 hours ago, according to foopy17:/build/tegra-177/twistd.log:
  2012-01-29 00:49:44-0800 [-] Server Shut Down.

(I think this can happen if the slave's pid file gets removed before the process terminates, as check.sh and others trust the pid file.)

Comment 6

[u429623]

(In reply to Hal Wine [:hwine] from comment #5)
Oh, bother. disregard - Nick's stop_cp.sh did the job as intended. I misread the timestamps.

Comment 7

[Rail Aliiev [:rail]]

According to the tegra dashboard this tegra is up and running.