807646 - (tegra-093) tegra-093 problem tracking

Status: RESOLVED FIXED

(Infrastructure & Operations Graveyard :: CIDuty, task)

Description

[[e-mail] (:bhearsum)]

Ping is failing. Trying a pdu reboot on it.

Comment 1

[[e-mail] (:bhearsum)]

WTF. The dashboard  claimed the ping check was failing and that it was down, but it was actually running a job when I PDU rebooted it. This tegra seems fine, modulo the purple build as a result of me interrupting the job.

Comment 2

[Justin Wood (:Callek)]

(In reply to Ben Hearsum [:bhearsum] from comment #1)
> WTF. The dashboard  claimed the ping check was failing and that it was down,
> but it was actually running a job when I PDU rebooted it. This tegra seems
> fine, modulo the purple build as a result of me interrupting the job.

Just to comment somewhere for now, I'll use here:

The dashboard at present is a *snapshot in time* with a relatively length delay. Where it will take a tegras status at a given moment, up/down does it have an error.flg etc. and report it.

That then gets generated by the dashboard into its table, which takes a few minutes, then gets rsync'd to the server that shows us the website, so there is a slight delay to it.

Meanwhile tegras in real jobs frequently go up/down while installing/uninstalling/etc. things, so its possible to see a DOWN at any given moment, for response as to DOWN on a tegra in the dashboard, see the "bar" at the right of the row, which will show if its been offline for a handful of checks in a row, or what.

Comment 3

[[e-mail] (:bhearsum)]

SUTAgent not present;

time for recovery

Comment 4

[Phil Ringnalda (:philor)]

The thing that confuses me about all that is that there's no way to reconcile any of that with https://secure.pub.build.mozilla.org/buildapi/recent/tegra-093?numbuilds=100 - it did not do a purple build on the morning of the 1st, it set retry on one at 04:58, but at 07:12 it was happily doing a run it had started at 07:09 and finished, green, at 07:42. At 06:26 on the 2nd, when it was "time for recovery" it had finished a green job 24 minutes before, and was 6 minutes away from taking another and doing it green. Then it is rumored to have been recovered sometime before 16:47 on the 2nd, which could conceivably have happened between 11:49 and 12:25, because that's the only time when it went more than 5 minutes between jobs. That's also the dividing line between when it went from being a pretty darn good tegra to being a pretty darn broken one. 25 builds since then, 52% green; the 25 builds before then were 84% green.

My best wild and ridiculous guess is that there are two slaves that both think they are named tegra-093.

Comment 5

[Justin Wood (:Callek)]

Ran ./stop_cp.sh

Comment 6

[Justin Wood (:Callek)]

Brought back to life

Comment 7

[Justin Wood (:Callek)]

Actually needed cpr, sending over to recovery for a reimage.

Comment 8

[Justin Wood (:Callek)]

No jobs taken on this device for >= 7 weeks

Comment 9

[Justin Wood (:Callek)]

(mass change: filter on tegraCallek02reboot2013)

I just rebooted this device, hoping that many of the ones I'm doing tonight come back automatically. I'll check back in tomorrow to see if it did, if it does not I'll triage next step manually on a per-device basis.

---
Command I used (with a manual patch to the fabric script to allow this command)

(fabric)[jwood@dev-master01 fabric]$  python manage_foopies.py -j15 -f devices.json `for i in 021 032 036 039 046  048 061 064 066 067 071 074 079 081 082 083 084 088 093 104 106 108 115 116 118 129 152 154 164 168 169 174 179 182 184 187 189 200 207 217 223 228 234 248 255 264 270 277 285 290 294 295 297 298 300 302 304 305 306 307 308 309 310 311 312 314 315 316 319 320 321 322 323 324 325 326 328 329 330 331 332 333 335 336 337 338 339 340 341 342 343 345 346 347 348 349 350 354 355 356 358 359 360 361 362 363 364 365 367 368 369; do echo '-D' tegra-$i; done` reboot_tegra

The command does the reboot, one-at-a-time from the foopy the device is connected from. with one ssh connection per foopy

Comment 10

[Justin Wood (:Callek)]

had to cycle clientproxy to bring this back

Comment 11

[[e-mail] (:bhearsum)]

Hasn't run a job in 15 days, 16:05:21

Comment 12

[[e-mail] (:bhearsum)]

recovery didn't help, dunno what to do

2013-04-05 06:20:43 tegra-093 p    online   active  OFFLINE :: error.flg [Automation Error: Unable to connect to device after 5 attempts]

Comment 13

[Justin Wood (:Callek)]

Sending this slave to recovery
-->Automated message.

Comment 14

[Justin Wood (:Callek)]

Sending this slave to recovery
-->Automated message.

Comment 15

[Vinh Hua [:vinh]]

flashed and reimaged

Comment 16

[Justin Wood (:Callek)]

Replace sdcard please

Comment 17

[[e-mail] (:bhearsum)]

back in production

Comment 18

[[e-mail] (:bhearsum)]

2014-01-16 10:24:20 tegra-093 p    online   active  OFFLINE :: error.flg [Automation Error: Unable to connect to device after 5 attempts] 

pdu reboot didn't help

Comment 19

[John Pech [:jpech-lbc]]

SD card have been wiped and tegra has been re-imaged.

Comment 20

[[e-mail] (:bhearsum)]

(In reply to John Pech [:jpech-lbc] from comment #19)
> SD card have been wiped and tegra has been re-imaged.

can we try again?

Comment 21

[Van Le [:van]]

SD card formatted, tegra reimaged and flashed.

[[e-mail]3 ~]$ fping tegra-093.tegra.releng.scl3.mozilla.com
tegra-093.tegra.releng.scl3.mozilla.com is alive

Comment 22

[Justin Wood (:Callek)]

Taking jobs again