what bhearsum did to get a bunch of tegras online
I attacked the pool of broken/dead Tegras this week, and with a lot of guidance from Callek, managed to get a bunch of them back online. The Tegra wiki page (https://wiki.mozilla.org/ReleaseEngineering:How_To:Android_Tegras) isn't very helpful in terms of mapping "state it's in" -> "action i need to take", so I wanted to let you all know the simple steps I took:
- Look for Tegras with alerts on http://admin1.infra.scl1.mozilla.com/nagios/cgi-bin/status.cgi?host=all&servicestatustypes=28&hoststatustypes=15&serviceprops=270346
- Find the Tegra on http://mobile-dashboard1.build.mtv1.mozilla.com/tegras/
- Click the "tegra-xxx" link to find the Tegra's Buildbot page
- If the Tegra is offline, try to PDU reboot it (https://wiki.mozilla.org/ReleaseEngineering:How_To:Android_Tegras#Reboot_a_Tegra_through_the_PDU)
- If the Tegra isn't online 30min later, open its bug and block it against the "tegra-recovery" bug. You may need to file one or both of these.
- If that doesn't work, talk to Callek ;).
And that's it! More than half the Tegras I touched came back online either through the PDU reboot or IT recovery. I implore you all to keep on top of Tegras when you're on buildduty - it's not as hard as it seems.