1106974 - slaveapi never successfully ssh reboots any talos-linux64-ix slave

Status: RESOLVED WONTFIX

(Release Engineering :: General, defect)

Description

[Phil Ringnalda (:philor)]

Look at the reboot history for any talos-linux64-ix slave (https://secure.pub.build.mozilla.org/builddata/reports/slave_health/slave.html?class=test&type=talos-linux64-ix&name=talos-linux64-ix-018 is handy because it has lots of reboots, having been dead for a couple of weeks, but even live slaves like https://secure.pub.build.mozilla.org/builddata/reports/slave_health/slave.html?class=test&type=talos-linux64-ix&name=talos-linux64-ix-070 will do), and you'll see that they never ever successfully ssh reboot.

Perhaps related, for those rare periods when slaverebooter is alive and trying to graceful them, they never ever show it whatever it is that it wants to see to call a graceful successful.

My gut feeling was that the same was not true of talos-linux32-ix, but looking at the ones which currently have reboot history (not many, we recently threw it on the floor), it looks like maybe they never ssh reboot or graceful either.

My theory is always that slaverebooter trips over its own feet by having too many difficult things to deal with, and since those two pools have massive over-capacity and thus at almost any time will have 100 slaves >5hrs idle, if my theory is actually correct this could be the reason why slaverebooter is almost never running anymore.

Comment 1

[Phil Ringnalda (:philor)]

Not fixed by bug 1124843 like I hoped it might be.