The idea is to have hosts still running in powercycle tasks so members of a Build team will be able to have live machines to diagnose the root cause of ssh failures in powercycle.

Run 20 powercycle tasks with extended host lifetime on ~ STM Daily Cron for easier debugging:

• run 20 powercycle_smoke_no_compile tasks - use generate.tasks
• extended host lifetime by increasing exec_timeout_secs and timeout_secs to 7 days in task config (do not modify the current powerycle tasks that will still continue running on other buildvariants as is)
• set timeout_teardown_secs to 7 days in host.create command
• add a sentinel task that will system fail when any of the tasks above will be running for more than 2 hours (current exec_timeout_secs of powercycle tasks) - use Evergreen API to get other tasks start time