Decrease shared runner job timeout to 1 hour before the migration
Noted in 2018-07-12 failover attempt
This is a step in failover.md:
# T minus one day
1. [ ] **PRODUCTION ONLY** **UNTESTED** 🔪 {+ Chef-Runner +}: Update GitLab shared runners to expire jobs after 1 hour
We don't have a merge request or procedure for this yet, and we need one.
Since we have shared runners in staging, I think we can change this so it's no longer a production only step too. This will allow us to test it and remove **UNTESTED**
as well
@tmaczukin do you have insight on how we'd do this? Is there a change @ahmadsherif or @alejandro can apply to make this happen?