[Feature Request] Restart all failed jobs

[yakanechi]

What happened:
Multiple builds that should succeed may fail due to network load, scm failure or etc....
When we want to restart failed job after a time interval, the current UI requires restarting jobs one by one, which is very time consuming.

This is duplicative of part of #2077 proposal

What you expected to happen:

• Allow only failed jobs to be restarted together.
• Alternatively, at restart time, users can select which jobs to restart, e.g., by checkbox.

How to reproduce it: