request a build and invite user crash if Redis is down
when Redis is restarted, builds requested during downtime are never restarted
Either we make the dependency on Redis stronger and block everything if Redis is down (which I guess would make sense), or we rescue those errors and add some worker or something to find what need to be re-enqueued when Redis restarts
Several things should be improved
Either we make the dependency on Redis stronger and block everything if Redis is down (which I guess would make sense), or we rescue those errors and add some worker or something to find what need to be re-enqueued when Redis restarts