I remember working the support queue _before_ this automatic re-routing mitigation system went in and it was a lifesaver. Having to run over to SRE and yell "look! look at grafana showing this big jump in 522s across the board for everything originating in ORD-XX where the next hop is ASYYYY! WHY ARE WE STILL SENDING TRAFFIC OVER THAT ARRRGHH please re-route and make the 522 tickets stop"
it's cool to see something large enough that the auto-healing mechanisms weren't able to handle it on their own, though shoutout to whoever was on the weekend support/SRE shift; that stuff was never fun to deal with when you were one of a few reduced staff on the weekend shifts
it's cool to see something large enough that the auto-healing mechanisms weren't able to handle it on their own, though shoutout to whoever was on the weekend support/SRE shift; that stuff was never fun to deal with when you were one of a few reduced staff on the weekend shifts