User-493622830 posted
We have been using Redis cache for scale out for sometime now well over a year. Our SignalR code hasn't changed for almost as long. About 4 months ago we started getting 503 errors when scaling.
Not sure what changed that when we scale nodes up or down SingalR often throws Service Unavailable and our clients get a white screen of death.
We upgraded from ver 2.2.0 to 2.2.2 today and it didn't help.
Any help would be appreciated.