You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
So we've been having an issue in our production servers, all of a sudden the number of web requests gets really high, enough that our prod servers crash and restarts. I haven't been able to find exactly why it's happening (cause I can't replicate it) but I have noticed locally a couple times that "OnReconnect" method gets called hundreds of time in a row...BY ONE USER. I haven't been able to find why it's happening or how to consistently replicate it. any ideas?
The text was updated successfully, but these errors were encountered:
After implementing a simple reconnect "strategy" that creates a new hub after a disconnect, I saw that this component produces 100s of error messages in chrome console if the signalR endpoint is not available (e.g. when the UI is started in dev mode without a backend).
I'm assuming this is the same problem as this one here. It lead primarily to client-side issues for me because the chrome tab freezes after 2 minutes.
Based on this approach here (https://stackoverflow.com/a/23407156) I'm now not creating a new hub every 5 seconds but instead trigger a HTTP request to the signalR endpoint. Only if that one is successful (status 200 on GET/HEAD request to /signalr/negotiate) I will create a new hub.
I have not tested it extensively but on the first tries it feels like a big improvement to not have to kill the chrome tab after a signalr connection loss.
So we've been having an issue in our production servers, all of a sudden the number of web requests gets really high, enough that our prod servers crash and restarts. I haven't been able to find exactly why it's happening (cause I can't replicate it) but I have noticed locally a couple times that "OnReconnect" method gets called hundreds of time in a row...BY ONE USER. I haven't been able to find why it's happening or how to consistently replicate it. any ideas?
The text was updated successfully, but these errors were encountered: