Last post Nov 02, 2013 03:59 PM by davidfowl
Oct 30, 2013 09:02 PM|jkarcherla|LINK
Envirnoment : Windows 2008 R2, IIS 7.5, SignalR 0.5.3, SignalR.Redis.0.6.0
We seem to be having issues in our staging envirnoment integrating CDN Provider with SignalR. Futher looking into the details of the client level logs and server events, it looks like default transport is falling from SSE , Forever frame to Long Polling
(Potentially due to content buffering - reading through a existing article). When this happens, none of the server to client hub communications are working.This is the same behavior without CDN in the communication handshake in other envirnoments whenever
transport is falling back to Long Polling.
Is there any known issues with LongPolling and the current version of SignalR packages we are using ? Is anyone else experiencing this issue. Also, are there any logs that would indicate reconnection when transport fallsback to long polling.
Chrome --> SSE Fails switch to --> Long Polling --> Doesn’t fire post event
Firefox --> SSE Fails switch to --> Long Polling --> Doesn’t fire post event
IE --> Forever frame Fails switch to --> Long Polling --> Doesn’t fire post event
There are few HTTP 500 errors in IIS logs around these events when transport is falling back to long polling. Any additional direction would help.
Nov 02, 2013 03:59 PM|davidfowl|LINK
You should upgrade to a newer version of SignalR.