tags:

views:

217

answers:

1
+1  A: 

I always put a "heartbeat" message in my long-running WCF services. Then you can set Type #1 timeouts to a low value (2-3 times the heartbeat call frequency), and Type #2 timeouts become obvious.

Stephen Cleary
Can you be more specific? It sounds like you are either able to single out Type #1 timeouts (which I'm not), or you are somehow able to have long-running WCF services that work when ALL of your timeouts are set to small values. How do you achieve this?
Greg
Actually, after re-reading your question, I'd say that a callback is all that you may need (so you'd end up with a one-way request that is at some future time responded to by the one-way callback). The heartbeat would still be necessary to prevent idle timeouts while waiting for the request.
Stephen Cleary
I think you are assuming that the one-way request is successful and the server-side code (the processing of the request) is the culprit. See the clarification to Andrew's comment in my edit - I don't think this is the problem. I think the one-way request is just as likely to fail as the full-fledged operation.
Greg