Can you please read the contributing document and resubmit these patches https://github.com/FreeRADIUS/freeradius-server/blob/master/doc/developer/co... It's not appropriate to send large patches to the developers list. It makes it more difficult to review the patches in context, and means we can't offer line by line feedback. Arran Cudbard-Bell <a.cudbardb@freeradius.org> FreeRADIUS Development Team FD31 3077 42EC 7FCD 32FE 5EE2 56CF 27F9 30A8 CAA2

Jon Sexson wrote:

These patches add support for several options allowing configuration of reduced proxy response times on low-latency links. They are being submitted so that we can migrate from our current platform to FreeRADIUS.

That's always a good idea. :)

Due to SLAs with our Tier-1 Carriers, we require support for a higher precision response window and a reduced initial request transmission delay to be able to configure sub-second proxy response times.

Hmm... I'm not sure what patch #4 is for. The server doesn't have a transmission delay. The delay you're changing is just a timer for the main thread to service the requests. If the response_window is small, then the delay should be set below response_window, so that the proxied request will be timed out. Is that what the patch is doing?

We need to override response window for some Tier-1 carrier clients so the proxy server fails over to next home server faster to satisfy required response times of their tighter SLAs.

That makes sense.

Sometimes a given sgsn has a very short-lived response time problem which doesn't deserve putting the home server into the zombie mode, so we'd like to configure tolerance to a certain amount of response window misses, which "max_response_timeouts" allows us to do.

That makes sense, too. I think the name is a little confusing, but I'll look at that. Overall, the patches look reasonable and simple. Alan DeKok.

Have your upstream looked at status-server ? alan -- Sent from my Android device with K-9 Mail. Please excuse my brevity.