The revised SipRefreshManager has code to recover from transient errors when it is re-SUBSCRIBEing: It waits for a few seconds and sends the re-SUBSCRIBE again (with incremented CSeq). Currently, only two failure responses are treated as transient errors, 408 (Timeout) and 482 (Looped Request). (482 usually results from a delayed response causing an upstream proxy to re-send the request via a different transport.) (If the repeated re-SUBSCRIBE attempts fail, SipRefreshManager gives up on refreshing the subscription, whereupon SipSubscribeClient starts a new subscription.)
I am considering adding 500 and 503 to the list of transient errors. 503 is used only for reporting overload conditions at the immediately downstream node. 500 is used for several purposes, but one of them is reporting overload conditions at nodes that are further downstream. Comments? Dale _______________________________________________ sipx-dev mailing list [email protected] List Archive: http://list.sipfoundry.org/archive/sipx-dev Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev sipXecs IP PBX -- http://www.sipfoundry.org/
