Bugs item #3604147, was opened at 2013-02-11 11:12 Message generated for change (Comment added) made by bogdan_iancu You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=1086410&aid=3604147&group_id=232389
Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: modules >Group: 1.9.x Status: Open Resolution: None Priority: 5 Private: No Submitted By: Seth Schultz (sethschultz) >Assigned to: Bogdan-Andrei Iancu (bogdan_iancu) Summary: load_balance in failure_route Initial Comment: Hello, I call load_balance in my failure_route for sip_codes "(408)|[56][0-9]{2}". This works, however, once all of the destinations have been exhausted, it continually tries to load balance the call to the last address, instead of returning false. Please see the snippet below. failure_route[1] { ... if (t_was_cancelled()) { exit; } if (is_method("INVITE") && $tt == NULL && t_check_status("(403)|(404)|(408)|([56][0-9]{2})")) { if(load_balance("1", "pstn", "1")) { ... #this code gets continually called with the last address in the pool instead of hitting the else condition below } else { send_reply("$rs","$rr"); exit; } } ... } ---------------------------------------------------------------------- >Comment By: Bogdan-Andrei Iancu (bogdan_iancu) Date: 2013-02-16 01:30 Message: Hi Seth, Tested with trunk version (and 1.9) with this script: failure_route[lb] { if (load_balance("1","pstn","1") ) { xlog("L_DEBUG", "\n===DESTINATION IS $du===\n"); t_on_failure("lb"); t_relay(); } } route { load_balance("1","pstn","1"); xlog("L_DEBUG", "\n===DESTINATION IS $du===\n"); t_on_failure("lb"); t_relay(); exit; } I used 2 destinations in LB and it stopped after the second failure . See here the logs : Feb 16 11:28:52 [24870] DBG:load_balancer:do_load_balance: found requested (0) resource pstn Feb 16 11:28:52 [24870] DBG:load_balancer:do_load_balance: destination <sip:127.0.0.10> selected for LB set with free=100 (max=100) Feb 16 11:28:52 [24870] DBG:load_balancer:do_load_balance: destination <sip:127.0.0.15> selected for LB set with free=100 (max=100) Feb 16 11:28:52 [24870] DBG:load_balancer:do_load_balance: winning destination <sip:127.0.0.10> selected for LB set with free=100 ..... Feb 16 11:28:57 [24877] DBG:load_balancer:do_load_balance: found requested (0) resource pstn Feb 16 11:28:57 [24877] DBG:load_balancer:do_load_balance: sequential call of LB - previous selected dst is 1 Feb 16 11:28:57 [24877] DBG:load_balancer:do_load_balance: skipping destination <sip:127.0.0.10> (used=1 , disabled=0) Feb 16 11:28:57 [24877] DBG:load_balancer:do_load_balance: destination <sip:127.0.0.15> selected for LB set with free=100 (max=100) Feb 16 11:28:57 [24877] DBG:load_balancer:do_load_balance: winning destination <sip:127.0.0.15> selected for LB set with free=100 ..... Feb 16 11:29:03 [24877] DBG:load_balancer:do_load_balance: found requested (0) resource pstn Feb 16 11:29:03 [24877] DBG:load_balancer:do_load_balance: sequential call of LB - previous selected dst is 2 Feb 16 11:29:03 [24877] DBG:load_balancer:do_load_balance: skipping destination <sip:127.0.0.10> (used=1 , disabled=0) Feb 16 11:29:03 [24877] DBG:load_balancer:do_load_balance: skipping destination <sip:127.0.0.15> (used=1 , disabled=0) Feb 16 11:29:03 [24877] DBG:load_balancer:do_load_balance: no destination found Please consider the attached patch for more debugs - run with it and post the logs generated by the LB module Regards, Bogdan ---------------------------------------------------------------------- Comment By: Seth Schultz (sethschultz) Date: 2013-02-11 11:35 Message: I should note that I have tried it with both the 0 and the 1 algorithm. Both of them fail in the same way. Thanks, Seth ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=1086410&aid=3604147&group_id=232389 _______________________________________________ Devel mailing list Devel@lists.opensips.org http://lists.opensips.org/cgi-bin/mailman/listinfo/devel