Hi guys, we have a problem regarding High Availability cluster. two sipxecs servers are connected through HA cluster, and each one of them has their own trunk. our phones are registered randomly on both servers.
although the high availability works (when one server goes down, the second takes over all the calls immediatley), it seems that incoming calls don't always get through. to be more precise, exactly every second call goes through. by checking the logs, it seems that the remaining sipxecs server (one is down) does load balancing, that is, once it processes the call himself, once it sends the call to the other server (although it is down), thinking it will process the call. while this isn't a problem while both servers are up, it kinda makes HA system pointless, because we want to have all the calls going through no matter what. by the looks of it, one would say this is the way the sipxecs was designed to function, because this thing always happens. it even happened in the 4.2.1 version (now we are using 4.4.0). we want to be sure every call, outgoing and incoming, will get through. after all, that is why we are making the HA cluster in the first place. does anybody has any idea how to stop this load balancing of incoming calls? additional information: it didn't matter if the phone was registered with the server that was still up, or with the server that went down we would get the problem described above. p.s. we also had a problem with sipxbridges in HA. although these two problems seem unrelated, for more information on the other one check: http://track.sipfoundry.org/browse/XX-9582?focusedCommentId=55113#action_55113 Regards, -- Irena Dolovčak
_______________________________________________ sipx-users mailing list sipx-users@list.sipfoundry.org List Archive: http://list.sipfoundry.org/archive/sipx-users/
