On Fri, 2009-09-18 at 08:50 -0400, Joly, Robert (CAR:9D30) wrote: > > On Fri, 2009-09-18 at 07:39 -0400, Alfred Campbell wrote: > > > We have some folks over in Japan using sipXecs and two issues have > > > come > > > up: > > > 1. The autoattendant was failing replication with a > > Japanese character > > > named AA. Once I removed this it seemed to replicate. Just > > guessing > > > this is a bug? > > > 2. The proxy is suffering issues as well and I see users > > with Japanese > > > First/Last names however the User ID is numeric. The proxy > > is failing > > > to start with no good reason (no detail available in GUI). > > The only > > > thing in the proxy log are errors around CSE creation. > > > > > > Any thoughts? > > > > Do we have a snapshot? It would show the contents of the > > relevant configuration files. > > There is little info in the logs but the INSERT statements that the CSE > observer trips on suggests that the SIP signaling contains fields with > characters that may not be properly escaped. I'd like to get a sniffer > trace taken from the sipXecs box so that I can take a look at the > signaling as it flies on the wire using the following command 'tcpdump > -n -nn -s 0 -i any -w trace.cap'.
Looking at sipXproxy.log, I see that they are not logging at INFO level. Have these people not been informed how to obtain diagnostic information? sipXproxy.log does, however, log the failed database operations. If I pull it into Emacs without character-set translation, I see text strings which are not valid UTF-8 encodings. Since the SIP messages aren't logged, I can't tell where those string come from, but I suspect that they are from the SIP messages, and that the phones are generating incorrect UTF-8 for the non-Latin characters. This may portend difficulties, because it's quite possible that sipXconfig is creating config files for the phones that contain non-UTF-8 byte strings. Or maybe the users are entering them through the phones' web interfaces. In any case, there may be some serious difficult in ensuring that non-Latin characters are always translated into UTF-8 properly. 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/
