[Serusers] failure route and ERROR: t_should_relay: status rewrite by UAS: stored: 408, received: 487

[Jiri Kuthan]

Alan,

thanks -- the call-flow looks fine, I looked at the source code and it seems
to be just a stupid warning. 

-jiri

At 10:29 PM 12/19/2003, Alan Crosswell wrote:
>Jiri,
>
>The tcpdump is at http://www.columbia.edu/~alan/ser.tcpdump
>Here's the corresponding ser syslog:
>
>Dec 19 16:16:38 ren kernel: eth0: Setting promiscuous mode.
>Dec 19 16:16:38 ren kernel: device eth0 entered promiscuous mode
>Dec 19 16:16:56 ren /usr/sbin/ser[5693]: ACC: call missed: method=INVITE, i-uri=sip:17548 at 128.59.39.127;user=phone;phone-context=private, o-uri=sip:alan at 68.198.11.112:5060, call_id=798FB767-319F11D8-B06B89D2-F2E2B15E at 128.59.59.242, from="43754" <sip:43754 at 128.59.59.242>, code=408 Request Timeout
>Dec 19 16:16:56 ren /usr/sbin/ser[5693]: redirection to voicemail
>Dec 19 16:16:56 ren /usr/sbin/ser[5676]: ERROR: t_should_relay: status rewrite by UAS: stored: 408, received: 487
>Dec 19 16:16:56 ren /usr/sbin/ser[5681]: ERROR: t_should_relay: status rewrite by UAS: stored: 408, received: 487
>Dec 19 16:17:19 ren kernel: device eth0 left promiscuous mode
>
>The timestamp in the log corresponds to the 487's coming back from the 7960 phones.
>
>Asterisk is running on 128.59.39.163 in this trace...  (Ignore the couple of transactions from some failed REGISTERs for phones that we haven't configured into ser yet.)
>
>/a
>
>Alan Crosswell wrote:
>>
>>Jiri Kuthan wrote:
>>
>>>At 08:58 PM 12/19/2003, Alan Crosswell wrote:
>>>
>>>>I'm trying to do failure route to voicemail (which is working) but this error is logged:
>>>>
>>>>ERROR: t_should_relay: status rewrite by UAS: stored: 408, received: 487
>>>>
>>>>I googled this and see that this came up last in October (http://lists.iptel.org/pipermail/serusers/2003-October/002921.html) but I don't see any evidence of a solution in the thread.
>>>
>>>
>>>
>>>I think it is a msitake in log severity and it should be softened to a warning.
>>>What possibly happens is that a UAS times out at the same time when the call
>>>is canceled, and it sends 408 timeout and milisecond later a 487 in response
>>>to CANCEL... What UAS did you use?
>>
>>Both UAs are cisco 7960's (POS3-04-3-00).  I don't see why either would time out.  It's only 15 seconds for fr_inv_timer.  I also just tested with another singly-registered user and the error still happens so it's not a multiple dset thing.
>>
>>>
>>>Generaly, such issues are easier to capture if you keep track of all SIP passing
>>>
>>>your site. I think that's a very good practice. If you get me message dumps,
>>>I will try to verify what really happened.
>>
>>Part of the problem is since asterisk is running on localhost I won't be able to get captures for that.  I suppose I could move it to another machine just to help get some tcpdumps.  I'll do that and send along.
>>
>>>
>>>Regarding voicemail: SEMS supports voicemail2email. As the number of people
>>>interested in IVR grows, there is now a first IVR version of voicemail too
>>>-- but this piece of work has never been really tested.
>>I'll be happy to test it, but you need one of the canned IVR messages to be "Weasels have eaten our phone system" before you'll see a mass exodus from asterisk to sems;-)
>>
>>>-jiri 
>>
>>_______________________________________________
>>Serusers mailing list
>>serusers at lists.iptel.org
>>http://lists.iptel.org/mailman/listinfo/serusers

--
Jiri Kuthan            http://iptel.org/~jiri/