[OpenSIPS-Users] NAT fixup question

Bogdan-Andrei Iancu bogdan at voice-system.ro
Tue Oct 20 18:15:16 CEST 2009


HI Jeff,

Just a wild guess - looking at the RURI and ROUTE hdrs, I would say your 
opensips is doing strict routing and not loose routing - this may happen 
when the IP in RURI is recognized as local SIP domain....

So, have you added "64.111.17.11" IP as alias in script or in domain 
table ? if so, please remove it!

Regards,
Bogdan



Jeff Kronlage wrote:
> Please also note this only happens on reinvites - the initial invite is
> fine.
>
> -----Original Message-----
> From: users-bounces at lists.opensips.org
> [mailto:users-bounces at lists.opensips.org] On Behalf Of Jeff Kronlage
> Sent: Monday, October 19, 2009 10:33 PM
> To: OpenSIPS users mailling list
> Subject: Re: [OpenSIPS-Users] NAT fixup question
>
> Bogdan,
>
> I wish I could post something useful.  I've been tinkering with this all
> evening, the catch is that one of our sip providers does things a tad
> unusual and I have a number of normalization procedures in place that
> make it hard to output something useful for this.
>
> I can provide this information -
>
> The inbound packet is:
> 22:19:32.479151 IP (tos 0x0, ttl 249, id 1369, offset 0, flags [none],
> proto UDP (17), length 539) 64.111.16.10.45463 > 64.YYY.XX.XX.5060: SIP,
> length: 511
>         ACK sip:719330XXXX at 64.111.17.11:5060 SIP/2.0
>         Via: SIP/2.0/UDP 192.168.1.105:5060;branch=z9hG4bK-c6f3cdea
>         From:
> <sip:719358ZZZZ at proxy.sip.data102.com>;tag=5d371768e81ef5fci0
>         To: <sip:719330XXXX at proxy.sip.data102.com>;tag=3442594C-21D0
>         Call-ID: 948E57D9-BC6611DE-AE38B8A5-3ADA34D9 at 64.111.17.2
>         CSeq: 101 ACK
>         Max-Forwards: 70
>         Route:
> <sip:64.YYY.XX.XX;lr=on;ftag=3442594C-21D0;did=03f.4363e5f6>
>         Contact: 719358ZZZZ <sip:719358ZZZZ at 192.168.1.105:5060>
>         User-Agent: Linksys/SPA2102-3.3.6
>         Content-Length: 0
>
> After I receive this packet and loose_route() is called, the RURI
> (specifically the value of $ru, as confirmed via xlog) is set to:
> sip:64.YYY.XX.XX;lr=on;ftag=3444FCF0-2256;did=10a.beb781a3
> (note this is identical to the "Route" field - not sure how I missed
> that prior to your mentioning it)
>
> A debug value of 4 produces:
> DBG:rr:after_loose: Topmost route URI:
> 'sip:64.YYY.XX.XX;lr=on;ftag=33F412AC-1CD1;did=a35.17b53e66' is me
> (not sure if that is of any use)
>
> I'm certain this is too vague to produce a solid answer, but any idea
> where I might look next?
>
> Thanks,
>
> Jeff
>
> -----Original Message-----
> From: users-bounces at lists.opensips.org
> [mailto:users-bounces at lists.opensips.org] On Behalf Of Bogdan-Andrei
> Iancu
> Sent: Monday, October 19, 2009 9:21 PM
> To: OpenSIPS users mailling list
> Subject: Re: [OpenSIPS-Users] NAT fixup question
>
> RURI has nothing to do with the VIA part..
>
> Also RURI is not to be changed during loose_route(), only if you have a 
> strict router proxy in front of you....maybe you can post the inbound 
> and outbound request (to see how the loose_route() is done)
>
> Regards,
> Bogdan
>
> Jeff Kronlage wrote:
>   
>> The RURI.
>>
>> Thanks,
>>
>> Jeff
>>
>> -----Original Message-----
>> From: users-bounces at lists.opensips.org
>> [mailto:users-bounces at lists.opensips.org] On Behalf Of Bogdan-Andrei
>> Iancu
>> Sent: Monday, October 19, 2009 9:13 PM
>> To: OpenSIPS users mailling list
>> Subject: Re: [OpenSIPS-Users] NAT fixup question
>>
>> Jeff,
>>
>> the VIA hdr does not require mangling - the addition of the "received"
>>     
>
>   
>> param is enough to handle nat issues. So the VIA you posted is correct
>>     
>
>   
>> form NAT traversal point of view.
>>
>> Regarding the user part of the URI - what URI you are talking about? 
>> RURI ? TO / FROM uri? Contact URI ?
>>
>> Regards,
>> Bogdan
>>
>> Jeff Kronlage wrote:
>>   
>>     
>>> Thanks Bogdan,
>>>
>>> An unrelated question:
>>>
>>> Does anything special need to be done with "via" statements when
>>> implementing NAT transversal?
>>>
>>> Fix_nated_contact() takes care of the contact field for me, but I
>>>     
>>>       
>> still
>>   
>>     
>>> end up with:
>>> Via: SIP/2.0/UDP
>>>
>>>     
>>>       
> 192.168.1.105:5060;rport=42080;received=64.YYY.XX.XX;branch=z9hG4bK-e4e5
>   
>>   
>>     
>>> cd84
>>>
>>> I'm having some random problems with the user part of the URI
>>>       
> randomly
>   
>>> vanishing after I call loose_route() when NAT is involved, and I'm
>>> thinking these are related.
>>>
>>> Thanks,
>>>
>>> --
>>> Jeff Kronlage
>>> Senior IT Engineer, Data102
>>> 102 South Tejon, Suite #1250
>>> Colorado Springs, CO 80903
>>> (719) 387-0000 x 1335 direct
>>> (719) 578-8844 fax
>>> jeff at data102.com / http://www.data102.com
>>>
>>>
>>> -----Original Message-----
>>> From: users-bounces at lists.opensips.org
>>> [mailto:users-bounces at lists.opensips.org] On Behalf Of Bogdan-Andrei
>>> Iancu
>>> Sent: Monday, October 19, 2009 4:07 PM
>>> To: OpenSIPS users mailling list
>>> Subject: Re: [OpenSIPS-Users] Additional info on potential
>>>     
>>>       
>> registration
>>   
>>     
>>> issue
>>>
>>> so DB ONLY mode......simply ignore the warning (see its meaning in my
>>>       
>
>   
>>> previous post) . The contacts will still be shared, but the socket 
>>> information discarded.
>>>
>>> Regards,
>>> Bogdan
>>>
>>> Jeff Kronlage wrote:
>>>   
>>>     
>>>       
>>>> Usrloc mode is 3.  
>>>>
>>>> --
>>>> Jeff Kronlage
>>>> Senior IT Engineer, Data102
>>>> 102 South Tejon, Suite #1250
>>>> Colorado Springs, CO 80903
>>>> (719) 387-0000 x 1335 direct
>>>> (719) 578-8844 fax
>>>> jeff at data102.com / http://www.data102.com
>>>>
>>>> -----Original Message-----
>>>> From: users-bounces at lists.opensips.org
>>>> [mailto:users-bounces at lists.opensips.org] On Behalf Of Bogdan-Andrei
>>>> Iancu
>>>> Sent: Monday, October 19, 2009 11:32 AM
>>>> To: OpenSIPS users mailling list
>>>> Subject: Re: [OpenSIPS-Users] Additional info on potential
>>>>     
>>>>       
>>>>         
>>> registration
>>>   
>>>     
>>>       
>>>> issue
>>>>
>>>> So, this is the problem - each opensips instance loads only the
>>>>       
>>>>         
>> usrloc
>>   
>>     
>>>>     
>>>>       
>>>>         
>>>   
>>>     
>>>       
>>>> records that have the a local socket corresponding to that instance.
>>>>     
>>>>       
>>>>         
>>> In 
>>>   
>>>     
>>>       
>>>> other words, if the record was saved by the other instance, opensips
>>>>         
>
>   
>>>> will not load it.
>>>>
>>>> what db_mode do you use for usrloc?
>>>>
>>>> Regards,
>>>> Bogdan
>>>>
>>>> Jeff Kronlage wrote:
>>>>   
>>>>     
>>>>       
>>>>         
>>>>> Yes, shared location table over multiple servers.
>>>>>
>>>>> -----Original Message-----
>>>>> From: users-bounces at lists.opensips.org
>>>>> [mailto:users-bounces at lists.opensips.org] On Behalf Of
>>>>>           
> Bogdan-Andrei
>   
>>>>> Iancu
>>>>> Sent: Sunday, October 18, 2009 10:33 PM
>>>>> To: OpenSIPS users mailling list
>>>>> Subject: Re: [OpenSIPS-Users] Additional info on potential
>>>>>     
>>>>>       
>>>>>         
>>>>>           
>>>> registration
>>>>   
>>>>     
>>>>       
>>>>         
>>>>> issue
>>>>>
>>>>> Hi Jeff,
>>>>>
>>>>> Do you use a shared location table (via multiple registrar servers)
>>>>>         
>>>>>           
>> ?
>>   
>>     
>>>>> Regards,
>>>>> Bogdan
>>>>>
>>>>> Jeff Kronlage wrote:
>>>>>   
>>>>>     
>>>>>       
>>>>>         
>>>>>           
>>>>>> I'm getting this over and over in my syslog:
>>>>>>
>>>>>> WARNING:usrloc:get_all_db_ucontacts: non-local socket
>>>>>> <udp:HI.DDE.N.12:5060>...ignoring
>>>>>>
>>>>>> Thanks,
>>>>>>
>>>>>> Jeff
>>>>>>
>>>>>>
>>>>>>       
>>>>>>         
>>>>>>           
>>>>>>             




More information about the Users mailing list