[OpenSIPS-Users] lookup "b" flag - one registration at a time

[Bogdan-Andrei Iancu]

Hi Jeff,

Got your email - I used the SB you had and from script check_address() - 
to be able to check with your IPs . It worked

Also I tested using some IPs from our network versus 
check_source_address() and again, It worked.

Please apply the attached patched (just debugs) and use debug=4 to 
capture the logs to see what is going wrong in your case.

Regards,
Bogdan


Bogdan-Andrei Iancu wrote:
> I do not think this is related to the registrar update, but rather to 
> the "permissions" fixes from today (see rev 6539).
>
> I did some tests and seams to work ok - if you want, you can privately 
> send me the dump of the address table and the IP which does not match, 
> just to check what is going on.
>
> Regards,
> Bogdan
>
> Jeff Pyle wrote:
>   
>> Bogdan,
>>
>> I didn't catch it before, but it appears the update seems to have broken the following:
>>
>> if (check_source_address("1","$avp(s:peer_uuid)")) { 
>>   .. do stuff  ..
>> }
>>
>> The check_source_address returns false.  As a temporary workaround I've added lines the following lines:
>>
>> if ($si == '2.3.4.5') {
>>   .. do stuff ..
>> } else if ($si == '2.3.4.6') {
>>   .. do the same stuff .. {
>> } else if ...
>>
>>
>> - Jeff
>>
>>
>> On Jan 25, 2010, at 11:49 PM, Jeff Pyle wrote:
>>
>>   
>>     
>>> Bogdan,
>>>
>>> Early results have this update working perfectly.
>>>
>>>
>>> - Jeff
>>>
>>>
>>> On Jan 25, 2010, at 10:29 AM, Jeff Pyle wrote:
>>>
>>>     
>>>       
>>>> Bogdan,
>>>>
>>>> Will do.  Thanks.
>>>>
>>>>
>>>>
>>>> - Jeff
>>>>
>>>>
>>>> On Jan 25, 2010, at 9:35 AM, Bogdan-Andrei Iancu wrote:
>>>>
>>>>       
>>>>         
>>>>> Hi Jeff,
>>>>>
>>>>> See revision #6527 on trunk - if you could run some more tests on it and 
>>>>> report if works ok, it will be great.
>>>>>
>>>>> Regards,
>>>>> Bogdan
>>>>>
>>>>> Jeff Pyle wrote:
>>>>>         
>>>>>           
>>>>>> The "f" flag sounds fantastic.  Thanks.
>>>>>>
>>>>>>
>>>>>> - Jeff
>>>>>>
>>>>>>
>>>>>> On Jan 18, 2010, at 9:24 AM, Bogdan-Andrei Iancu wrote:
>>>>>>
>>>>>>
>>>>>>           
>>>>>>             
>>>>>>> Hi Jeff,
>>>>>>>
>>>>>>> Jeff Pyle wrote:
>>>>>>>
>>>>>>>             
>>>>>>>               
>>>>>>>> Iñaki,
>>>>>>>>
>>>>>>>> On Jan 9, 2010, at 5:00 PM, Iñaki Baz Castillo wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>               
>>>>>>>>                 
>>>>>>>>> El Sábado, 9 de Enero de 2010, Jeff Pyle escribió:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>                 
>>>>>>>>>                   
>>>>>>>>>> Hello,
>>>>>>>>>>
>>>>>>>>>> The docs say that when using the "b" flag with lookup() when multiple
>>>>>>>>>> records are present, it will load only the one with the highest q.  What
>>>>>>>>>> if the q is the same for all?  How does it decide which to use?
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>                   
>>>>>>>>>>                     
>>>>>>>>> I've not tested it with multiple users sharing same "q". however it should 
>>>>>>>>> fetch all the users with highest "q", not just one of them.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>                 
>>>>>>>>>                   
>>>>>>>> Perhaps I'm asking the wrong question.  I'm looking to allow only one registration per user in the sense that if a second successful registration comes in it will replace tne existing one.  My approach so far is to use a max_contacts=2 and the lookup() function with the "b" flag to retrieve only one. 
>>>>>>>>
>>>>>>>>               
>>>>>>>>                 
>>>>>>> maybe without the "b" flag as the "b" flag will return you all the 
>>>>>>> registered contacts.
>>>>>>>
>>>>>>>             
>>>>>>>               
>>>>>>>> max_contacts=1 returns a 503 to the new "replacement" registration request, so that's out.
>>>>>>>>
>>>>>>>> Perhaps the hot ticket is to run an all-DB mode running a manual mysql query with avp_db_query after successful REGISTER authentication but before the save() so we can remove any existing registrations before the new one is saved.  Thoughts?
>>>>>>>>
>>>>>>>>
>>>>>>>>               
>>>>>>>>                 
>>>>>>> No way - the SIP contact matching is much to complicated to do it at DB 
>>>>>>> level.
>>>>>>>
>>>>>>>
>>>>>>> As I found that kind of behaviour was more and more asked by people, I 
>>>>>>> will add a new flag "f" to force at save() time the override of the 
>>>>>>> existing contacts if the max_contacts() was exceeded.
>>>>>>>
>>>>>>> Regards,
>>>>>>> Bogdan
>>>>>>>
>>>>>>>             
>>>>>>>               
>>>>>>>> - Jeff
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Users mailing list
>>>>>>>> Users at lists.opensips.org
>>>>>>>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>               
>>>>>>>>                 
>>>>>>> -- 
>>>>>>> Bogdan-Andrei Iancu
>>>>>>> www.voice-system.ro
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Users mailing list
>>>>>>> Users at lists.opensips.org
>>>>>>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>>>>>>>
>>>>>>>             
>>>>>>>               
>>>>>> Regards,
>>>>>> --------
>>>>>> Jeff Pyle
>>>>>> Director, Voice Engineering
>>>>>> Fidelity Voice & Data | 23250 Chagrin Blvd, Suite 250 | Beachwood, Ohio 44122
>>>>>> P: 216-245-4106
>>>>>> F: 216-595-0706
>>>>>> E: jpyle at fidelityvoice.com
>>>>>>
>>>>>> Visit us at http://www.fidelityvoice.com
>>>>>>
>>>>>> 2008 & 2009 Inductee to the prestigious Weatherhead 100
>>>>>>
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------
>>>>>>
>>>>>> ------------------------------------------------------------------------
>>>>>>
>>>>>>           
>>>>>>             
>
>
>   


-- 
Bogdan-Andrei Iancu
www.voice-system.ro

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: permission.diff
Url: http://lists.opensips.org/pipermail/users/attachments/20100127/47bf3b34/attachment.txt