[Spce-user] Multiple ATA registrations
skchopperguy at gmail.com
Mon Nov 19 01:07:48 EST 2012
Ah ok. So the device is following RFC 3261 then, sending proper callid
and incremental CSeq with the request. Thinking out loud, its as though the
command from kamailio to the DB is being duplicated for some reason but
only one AOR is being kept in cache.
I don't know what the next step to troubleshoot would be. Jon or Andreas
will be better suited to assist further.
On Sun, Nov 18, 2012 at 9:35 PM, Dave Massey <dave at optionsdsl.ca> wrote:
> The cseq is indeed incrementing each 60 seconds, but all the duplicate
> entries in the database do too.
> I should have sent an SQL following another 60 seconds, to demonstrate it
> but I didnt think of it :(
> On 2012-11-19, at 12:26 AM, Skyler <skchopperguy at gmail.com> wrote:
> Hi Dave,
> Looking at the sql you sent, I see the callid, last_modified and cseq are
> identical in all entries for ddufresne and test2. As you stated, the
> multiple entries in the DB are occurring whether the device reboots or it
> sits powered on (re-registering every 60 seconds) over time.
> Quoting Andreas from earlier in this thread:
> "If you register a contact for the first time, the server takes the
> call-id and cseq and stores the binding. If a UA wants to refresh this
> binding, it MUST send the same call-id, and it MUST increment the cseq.
> Failing to increment the cseq leads to an error message from the server,
> and failing to use the same call-id leads to an additional entry in the
> location table."
> So, the device is indeed not following RFC 3261 - 12.2.1 UAC Behavior;
> it is clearly ignoring the requirement where "the value of the local
> sequence number MUST be incremented by one, and this value MUST be placed
> into the CSeq header field."
> IMO, Kamailio is handling the CSeq error on the server-side by
> maintaining 1 entry in cache to avoid this UA's problem to avoid call
> failures, however, the location entry deletion is not occurring in the
> DB...(maybe on purpose?) so you/we can find problem UA's like this.
> On Sun, Nov 18, 2012 at 8:46 PM, Dave Massey <dave at optionsdsl.ca> wrote:
>> I was looking at RFC 3261 for call-id requiring to remain the same across
>> reboots, and I cant find anything stating it is :(
>> Im just looking for ammunition to give the developers.
>> Ive confirmed that it doesn't with this UA, but at this point, all they
>> are going to say, is "so?"
>> On 2012-11-15, at 9:28 AM, Jon Bonilla (Manwe) <jbonilla at sipwise.com>
>> > El Thu, 15 Nov 2012 09:21:40 -0500
>> > Dave Massey <dave at optionsdsl.ca> escribió:
>> >> This is probably exactly whats happening. :(
>> > If you can confirm this issue, probably you could ask the device
>> developers to
>> > update the firmware to fix it.
>> > _______________________________________________
>> > Spce-user mailing list
>> > Spce-user at lists.sipwise.com
>> > http://lists.sipwise.com/listinfo/spce-user
>> Spce-user mailing list
>> Spce-user at lists.sipwise.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Spce-user