[Spce-user] R: mediator stopped

[Alessandro Romano]

Hi

I’ve the same problem do you have solved it ?

 

BR

 

Da: Spce-user <spce-user-bounces at lists.sipwise.com> Per conto di qabane me
Inviato: lunedì 11 marzo 2019 08.51
A: Andrew Pogrebennyk <apogrebennyk at sipwise.com>
Cc: Spce-user <spce-user at lists.sipwise.com>
Oggetto: Re: [Spce-user] mediator stopped

 

Hi Andrew

 

Could you give me a hint as to how to remove this from the redis DB. I have done a bit of searching on Redis but am pretty clueless on that and can't figure it out..

 

Cheers

 

On Wed, Mar 6, 2019 at 2:17 PM qabane me <qabaneitsolutions at gmail.com <mailto:qabaneitsolutions at gmail.com> > wrote:

Hi Andrew

 

Ah that explains why I don't see it in the DB. 

 

I am not sure what to look for in the kamailio.proxy log other than that caller id? there certainly isn't anything today and we have already purged older ones. 

 

I also have no idea where that long callid comes from and all seems to be running fine otherwise. But it would be swell of course if we can actually find what caused it.

In the meantime - aside from the cause, how do I remove it :-) ?

 

Regards

 

 

On Wed, Mar 6, 2019 at 1:42 PM Andrew Pogrebennyk <apogrebennyk at sipwise.com <mailto:apogrebennyk at sipwise.com> > wrote:

On 3/6/19 10:58 AM, qabane me wrote:
> Hi
> 
> Ok I have done a bit of digging as follows:
> 
>  select * from cdr WHERE call_id LIKE
> '%$116d5af1-b6d4-1237-df83-a6958035e26f%';
> 
> Under the assumption that I will find the offending call there and once
> I do, get rid of it. The result is empty though so clearly I am not
> looking in the right place?

Mediator works with acc records so you don't get a CDR if there is an
error fetching acc.
The acc records are stored in the redis DB with index 21 in most recent
versions. The mediator also tries to correlate the start/stop records
from redis with those from mysql, in case the user has switched the
engine or just updated the system. In that case, you might get some
errors in the cdr.log which aren't actually a problem, but in your case
my wild guess is that some system misconfiguration is causing such a
huge call-id suffix that it leads to inaccurate SQL query when trying to
fetch acc.

Can you check the kamailio-proxy.log and/or acc records and tell us how
in your setup, you obtained such a long call-id
116d5af1-b6d4-1237-df83-a6958035e26f_b2b-1_b2b-1_b2b-1_b2b-1....etc
because there is probably a loop going on somewhere in the system, e.g.
circular call forwarding involving several users. I am especially
interested in the kamailio-proxy.log because in most cases we do manage
to detect and break the loop in case of misconfiguration.

BR,
Andrew

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sipwise.com/pipermail/spce-user_lists.sipwise.com/attachments/20190311/542a073b/attachment-0001.html>