[Spce-user] Problems with peergroup server change and cdr display

Vladimir Grujic hyperbabac at gmail.com
Mon Jun 18 12:34:37 EDT 2012


It's latest and greatest 2.5.

:)

Vladimir


On Mon, Jun 18, 2012 at 6:32 PM, Andrew Pogrebennyk <
apogrebennyk at sipwise.com> wrote:

> ** We've fixed a related bug back in April (fax vs. reminder
> identification), have you updated your system recently?
>
> --
> Sent from my Android phone with K-9 Mail. Please excuse my brevity.
>
>
> Vladimir Grujic <hyperbabac at gmail.com> wrote:
>>
>> I've found what was causing cdr to stop working. It's the mediator
>> process failing to recognize call_type field.(Error inserting cdrs : Data
>> truncated for column 'call_type' at row (sometimes 1 or 2 or 6))
>>  for calls originating from the asterisk server (prepended with fax-
>> prefix for the moment). The mediator process dies instantly upon receiving
>> such a record in kamailio table. I need to originate some calls from
>> asterisk and i have used the fax server configuration logic for that. I see
>> i will need to make a proper route for my calls from asterisk.
>>
>> As for the nat detection i will make a trace calls and see what's going
>> on.
>>
>>
>> Vladimir
>>
>>
>>
>> On Mon, Jun 18, 2012 at 5:14 PM, Andreas Granig <agranig at sipwise.com>wrote:
>>
>>> Hi,
>>>
>>> Please ALWAYS reply to the list as well for others to see it (unless
>>> it's log messages we're requesting for in-depth trouble shooting).
>>>
>>> On 06/17/2012 09:53 PM, Vladimir Grujic wrote:
>>> > now even more strange thing is happening. I have a call with some
>>> > call_id in the cdr table but there is no entries for that call_id in
>>> the
>>> > log file for that date.
>>>
>>> Then rsyslog is not running on your machine, or there is some issue with
>>> your storage. Also make sure to use "zgrep" instead of "grep" when
>>> searching in /var/log/ngcp/old/, because the log files there are gzipped.
>>>
>>> > Will do some more tests (maybe i have accidentally change something in
>>> > proxy.conf for kamailio, because a had to add some app to asterisk and
>>> > route some numbers to it)
>>> > Another thing i have noticed is bad nat detection. I have various cpe
>>> > devices at customer premises (adsl, cdma evdo, wimax, wifi) and in
>>> > majority of cases i see that subscriber use of  nat is no on his status
>>> > page and i am certain that his sip is behind the cpe. The another
>>> > problem are some cpe devices with sip clients on it . I have a peer
>>> > defined and forced use of proxy for that peer. The sip client is at
>>> > public ip and forced to use proxy, nevertheless the rtp stream is
>>> trying
>>> > to connect from peer to cpe directly. I will elaborate all that quite
>>> > extensively on another mail which i will send in front of the company i
>>> > work for in the next couple of days (some form o enterprise business
>>> > inquiry).
>>>
>>> NAT detection is done based on the comparison of the source ip the
>>> request is coming from with the IP addresses in various header fields
>>> (e.g. Via and Contact). Do a tcpdump or ngrep to see the content of the
>>> SIP messages.
>>>
>>> Andreas
>>>
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sipwise.com/pipermail/spce-user_lists.sipwise.com/attachments/20120618/14db73ba/attachment-0001.html>


More information about the Spce-user mailing list