[Spce-user] Strange rate-o-mat behaviour v2.8

George Mason george.mason at xoomtalk.com
Mon Sep 1 11:23:10 EDT 2014


Hi,

Having some trouble understanding carrier rating and interested to
understand what's going on. I have created contracts for my SIP peerings
with costs for destinations. I have then assigned these to the peering
groups and made some test calls.

I can see that rate-o-mat is rating the calls and I can see in the cdr
table that the calls have a source_carrier_cost assigned after it has
run. However I cannot understand how it is calculating the costs, they
appear to have no relation to the costs in the billing fees.

An example as follows:

I set a contract with calls to 44 to have a 1 cent per second billing
fee. I then made a 20 secs call. The entry in the cdr table is as follows:

+----+-------------+--------------+------------------+---------------------+----------+-----------+---------------------+---------------------+---------------+
| id | source_user | source_cli   | destination_user |
update_time         | duration | call_type | source_carrier_cost |
rated_at            | rating_status |
+----+-------------+--------------+------------------+---------------------+----------+-----------+---------------------+---------------------+---------------+
|  8 | 225544      | +12127450200 | +441273900741    | 2014-09-01
11:02:01 |   26.170 | call      |                0.02 | 2014-09-01
11:02:01 | ok            |
+----+-------------+--------------+------------------+---------------------+----------+-----------+---------------------+---------------------+---------------+
1 row in set (0.00 sec)

I cannot see where the system is getting this cost from - it appears to
be using a cost of 5c per minute, which I don't have anywhere. I have
also tried to change the fee in the billing fee for the contract but
this had no effect.

Can I increase rate-o-mat's logging to see what it is doing? Or is there
some other diagnostic steps I can take?

Thanks in advance

George



More information about the Spce-user mailing list