[SR-Users] no event received for call disconnection

Daniel-Constantin Mierla miconda at gmail.com
Fri Nov 3 09:26:54 CET 2017


Hello,

can you try with latest master? I just pushed a patch that should handle
this case. Let me know if works or not.

Cheers,
Daniel


On 01.11.17 11:57, Aleksandar Sosic wrote:
> Thank you Daniel!
> --
> Aleksandar Sosic
> mail: alex.sosic at timenet.it
> skype: alex.sosic
> cell: +385 91 2505 146
>
>
> On Wed, Nov 1, 2017 at 10:39 AM, Daniel-Constantin Mierla
> <miconda at gmail.com> wrote:
>> Hello,
>>
>> ok, now the cfg execution flow is easier to understand. The evapi
>> receiver was not designed to write back directly, I will try to find a
>> solution for this case.
>>
>> Cheers,
>> Daniel
>>
>> On 31.10.17 15:03, Aleksandar Sosic wrote:
>>> Hi Daniel,
>>>
>>> so the kamailio.conf sections that matters are here:
>>> https://gist.github.com/alexsosic/5a33d314d2db1832fd04166fe6615227
>>>
>>> Thanks for your time!
>>> --
>>> Aleksandar Sosic
>>> mail: alex.sosic at timenet.it
>>> skype: alex.sosic
>>> cell: +385 91 2505 146
>>>
>>>
>>> On Tue, Oct 31, 2017 at 12:51 PM, Daniel-Constantin Mierla
>>> <miconda at gmail.com> wrote:
>>>> It is a bit hard to guess the logic you have there, but seems like
>>>> cgrates sends a disconnect request via event api, which via some actions
>>>> triggers a dialog end, which then wants to send back an event to cgrates
>>>> that call is ending? Like a loop back communication with evapi receiver ...
>>>>
>>>> Maybe you can share the config file (or the relevant parts).
>>>>
>>>> Cheers,
>>>> Daniel
>>>>
>>>> On 31.10.17 12:18, Aleksandar Sosic wrote:
>>>>> 17(114) NOTICE: evapi [evapi_dispatch.c:339]: evapi_recv_client(): {0}
>>>>> [127.0.0.1:41154] - received
>>>>> [99:{"Event":"CGR_SESSION_DISCONNECT","HashEntry":"715","HashId":"1885","Reason":"-INSUFFICIENT_FUNDS"},]
>>>>> (103) (0)
>>>>>
>>>>> It seems like an evapi recieved command.
>>>>> That triggers that route.
>>>>>
>>>>> After that I do have this errors in kamailio logs:
>>>>> 17(114) ERROR: <core> [core/parser/parse_from.c:107]:
>>>>> parse_from_uri(): failed to parse From uri
>>>>> 17(114) ERROR: pv [pv_core.c:414]: pv_get_xto_attr(): cannot parse From URI
>>>>> 17(114) ERROR: <core> [core/parser/parse_to.c:81]: parse_to_uri():
>>>>> failed to parse To uri
>>>>> 17(114) ERROR: pv [pv_core.c:408]: pv_get_xto_attr(): cannot parse To URI
>>>>> --
>>>>> Aleksandar Sosic
>>>>> mail: alex.sosic at timenet.it
>>>>> skype: alex.sosic
>>>>> cell: +385 91 2505 146
>>>>>
>>>>>
>>>>> On Tue, Oct 31, 2017 at 12:04 PM, Daniel-Constantin Mierla
>>>>> <miconda at gmail.com> wrote:
>>>>>> Hello,
>>>>>>
>>>>>> when is the route[CGR_SESSION_DISCONNECT] executed? In other words,
>>>>>> where do you have route(CGR_SESSION_DISCONNECT) in your config or is
>>>>>> there any modparam with CGR_SESSION_DISCONNECT in it?
>>>>>>
>>>>>> Cheers,
>>>>>> Daniel
>>>>>>
>>>>>>
>>>>>> On 31.10.17 11:45, Aleksandar Sosic wrote:
>>>>>>> Hi Daniel,
>>>>>>>
>>>>>>> It's done by this cgrates route:
>>>>>>> ```
>>>>>>> # CGRateS request for session disconnect
>>>>>>> route[CGR_SESSION_DISCONNECT] {
>>>>>>>   json_get_field("$evapi(msg)", "HashEntry", "$var(HashEntry)");
>>>>>>>   json_get_field("$evapi(msg)", "HashId", "$var(HashId)");
>>>>>>>   json_get_field("$evapi(msg)", "Reason", "$var(Reason)");
>>>>>>>   jsonrpc_exec('{"jsonrpc":"2.0","id":1,
>>>>>>> "method":"dlg.end_dlg","params":[$(var(HashEntry){s.rm,"}),$(var(HashId){s.rm,"})]}');
>>>>>>> }
>>>>>>> ```
>>>>>>>
>>>>>>> It executes a dlg.end_dlg via jsonrpc.
>>>>>>> --
>>>>>>> Aleksandar Sosic
>>>>>>> mail: alex.sosic at timenet.it
>>>>>>> skype: alex.sosic
>>>>>>> cell: +385 91 2505 146
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Oct 30, 2017 at 5:57 PM, Daniel-Constantin Mierla
>>>>>>> <miconda at gmail.com> wrote:
>>>>>>>> Hello,
>>>>>>>>
>>>>>>>>
>>>>>>>> On 30.10.17 17:29, Aleksandar Sosic wrote:
>>>>>>>>> Hi Everyone,
>>>>>>>>>
>>>>>>>>> we're trying to implement cgrates in our kamailio nodes.
>>>>>>>>> Everything is working fine except for the fact that kamailio is not
>>>>>>>>> signaling the call end to cgrates for prepaid users... The versions,
>>>>>>>>> logs and confs are:
>>>>>>>>>
>>>>>>>>> ```
>>>>>>>>> # kamailio -v
>>>>>>>>> version: kamailio 5.0.3 (x86_64/linux)
>>>>>>>>> flags: STATS: Off, USE_TCP, USE_TLS, USE_SCTP, TLS_HOOKS,
>>>>>>>>> DISABLE_NAGLE, USE_MCAST, DNS_IP_HACK, SHM_MEM, SHM_MMAP, PKG_MALLOC,
>>>>>>>>> Q_MALLOC, F_MALLOC, TLSF_MALLOC, DBG_SR_MEMORY, USE_FUTEX,
>>>>>>>>> FAST_LOCK-ADAPTIVE_WAIT, USE_DNS_CACHE, USE_DNS_FAILOVER, USE_NAPTR,
>>>>>>>>> USE_DST_BLACKLIST, HAVE_RESOLV_RES
>>>>>>>>> ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16,
>>>>>>>>> MAX_URI_SIZE 1024, BUF_SIZE 65535, DEFAULT PKG_SIZE 8MB
>>>>>>>>> poll method support: poll, epoll_lt, epoll_et, sigio_rt, select.
>>>>>>>>> id: unknown
>>>>>>>>> compiled with gcc 5.3.1
>>>>>>>>> ```
>>>>>>>>>
>>>>>>>>> ```
>>>>>>>>> # cgr-engine --version
>>>>>>>>> CGRateS 0.9.1~rc8 git+15a0793 (2017-10-23T12:15:40+02:00)
>>>>>>>>> ```
>>>>>>>>>
>>>>>>>>> The kamailio conf regarding our problem is:
>>>>>>>>> ```
>>>>>>>>> # Inform CGRateS about CALL_END (stop debit loops, perform accounting
>>>>>>>>> if desired in this way)
>>>>>>>>> route[CGR_CALL_END] {
>>>>>>>>>   if $sht(cgrconn=>cgr) == $null {
>>>>>>>>>     xlog("Charging controller unreachable");
>>>>>>>>>     exit;
>>>>>>>>>   }
>>>>>>>>>   $var(callDur) = $TS - $dlg(start_ts);
>>>>>>>>>   evapi_relay("{\"event\":\"CGR_CALL_END\",
>>>>>>>>>     \"callid\":\"$dlg(callid)\",
>>>>>>>>>     \"from_tag\":\"$dlg(from_tag)\",
>>>>>>>>>     \"cgr_reqtype\":\"$dlg_var(cgrReqType)\",
>>>>>>>>>     \"cgr_tenant\":\"$dlg_var(cgrTenant)\",
>>>>>>>>>     \"cgr_account\":\"$dlg_var(cgrAccount)\",
>>>>>>>>>     \"cgr_destination\":\"$dlg_var(cgrDestination)\",
>>>>>>>>>     \"cgr_answertime\":\"$dlg(start_ts)\",
>>>>>>>>>     \"cgr_duration\":\"$var(callDur)\",
>>>>>>>>>     \"cgr_supplier\":\"$dlg_var(cgrSupplier)\",
>>>>>>>>>     \"cgr_disconnectcause\":\"$T_reply_code\"}");
>>>>>>>>> }
>>>>>>>>> ```
>>>>>>>>>
>>>>>>>>> The kamailio logs:
>>>>>>>>>
>>>>>>>>> ```
>>>>>>>>> 17(114) ERROR: evapi [evapi_dispatch.c:707]: _evapi_relay(): failed to
>>>>>>>>> pass the pointer to evapi dispatcher
>>>>>>>>> 17(114) ERROR: evapi [evapi_mod.c:261]: w_evapi_relay(): failed to
>>>>>>>>> relay event: {"event":"CGR_CALL_END",
>>>>>>>>> [...]
>>>>>>>> does this happen on BYE or on dialog timeout?
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> Daniel
>>>>>>>>
>>>>>>>> --
>>>>>>>> Daniel-Constantin Mierla
>>>>>>>> www.twitter.com/miconda -- www.linkedin.com/in/miconda
>>>>>>>> Kamailio Advanced Training, Nov 13-15, 2017, in Berlin - www.asipto.com
>>>>>>>> Kamailio World Conference - www.kamailioworld.com
>>>>>>>>
>>>>>> --
>>>>>> Daniel-Constantin Mierla
>>>>>> www.twitter.com/miconda -- www.linkedin.com/in/miconda
>>>>>> Kamailio Advanced Training, Nov 13-15, 2017, in Berlin - www.asipto.com
>>>>>> Kamailio World Conference - www.kamailioworld.com
>>>>>>
>>>> --
>>>> Daniel-Constantin Mierla
>>>> www.twitter.com/miconda -- www.linkedin.com/in/miconda
>>>> Kamailio Advanced Training, Nov 13-15, 2017, in Berlin - www.asipto.com
>>>> Kamailio World Conference - www.kamailioworld.com
>>>>
>> --
>> Daniel-Constantin Mierla
>> www.twitter.com/miconda -- www.linkedin.com/in/miconda
>> Kamailio Advanced Training, Nov 13-15, 2017, in Berlin - www.asipto.com
>> Kamailio World Conference - www.kamailioworld.com
>>

-- 
Daniel-Constantin Mierla
www.twitter.com/miconda -- www.linkedin.com/in/miconda
Kamailio Advanced Training, Nov 13-15, 2017, in Berlin - www.asipto.com
Kamailio World Conference - www.kamailioworld.com




More information about the sr-users mailing list