[asterisk-bugs] [JIRA] (ASTERISK-24440) Call leak in Confbridge

Ben Klang (JIRA) noreply at issues.asterisk.org
Sat Oct 18 23:33:28 CDT 2014


     [ https://issues.asterisk.org/jira/browse/ASTERISK-24440?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ben Klang updated ASTERISK-24440:
---------------------------------

    Attachment: simple_call.xml

> Call leak in Confbridge
> -----------------------
>
>                 Key: ASTERISK-24440
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-24440
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Bridges/bridge_simple
>    Affects Versions: 11.13.0
>            Reporter: Ben Klang
>         Attachments: simple_call.xml
>
>
> I've been testing various Asterisk call behaviors with SIPp and with a no-media timeout set to 10 seconds. In every case, if I kill SIPp, Asterisk will eventually terminate the call for lack of media and eventually reach zero calls. However, if I have SIPp call into ConfBridge and then overload Asterisk, those calls are not cleaned up, even 30 minutes after SIPp has been killed.
> There is corresponding memory growth which makes me think that the channels aren't getting cleaned up like they should.  In "core show channels" output I see many thousands of calls, even though SIPp is limited to 1,000 concurrent calls.
> Some additional notes:
> * I do see messages indicating that calls should be getting disconnected, like this: [Oct 19 00:10:09] NOTICE[5874]: chan_sip.c:28942 check_rtp_timeout: Disconnecting call 'SIP/10.3.18.108-00000bda' for lack of RTP activity in 11 seconds
> * Many minutes after the test has ended and SIP traffic has stopped, the console is still flooded with messages like this: [Oct 19 00:12:26] WARNING[5874]: chan_sip.c:4258 __sip_autodestruct: Autodestruct on dialog '2472-31726 at 10.3.18.108' with owner SIP/10.3.18.108-000009a7 in place (Method: BYE). Rescheduling destruction for 10000 ms
> I've attached the SIPp XML that I'm using for this test.



--
This message was sent by Atlassian JIRA
(v6.2#6252)



More information about the asterisk-bugs mailing list