[asterisk-bugs] [JIRA] (ASTERISK-25302) Error:Oh dear... we couldn't allocate a port for RTP instance

[Vladimir Yordanov (JIRA)]

    [ https://issues.asterisk.org/jira/browse/ASTERISK-25302?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=227170#comment-227170 ] 

Vladimir Yordanov commented on ASTERISK-25302:
----------------------------------------------

Thank you for your reply. We have started gathering debug logs and will post them shortly.
About the reproduction scenario:
- We try to perform 40000 calls for 3 hours. We are using 150 lines. Calls are equaly distributed.
After some period of active usage and making calls our PBX stops sending ACK after our provider has sent 183 ringing and 200 OK with SDP.
This causes provider SBC to resend the message again and again until the SBC gives up and the call fails.
After that moment:
- Asterisk stops releasing ports and occupies more and more until all (10000-20000) become occupied
- our schedulers stop receiving hang up signal
- lines remain blocked
- if you make a test call the call is done but the messages are not played

We have posted in the forums but there are no replies or similar problems that can help us.




> Error:Oh dear... we couldn't allocate a port for RTP instance 
> --------------------------------------------------------------
>
>                 Key: ASTERISK-25302
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-25302
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>    Affects Versions: 13.4.0
>         Environment: CentOS 7, 150 lines
>            Reporter: Vladimir Yordanov
>
> After a period of constant work (calls) this problem appears:
> "Oh dear... we couldn't allocate a port for RTP instance ..."
> and Asterisk  stops processing calls.
> The whole port range 10000-20000 appears blocked.
> Only restart of Asterisk frees the ports.
> Any ideas how this coud be resolved?



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