[asterisk-bugs] [JIRA] (ASTERISK-28543) When Asterisk cannot connect to SIP socket it starts to flood with "Bad descriptor" errors and hangs

[John Smith (JIRA)]

     [ https://issues.asterisk.org/jira/browse/ASTERISK-28543?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

John Smith updated ASTERISK-28543:
----------------------------------

    Description: 
    When some endpoint goes offline the following messages appear in the console:

 ERROR[2889]: tcptls.c:553 ast_tcptls_client_start: Unable to connect SIP socket to IP:5060: Connection refused 

followed by 

ERROR[2751]: iostream.c:569 ast_iostream_close: close() failed: Bad file descriptor

Error occurs only for endpoints with protocol=tcp in sip.conf (and probably tls) after some time calls started to drop with "Bad file descriptor" errors and finally after some time Asterisk becomes unresponsible: all endpoints lose their registrations, no calls can be made/received. 
Service restart does not help. 
Server restart can help for a short time. 
Workaround is to remove offline endpoint from sip.conf

sip.conf example:

[phone1] 
type=friend 
host=10.10.8.8 
port=5060 
dtmfmode=rfc2833
context=office
directmedia=no
qualify=yes 
transport=tcp 
allow=!all,alaw,ulaw,g722 


  was:
    When some endpoint goes offline the following messages appear in the console: ERROR[2889]: tcptls.c:553 ast_tcptls_client_start: Unable to connect SIP socket to IP:5060: Connection refused followed by ERROR[2751]: iostream.c:569 ast_iostream_close: close() failed: Bad file descriptor Error occurs only for endpoints with protocol=tcp in sip.conf (and probably tls) after some time calls started to drop with "Bad file descriptor" errors and finally after some time Asterisk becomes unresponsible: all endpoints lose their registrations, no calls can be made/received. Service restart does not help. Server restart can help on a short time. Workaround is to remove offline endpoint from sip.conf sip.conf example: [sfb] type=friend host=10.30.8.90 port=5060 dtmfmode=rfc2833 context=office directmedia=no qualify=yes transport=tcp allow=!all,alaw,ulaw,g722 



>     When Asterisk cannot connect to SIP socket it starts to flood with "Bad descriptor" errors and hangs
> --------------------------------------------------------------------------------------------------------
>
>                 Key: ASTERISK-28543
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-28543
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: General
>    Affects Versions: 16.3.0, 16.5.1
>         Environment:     CentOS Linux release 7.6.1810 (Core) x64
>            Reporter: John Smith
>         Attachments: 2019-09-18_10-4303.png, MobaXterm_Personal_11.1_2019-09-16_21-2648.png
>
>
>     When some endpoint goes offline the following messages appear in the console:
>  ERROR[2889]: tcptls.c:553 ast_tcptls_client_start: Unable to connect SIP socket to IP:5060: Connection refused 
> followed by 
> ERROR[2751]: iostream.c:569 ast_iostream_close: close() failed: Bad file descriptor
> Error occurs only for endpoints with protocol=tcp in sip.conf (and probably tls) after some time calls started to drop with "Bad file descriptor" errors and finally after some time Asterisk becomes unresponsible: all endpoints lose their registrations, no calls can be made/received. 
> Service restart does not help. 
> Server restart can help for a short time. 
> Workaround is to remove offline endpoint from sip.conf
> sip.conf example:
> [phone1] 
> type=friend 
> host=10.10.8.8 
> port=5060 
> dtmfmode=rfc2833
> context=office
> directmedia=no
> qualify=yes 
> transport=tcp 
> allow=!all,alaw,ulaw,g722 



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