[asterisk-bugs] [JIRA] (ASTERISK-20037) func_odbc timeout quite long when connection drops during query - possibly need to expose query timeout attribute to make it configurable

Martin Tomec (JIRA) noreply at issues.asterisk.org
Wed May 13 06:25:33 CDT 2015


    [ https://issues.asterisk.org/jira/browse/ASTERISK-20037?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=226197#comment-226197 ] 

Martin Tomec commented on ASTERISK-20037:
-----------------------------------------

I am not sure if this is related, but I submited patch to ASTERISK-22708. If you set negative_connection_cache, asterisk should not connect to dead DB for specified amount of time.

> func_odbc timeout quite long when connection drops during query - possibly need to expose query timeout attribute to make it configurable
> -----------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: ASTERISK-20037
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-20037
>             Project: Asterisk
>          Issue Type: Improvement
>      Security Level: None
>          Components: Functions/func_odbc, Resources/res_odbc
>    Affects Versions: 1.8.9.3
>         Environment: Asterisk 1.8.9.2
> CentOS 5.5 Final
> MyODBC 5.1.9
>            Reporter: Steven T. Wheeler
>         Attachments: buglog.log
>
>
> When using UnixODBC to connect Asterisk to a MySQL database server we are encountering issues when the database goes down.  We have designed the system with a redundant database server, but Asterisk locks up when the primary database goes into a hung state where it neither accepts or rejects the connections.  We simulate this condition using this iptables rule on the Asterisk server: -A OUTPUT -d x.x.x.x -j DROP.  We have setup connection timeouts which work well if the connections are not established when the primary database server goes down, but it doesn't look like Asterisk implements any sort of query timeout.
> Am I missing something or is this a known issue with Asterisk?



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



More information about the asterisk-bugs mailing list