[logback-dev] [JIRA] Commented: (LBCLASSIC-57) SyslogAppender fails sporadically

[Anton Tagunov (JIRA)]

    [ http://jira.qos.ch/browse/LBCLASSIC-57?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=10620#action_10620 ] 

Anton Tagunov commented on LBCLASSIC-57:
----------------------------------------

We've traced this down to a very long log message "killing" our logging.
After we attempt to log a 1.85Mb message via SyslogAppender it stops working.

What is surprising is that even configuration reload which recreates DatagramSocket does not help.
It looks as if we messes up something in the data structures of libc on top of which jvm operates.

> SyslogAppender fails sporadically
> ---------------------------------
>
>                 Key: LBCLASSIC-57
>                 URL: http://jira.qos.ch/browse/LBCLASSIC-57
>             Project: logback-classic
>          Issue Type: Bug
>         Environment: $ uname -a
> Linux ... 2.6.9-67.ELsmp #1 SMP Wed Nov 7 13:56:44 EST 2007 x86_64 x86_64 x86_64 GNU/Linux
> $ java -version
> java version "1.6.0_03"
> Java(TM) SE Runtime Environment (build 1.6.0_03-b05)
> Java HotSpot(TM) Server VM (build 1.6.0_03-b05, mixed mode)
> multiprocessor PC box
>            Reporter: Anton Tagunov
>            Assignee: Logback dev list
>
> * several JVM-s per Linux box
> * 5 SyslogAppender-s logging to syslog-ng on 127.0.0.1
> Once in a while one of the SyslogAppender-s (the busiest one with longest messages) stops logging completely
> Other SyslogAppender-ss continue working fine.
> /usr/sbin/tcpdump has confirmed upd packets reach
> syslog-ng port via 'lo' network interface for remaining appends
> /usr/sbin/tcpdump has confirmed NO udp packets reach
> syslog-ng port via 'lo' network interface for the failing appender
> We have JMX enabled. JConsole does show StatusManager messages.
> There are no error messages there.
> Just the messages emitted during initial logback configuration
> from logback.xml on the classpath.
> We had invoked 'reload()' method via JMX.
> StatusMananger messages observed via JMX show that reload had happened.
> Full re-configuration had happened.
> All 5 SyslogAppenders had been apparently re-created.
> Yet NO udp messages had come from the SyslogAppender
> replacing those which had previously failed.
> Reloading did not help at all.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira