logging-log4j-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sebastian Oerding (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (LOG4J2-290) System.out mixed up with other input streams
Date Wed, 26 Jun 2013 09:31:25 GMT

    [ https://issues.apache.org/jira/browse/LOG4J2-290?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13693865#comment-13693865

Sebastian Oerding commented on LOG4J2-290:


I tried to create a test for a few hours. However I struggled to create an own appender and
my boss won't be very happy if I use a whole day to write a Junit tests ... .

However I struggled with writing the tests as the generation of the Log4j2Plugins.dat seems
to be broken a bit and furthermore there are new problems comparing beta7 with beta6.

1. I invoked the PluginManager.main(new String[] { ".\\src/config", "log4j2.socket" }); in
a class that defines a Logger. Then the generated Log4j2Plugins.dat only contains all the
elements of log4j2 but not my own implementation of an appender. When I remove the Logger
definition from the test class such that no Log4j2 is invoked before the test executes the
generated Log4j2Plugins.dat contains only my appender as expected.

2. I named my appender MyAppender (the class name and the defined name attribute in @Plugin(name
= "MyAppender", ...). However in the Log4j2Plugins.dat the name is completely changed to lowercase.
However using that name in different spellings ("MyAppender", "myappender" and "MyAppender")
resulted in an "Unable to locate Appender APPENDER_NAME for logger".

3. I finally managed to get it working by using the Log4j2Plugins.dat created with log4j2
beta6, renaming the MyAppender class and putting it into the directory accordingly to the
copied Log4j2Plugins.dat

I will try to add the source code. Is there any possibility adding source without putting
it into a comment? otherwise see the next comment for the code.
> System.out mixed up with other input streams
> --------------------------------------------
>                 Key: LOG4J2-290
>                 URL: https://issues.apache.org/jira/browse/LOG4J2-290
>             Project: Log4j 2
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 2.0-beta7
>         Environment: Windows 7 64 bit, Eclipse 4.2.1, java JDK
>            Reporter: Sebastian Oerding
> Hi,
> I've noticed a strange behaviour which seems to be a bug in log4j 2. Unfortunately I'm
not totally sure if the problem was caused by my code, log4j or the JDK.
> I've written my own appender which works as expected. I've have implemented some stuff
to send log events over the network (the socket appender does not fit my needs).
> I use log4j 2 to get the log events in my own appender. There I filter the events and
send them over the network using java.net.sockets.
> For unit tests I simulate a client and server in the same JVM using sockets.
> Furthermore the console appender is activated in the log4j2.xml. As the logging is done
asynchronously the client are threads awaiting log events endlessly until the whole client
is stopped.
> 1. Now, when I output something with System.out.println sometimes it seems to be forwarded
to other appenders. Soory but at the moment I lack a better description.
> 2. I had the case that something was logged on the server using LOG.debug(...) (where
LOG was my Logger instance). However this log seem to never reach my appender but it appears
at the client's socket input stream.
> Hence either the input streams got somehow mixed up or due to a mysterious reason data
written to system.out is also written to the server's socket output stream (and received on
the client) or directly written to the client's input stream.
> I would not claim that my code has no bugs (actually I found zombie threads due to this)
but I can not imagine how I should have caused such problems as I am working nowhere with
System.in / System.out besides invoking System.out.println. This problems occurred only rarely
but was reproducible.
> Unfortunately I can not provide the complete code to reproduce this problem.
> I also feel unfortunate for the vague description.
> However if you have concrete questions, I may be able to be more precise or to provide
code snippets.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

To unsubscribe, e-mail: log4j-dev-unsubscribe@logging.apache.org
For additional commands, e-mail: log4j-dev-help@logging.apache.org

View raw message