You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@logging.apache.org by "Chris Martin (JIRA)" <ji...@apache.org> on 2018/10/17 06:38:00 UTC
[jira] [Commented] (LOG4J2-1479) Help Apache Spark project to
upgrade from Log4j 1.x to 2.x
[ https://issues.apache.org/jira/browse/LOG4J2-1479?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16653048#comment-16653048 ]
Chris Martin commented on LOG4J2-1479:
--------------------------------------
Hi- [~mikaelstaldal]
I've started work on the spark side to upgrade us to log4j2. Most of it is pretty straightforward with the exception of which [Logging.scala |https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/internal/Logging.scala]does a bit of messing about with the internals of log4j.
Fundamentally this class does the following:
* Determines if log4j is available
* Determines if log4j has been initialized yet
* Initializes log4j with a default config if one hasn't been specified
* Overwrites logger levels
* Resets the log4j system so it can be reinitialised form a new config at a later point.
I'd be interested in hearing your (or anyone else from the log4j2 project)'s opinion as to how this could be best achieved using log4j2. I'm fairly sure that the sort of thing this class does is not the approved mechanism for using log4j, however for the moment I'd like to preserve the functionality of the logging trait as much as possible.
thanks,
Chris
> Help Apache Spark project to upgrade from Log4j 1.x to 2.x
> ----------------------------------------------------------
>
> Key: LOG4J2-1479
> URL: https://issues.apache.org/jira/browse/LOG4J2-1479
> Project: Log4j 2
> Issue Type: Sub-task
> Reporter: Mikael Ståldal
> Priority: Major
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)