[
https://issues.apache.org/jira/browse/HADOOP-6884?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12899461#action_12899461
]
Luke Lu commented on HADOOP-6884:
---------------------------------
Did some digging on the aspectj wrapping approach. Looks like it's a no go
performance wise, as it cannot solve the arguments building cost issue without
fixing aspectj itself, as someone else tried to do exactly the same thing here:
http://dev.eclipse.org/mhonarc/lists/aspectj-users/msg09294.html
Looks the most reasonable current course of action is just committing the patch
as it looks correct, low risk and significantly lower gc stress (especially in
namenode code, according to Suresh in offline discussions.)
We should file a separate jira to explore the switching to slf4j api (still
using log4j as backend) approach. One open issue even with slf4j API is that it
doesn't solve the autoboxing cost issue for primitive types, which we use a lot
the logs:
http://qos.ch/pipermail/slf4j-dev/2005-August/000241.html (old but still
reflecting the current API design.)
> Add LOG.isDebugEnabled() guard for each LOG.debug("...")
> --------------------------------------------------------
>
> Key: HADOOP-6884
> URL: https://issues.apache.org/jira/browse/HADOOP-6884
> Project: Hadoop Common
> Issue Type: Improvement
> Affects Versions: 0.22.0
> Reporter: Erik Steffl
> Assignee: Erik Steffl
> Fix For: 0.22.0
>
> Attachments: HADOOP-6884-0.22-1.patch, HADOOP-6884-0.22.patch
>
>
> Each LOG.debug("...") should be executed only if LOG.isDebugEnabled() is
> true, in some cases it's expensive to construct the string that is being
> printed to log. It's much easier to always use LOG.isDebugEnabled() because
> it's easier to check (rather than in each case reason whether it's necessary
> or not).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.