Home » Java » Enterprise Java » SiftingAppender: logging different threads to different log files

About Tomasz Nurkiewicz

Java EE developer, Scala enthusiast. Enjoying data analysis and visualization. Strongly believes in the power of testing and automation.

SiftingAppender: logging different threads to different log files

One novel feature of Logback is SiftingAppender (JavaDoc). In short it’s a proxy appender that creates one child appender per each unique value of a given runtime property. Typically this property is taken from MDC. Here is an example based on the official documentation linked above:
 
 
 
 
 
 
 

<?xml version="1.0" encoding="UTF-8"?>
<configuration>
 
    <appender name="SIFT" class="ch.qos.logback.classic.sift.SiftingAppender">
        <discriminator>
            <key>userid</key>
            <defaultValue>unknown</defaultValue>
        </discriminator>
        <sift>
            <appender name="FILE-${userid}" class="ch.qos.logback.core.FileAppender">
                <file>user-${userid}.log</file>
                <layout class="ch.qos.logback.classic.PatternLayout">
                    <pattern>%d{HH:mm:ss:SSS} | %-5level | %thread | %logger{20} | %msg%n%rEx</pattern>
                </layout>
            </appender>
        </sift>
    </appender>
 
    <root level="ALL">
        <appender-ref ref="SIFT" />
    </root>
</configuration>

Notice that the <file> property is parameterized with ${userid} property. Where does this property come from? It has to be placed in MDC. For example in a web application using Spring Security I tend to use a servlet filter with a help of SecurityContextHolder:

import javax.servlet._
import org.slf4j.MDC
import org.springframework.security.core.context.SecurityContextHolder
import org.springframework.security.core.userdetails.UserDetails
 
class UserIdFilter extends Filter
{
    def init(filterConfig: FilterConfig) {}
 
    def doFilter(request: ServletRequest, response: ServletResponse, chain: FilterChain) {
        val userid = Option(
            SecurityContextHolder.getContext.getAuthentication
        ).collect{case u: UserDetails => u.getUsername}
 
        MDC.put("userid", userid.orNull)
        try {
            chain.doFilter(request, response)
        } finally {
            MDC.remove("userid")
        }
 
    }
 
    def destroy() {}
}

Just make sure this filter is applied after Spring Security filter. But that’s not the point. The presence of ${userid} placeholder in the file name causes sifting appender to create one child appender for each different value of this property (thus: different user names). Running your web application with this configuration will quickly create several log files like user-alice.log, user-bob.log and user-unknown.log in case of MDC property not set. Another use case is using thread name rather than MDC property. Unfortunately this is not built in, but can be easily plugged in using custom Discriminator as opposed to default MDCBasedDiscriminator:

public class ThreadNameBasedDiscriminator implements Discriminator<ILoggingEvent> {
 
    private static final String KEY = "threadName";
 
    private boolean started;
 
    @Override
    public String getDiscriminatingValue(ILoggingEvent iLoggingEvent) {
        return Thread.currentThread().getName();
    }
 
    @Override
    public String getKey() {
        return KEY;
    }
 
    public void start() {
        started = true;
    }
 
    public void stop() {
        started = false;
    }
 
    public boolean isStarted() {
        return started;
    }
}

Now we have to instruct logback.xml to use our custom discriminator:

<appender name="SIFT" class="ch.qos.logback.classic.sift.SiftingAppender">
    <discriminator class="com.blogspot.nurkiewicz.ThreadNameBasedDiscriminator"/>
    <sift>
        <appender class="ch.qos.logback.core.FileAppender">
            <file>app-${threadName}.log</file>
            <layout class="ch.qos.logback.classic.PatternLayout">
                <pattern>%d{HH:mm:ss:SSS} | %-5level | %logger{20} | %msg%n%rEx</pattern>
            </layout>
        </appender>
    </sift>
</appender>

Note that we no longer put %thread in PatternLayout – it is unnecessary as thread name is part of the log file name:

  • app-main.log
  • app-http-nio-8080-exec-1.log
  • app-taskScheduler-1
  • app-ForkJoinPool-1-worker-1.log
  • …and so forth

This is probably not the most convenient setup for server application, but on desktop where you have a limited number of focused threads like EDT, IO thread, etc. it might be a vital alternative.
 

Do you want to know how to develop your skillset to become a Java Rockstar?

Subscribe to our newsletter to start Rocking right now!

To get you started we give you our best selling eBooks for FREE!

1. JPA Mini Book

2. JVM Troubleshooting Guide

3. JUnit Tutorial for Unit Testing

4. Java Annotations Tutorial

5. Java Interview Questions

6. Spring Interview Questions

7. Android UI Design

and many more ....

 

5 comments

  1. Do you have the project on github

    • It’s just one class, you are free to copy these 28 lines to your project. Moreover I don’t have access to any maven repository. However I’m a Logback commiter so if there is as significant interest in this small feature, I can add it to core library. Would you like to create a ticket for that?

  2. How would SiftingAppender behave if the logs would go to the same logfile but only the layout/pattern would change (in your case: %d{HH:mm:ss:SSS} | %5-Level | ${threadName} | %logger{20} |%msg%n%rEX)? Wouldn’t that raise some concurrent access problems to the same log file? Thanks.

  3. Thanx alot ….. worked for me.:)

Leave a Reply

Your email address will not be published. Required fields are marked *

*


− nine = 0

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Want to take your Java Skills to the next level?
Grab our programming books for FREE!
  • Save time by leveraging our field-tested solutions to common problems.
  • The books cover a wide range of topics, from JPA and JUnit, to JMeter and Android.
  • Each book comes as a standalone guide (with source code provided), so that you use it as reference.
Last Step ...

Where should we send the free eBooks?

Good Work!
To download the books, please verify your email address by following the instructions found on the email we just sent you.