Bridging legacy APIs

Often, some of the components you depend on rely on a logging API other than SLF4J. You may also assume that these components will not switch to SLF4J in the immediate future. To deal with such circumstances, SLF4J ships with several bridging modules which redirect calls made to log4j, JCL and java.util.logging APIs to behave as if they were made to the SLF4J API instead. The figure below illustrates the idea.

Please note that for source code under your control, you really should use the slf4j-migrator. The binary-based solutions described in this page are appropriate for software beyond your control.

click to enlarge

Gradual migration to SLF4J from Jakarta Commons Logging (JCL)


To ease migration to SLF4J from JCL, SLF4J distributions include the jar file jcl-over-slf4j.jar. This jar file is intended as a drop-in replacement for JCL version 1.1.1. It implements the public API of JCL but using SLF4J underneath, hence the name "JCL over SLF4J."

Our JCL over SLF4J implementation will allow you to migrate to SLF4J gradually, especially if some of the libraries your software depends on continue to use JCL for the foreseeable future. You can immediately enjoy the benefits of SLF4J's reliability and preserve backward compatibility at the same time. Just replace commons-logging.jar with jcl-over-slf4j.jar. Subsequently, the selection of the underlying logging framework will be done by SLF4J instead of JCL but without the class loader headaches plaguing JCL. The underlying logging framework can be any of the frameworks supported by SLF4J. Often times, replacing commons-logging.jar with jcl-over-slf4j.jar will immediately and permanently solve class loader issues related to commons logging.


Some of our users after having switched to SLF4J API realize that in some contexts the use of JCL is mandatory and their use of SLF4J can be a problem. For this uncommon but important case, SLF4J offers a JCL binding, found in the file slf4j-jcl.jar. The JCL binding will delegate all logging calls made through SLF4J API to JCL. Thus, if for some reason an existing application must use JCL, your part of that application can still code against the SLF4J API in a manner transparent to the larger application environment. Your choice of SLF4J API will be invisible to the rest of the application which can continue to use JCL.

jcl-over-slf4j.jar should not be confused with slf4j-jcl.jar

JCL-over-SLF4J, i.e. jcl-over-slf4j.jar, comes in handy in situations where JCL needs to be supported for backward compatibility reasons. It can be used to fix problems associated with JCL, without necessarily adopting the SLF4J API, a decision which can be deferred to a later time.

On the other hand, slf4j-jcl.jar is useful after you have already adopted the SLF4J API for your component which needs to be embedded in a larger application environment where JCL is a formal requirement. Your software component can still use SLF4J API without disrupting the larger application. Indeed, slf4j-jcl.jar will delegate all logging decisions to JCL so that the dependency on SLF4J API by your component will be transparent to the larger whole.

Please note that jcl-over-slf4j.jar and slf4j-jcl.jar cannot be deployed at the same time. The former jar file will cause JCL to delegate the choice of the logging system to SLF4J and the latter jar file will cause SLF4J to delegate the choice of the logging system to JCL, resulting in an infinite loop.


SLF4J ship with a module called log4j-over-slf4j. It allows log4j users to migrate existing applications to SLF4J without changing a single line of code but simply by replacing the log4j.jar file with log4j-over-slf4j.jar, as described below.

How does it work?

The log4j-over-slf4j module contains replacements of most widely used log4j classes, namely org.apache.log4j.Category, org.apache.log4j.Logger, org.apache.log4j.Priority, org.apache.log4j.Level, org.apache.log4j.MDC, and org.apache.log4j.BasicConfigurator. These replacement classes redirect all work to their corresponding SLF4J classes.

To use log4j-over-slf4j in your own application, the first step is to locate and then to replace log4j.jar with log4j-over-slf4j.jar. Note that you still need an SLF4J binding and its dependencies for log4j-over-slf4j to work properly.

In most situations, replacing a jar file is all it takes in order to migrate from log4j to SLF4J.

Note that as a result of this migration, log4j configuration files will no longer be picked up. If you need to migrate your file to logback, the log4j translator might be of help. For configuring logback, please refer to its manual.

When does it not work?

The log4j-over-slf4j module will not work when the application calls log4j components that are not present in the bridge. For example, when application code directly references log4j appenders, filters or the PropertyConfigurator, then log4j-over-slf4j would be an insufficient replacement for log4j. However, when log4j is configured through a configuration file, be it or log4j.xml, the log4j-over-slf4j module should just work fine.

What about the overhead?

There overhead of using log4j-over-slf4j instead of log4j directly is relatively small. Given that log4j-over-slf4j immediately delegates all work to SLF4J, the CPU overhead should be negligible, in the order of a few nanoseconds. There is a memory overhead corresponding to an entry in a hashmap per logger, which should be usually acceptable even for very large applications consisting of several thousand loggers. Moreover, if you choose logback as your underlying logging system, and given that logback is both much faster and more memory-efficient than log4j, the gains made by using logback should compensate for the overhead of using log4j-over-slf4j instead of log4j directly.

log4j-over-slf4j.jar and slf4j-log4j12.jar cannot be present simultaneously

The presence of slf4j-log4j12.jar, that is the log4j binding for SLF4J, will force all SLF4J calls to be delegated to log4j. The presence of log4j-over-slf4j.jar will in turn delegate all log4j API calls to their SLF4J equivalents. If both are present simultaneously, slf4j calls will be delegated to log4j, and log4j calls redirected to SLF4j, resulting in an endless loop.

jul-to-slf4j bridge

The jul-to-slf4j module includes a java.util.logging (jul) handler, namely SLF4JBridgeHandler, which routes all incoming jul records to the SLF4j API. Please see SLF4JBridgeHandler javadocs for usage instructions.

Note on performance Contrary to other bridging modules, namely jcl-over-slf4j and log4j-over-slf4j, which reimplement JCL and respectively log4j, the jul-to-slf4j module does not reimplement the java.util.logging because packages under the java.* namespace cannot be replaced. Instead, jul-to-slf4j translates LogRecord objects into their SLF4J equivalent. Please note this translation process incurs the cost of constructing a LogRecord instance regardless of whether the SLF4J logger is disabled for the given level or nor. Consequently, j.u.l. to SLF4J translation can seriously increase the cost of disabled logging statements (60 fold or 6000%) and measurably impact the performance of enabled log statements (20% overall increase). As of logback version 0.9.25, it is possible to completely eliminate the 60 fold translation overhead for disabled log statements with the help of LevelChangePropagator.

If you are concerned about application performance, then use of SLF4JBridgeHandler is appropriate only if any one of the following two conditions is true:

  1. few j.u.l. logging statements are in play
  2. LevelChangePropagator has been installed

jul-to-slf4j.jar and slf4j-jdk14.jar cannot be present simultaneously

The presence of slf4j-jdk14.jar, that is the jul binding for SLF4J, will force SLF4J calls to be delegated to jul. On the other hand, the presence of jul-to-slf4j.jar, plus the installation of SLF4JBridgeHandler, by invoking "SLF4JBridgeHandler.install()" will route jul records to SLF4J. Thus, if both jar are present simultaneously (and SLF4JBridgeHandler is installed), slf4j calls will be delegated to jul and jul records will be routed to SLF4J, resulting in an endless loop.