blob: ae5d07271a10063f6d6fcacac50fb512df6209dd [file] [log] [blame]
Licensed to the Apache Software Foundation (ASF) under one or more
contributor license agreements. See the NOTICE file distributed with
this work for additional information regarding copyright ownership.
The ASF licenses this file to You under the Apache License, Version 2.0
(the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
See the License for the specific language governing permissions and
limitations under the License.
Benchmarking Tomcat
Tomcat benchmarks are more than welcome, and people who are interested are
encouraged to post results.
This guide is designed to be a FAQ on how to configure Tomcat for maximum
- Please upgrade to the latest available version of the Tomcat branch you
would like to benchmark. Small incremental performance improvements are often
made in each release.
- Use the most recent JDK available whenever possible. Usually, results are
dependent on the JDK, so results obtained with different JDK should not be
compared directly.
- When benchmarking Jasper, set it in production mode. To do that, add a
"development" init parameter (with value set to "false") in the Jasper
section of $CATALINA_HOME/conf/web.xml. This increases maximum throughtput
of JSPs by about 20%, as well as removing a synchronization point.
- When benchmarking a web application which creates sessions on each request,
be careful about not exhausting memory by having too many active sessions. If
a max number of sessions is set in the manager to avoid memory problems, this
will lead to requests failing when session creation is denied. To avoid that,
either limit the total number of requests, or increase the JVM memory.
- Do not enable access logging (unless that is required by the benchmark),
or any debug log.
- Please post the benchmark web application used along with the results, so
that people may attempt to reproduce the results.
- Definition of a standard webapp workload may be desirable, and contibutors
interested in doing that are welcome to discuss it on the Tomcat developer
mailing list.