DDevelopment & Design

Glimpse on Tomcat performance tuning.

  • byRafal P
  • August 23, 2010
  • 3 minute read

Production environment ;-)

Have You ever wondered about Tomcat configuration in production environment, or just let “this things” to the admins, or even worse, don’t care at all about it? If the answer is “Tomcat configuration ? I/We/Our client just installs tomcat and deploy our application. Why border about any additional configuration ?” You should read this post.

I will not write about all Tomcat’s configuration. It’s pointless. I just want to show some problems with performance with default Tomcat’s configuration in production enviroment. Especially if You are using Tomcat in as web server in internet, with many simultaneous clients and connections. In such cases performance and high responsivity is important.

1. Let’s start from logs. Standard Tomcat’s logs are configured to appear in two places: file and console. In production it’s pointless to have duplicate logs so first thing to gain some speed boost is to replace following line from logging.properties:

.handlers = 1catalina.org.apache.juli.FileHandler, java.util.logging.ConsoleHandler with this one: .handlers = 1catalina.org.apache.juli.FileHandler

2. Second thing to do with logs is to set max file size and protection from overflow. It’s also very easy. Just add new handler like following one:

catalina.java.util.logging.FileHandler

and configure it like this (max 4 filesx10Mb):

1catalina.java.util.logging.FileHandler.pattern = ${catalina.base}/logs/catalina.%g.log 1catalina.java.util.logging.FileHandler.limit = 10000000 1catalina.java.util.logging.FileHandler.count = 4

3. Last thing You have TO HAVE in production environment are asynchronous logs. Synchronous logging is far more time consuming then asynchronous one. Especially when You have numerous clients. Check if Your Tomcat is configured in proper way (I won’t write about this. Just search in web about log4j configuration. It’s lot of this there.)

4. That’s all about logging. Now something much more influent on connection speed-connectors. They are configured in server.xml under node.

Tomcat have 3 main connectors:

– BIO – Blocking Java connector which is default one

– APR – Uses native C code fo IO (very fast)

– NIO – Non blocking connectror in Java (also faster than default)

The first BIO connector (“org.apache.coyote.http11.Http11Protocol”) is set as default one. Why ? Becouse in many cases such configuration it’s enough. Tomcat usually is used in intranets where it’s not required to handle high traffic volume. Moreover BIO connector is very stable.

But if our applications have to serve many http requests the blocking connector isn’t the best choice. So here comes ARP and NIO connector.

The first one (org.apache.coyote.http11.Http11AprProtocol) requires to compile native library (just search in google for ARP) and could be less stable than BIO connector. In exchange ARP connector is very fast, could handle requests simultanously in non blocking mode, have pooling of unlimited size and could handle unlimited threads (in theory, becouse threads are limited with CPU power)

Last connector – NIO (org.apache.coyote.http11.Http11NioProtocol) is something between ARP and BIO. It’s good choice if You don’t want to compile native libraries. NIO connector is also non blocking, little slower in reading static content than ARP, but far more configurable (pool size, no of threads etc).

5. Ok, so now We know, which connector should we choose, but every connector have to be set up in proper way. There are several parameters but the important ones are:

– maxThreads – typical from 150-800 (For BIO this is max nr of open connections)

– maxKeepAliveRequests – typical 1 or 100-250. For BIO this should be set to 1 to disable keep alive (only if we have high concurency and not using SSL). BIO connector automatically disables keep alive for high connection traffic

– connectionTimeout – typical 2000-60000 WARNING: default Tomcat has it set to 20 000! It’s to high for production environment. Good choice is to decrese it to 3000-5000 unless Your production env is working with slow clients. This parameters describes max time between TCP packets during blocking read/write

6. This is “almost” the end of tunning Tomcat for production. The last thing is to configure cache. Default cache is configured to 10 MB. You can set this a little more if You have a lot of static content. Also cache revalidation (standard 5 sec) should be tuned. How ? It’s difficult to say. The best way is to tune this parameters by own during tests.

That’s all. I hope I realized to everyone why not rely on standard Tomcat configuration.

— Previous article

Meetbsd 2010

Next article —

<!--:en-->Playing with maven release plugin<!--:-->

You May Also Like
DDevelopment & Design

Sample for lift-ng: Micro-burn 1.0.0 released

During a last few evenings in my free time I've worked on mini-application called micro-burn. The idea of it appear from work with Agile Jira in our commercial project. This is a great tool for agile projects management. It has inline tasks edition, drag & drop board, reports and many more, but it also have a few drawbacks that turn down our team motivation.

Motivation

From time to time our sprints scope is changing. It is not a big deal because we are trying to be agile :-) but Jira's burndowchart in this situation draw a peek. Because in fact that chart shows scope changes not a real burndown. It means, that chart cannot break down an x-axis if we really do more than we were planned – it always stop on at most zero.

Also for better progress monitoring we've started to split our user stories to technical tasks and estimating them. Original burndowchart doesn't show points from technical tasks. I can find motivation of this – user story almost finished isn't finished at all until user can use it. But in the other hand, if we know which tasks is problematic we can do some teamwork to move it on.

So I realize that it is a good opportunity to try some new approaches and tools.

Tools

I've started with lift framework. In the World of Single Page Applications, this framework has more than simple interface for serving REST services. It comes with awesome Comet support. Comet is a replacement for WebSockets that run on all browsers. It supports long polling and transparent fallback to short polling if limit of client connections exceed. In backend you can handle pushes in CometActor. For further reading take a look at Roundtrip promises

But lift framework is also a kind of framework of frameworks. You can handle own abstraction of CometActors and push to client javascript that shorten up your way from server to client. So it was the trigger for author of lift-ng to make a lift with Angular integration that is build on top of lift. It provides AngularActors from which you can emit/broadcast events to scope of controller. NgModelBinders that synchronize your backend model with client scope in a few lines! I've used them to send project state (all sprints and thier details) to client and notify him about scrum board changes. My actor doing all of this hard work looks pretty small:

Lift-ng also provides factories for creating of Angular services. Services could respond with futures that are transformed to Angular promises in-fly. This is all what was need to serve sprint history:

And on the client side - use of service:


In my opinion this two frameworks gives a huge boost in developing of web applications. You have the power of strongly typing with Scala, you can design your domain on Actors and all of this with simplicity of node.js – lack of json trasforming boilerplate and dynamic application reload.

DDD + Event Sourcing

I've also tried a few fresh approaches to DDD. I've organize domain objects in actors. There are SprintActors with encapsulate sprint aggregate root. Task changes are stored as events which are computed as a difference between two boards states. When it should be provided a history of sprint, next board states are computed from initial state and sequence of events. So I realize that the best way to keep this kind of event sourcing approach tested is to make random tests. This is a test doing random changes at board, calculating events and checking if initial state + events is equals to previously created state:



First look

Screenshot of first version:


If you want to look at this closer, check the source code or download ready to run fatjar on github.During a last few evenings in my free time I've worked on mini-application called micro-burn. The idea of it appear from work with Agile Jira in our commercial project. This is a great tool for agile projects management. It has inline tasks edition, drag & drop board, reports and many more, but it also have a few drawbacks that turn down our team motivation.
DDevelopment & Design

Mentoring in Software Craftsmanship

Maria Diaconu and  Alexandru Bolboaca are both strong supporters of Software Craftsmanship and Agile movements in Romania, with years of experience as developers, leaders, architects, managers and instructors. On their lecture at Agile Central Eur...Maria Diaconu and  Alexandru Bolboaca are both strong supporters of Software Craftsmanship and Agile movements in Romania, with years of experience as developers, leaders, architects, managers and instructors. On their lecture at Agile Central Eur...