New J2EE vs .NET Performance Comparison Performed -- PANIC in Java J2EE-land!

The Middleware Companhy ^ | 10/28/2002 | The Middleware Company

[old-ager]

The Middleware Company has performed a new comparison of the performance and scalability of J2EE and .NET based on the familiar Pet Store application. This time, the Middleware Company has re-coded the J2EE Petstore and optimized the implementation for performance. In the comparison, a new implementation of the .NET Pet Shop has also been tested. This implementation uses dynamic SQL instead of stored procedures, and like the J2EE equivalent is an object-oriented, logical 3-tier implementation following Microsoft's recommended design pattern for building scalable Web applications.

In addition, the new performance and scalability comparison includes new comparative performance data on .NET and J2EE for XML Web Services as well as distributed transactions handled via .NET/COM+ and J2EE/JTA. The performance study was conducted by the Middleware Company on the same hardware and test lab for all tested solutions, and the results are certified by the Middleware Company.

CLICK HERE FOR A LINK TO THE DISCUSSION, WHICH HAS A LINK TO A PDF OF THE ACTUAL REPORT

[old-ager]

Even after a rewrite of the Java J2EE Petstore by one of the top J2EE consultants, C# .NET Pet Shop is still much faster, handles more than twice the user load, requires far less code (2k lines vs 14k lines), and comes in at half the $ / transaction. Hats off to the Middleware Company for their objectivity and honesty. Java Religionists, start your FUD-throwers! ;-)

[old-ager]

[dheretic]

The only problem with DotNET is that there isn't a MS-independent implementation. Mono:: looks promising though and Borland announced that they may be using it as the foundation for their DotNET projects on UNIX. Of course as we all know only godless communists and flaming homosexuals use UNIX whereas only good, clean-cut, god-fearing capitalists use Microsoft products (lol that's such an oxymoron, "god-fearing capitalist....")

[ThinkDifferent]

Refuted here. Basically the J2EE implementation was not optimized at all, and used a number of clearly bad design techniques.

[old-ager]

"foundation for their DotNET projects on UNIX"

That's interesting. I knew they were going to support .NET in C++ and Delphi, but assumed they meant .NET on Windows.

[old-ager]

"Basically the J2EE implementation was not optimized at all, and used a number of clearly bad design techniques."

But would you say that such bad designs are really unusual for the Java/J2EE culture? I would say that the kind of bad design you see is typical, and not only that, is encouraged. For example, if Entity Beans were not suited to this project, for what would they be suited? I think the answer is that J2EE itself encourages bad design. Yes, you may be able to create a somewhat fast and efficient system using J2EE, but only by violating most of the "community norms".

[D-fendr]

these web applications will only run on Windows boxes?

[BuddhaBoy]

I am a huge fan of .NET.

But in fairness, the J2EE implementation was anything BUT optimized. Many basic architectural mistakes were made, and for the sake of credibility, this whole set of results should be thrown out.

It doesnt benefit Microsoft to show an optimized solution against one who's coders didnt know what they were doing.

[old-ager]

"But in fairness, the J2EE implementation was anything BUT optimized"

A couple of points: 1. This version of the MS implementation has been somewhat de-optimized because of feedback from Java people. 2. I don't think the claim is that the J2EE implementation is optimized -- only that it's per community norms and recommendations. The fact is that best practices in the MS part of the world are already geared towards performance, but best practices in the Java world are geared towards ideology and groupthink.

[old-ager]

"It doesnt benefit Microsoft to show an optimized solution against one who's coders didnt know what they were doing."

The problem with this kind of statement is that it implies that there is some way to make a J2EE box perform as well as a .NET box. I don't think there's any way to do this! If it's so obvious, where is such a rewrite posted and benchmarked? Show us.

[BuddhaBoy]

I agree with you that it isnt possible to make a J2EE box perform as well as .NET, but I think Microsoft needs to make this case in a more fair manner.

The J2EE implementation is some of the stupidest coding I have ever seen, even for Java. This is true because most people love Java for the same reasons that many love Visual Basic. It allows idiots to be productive.

As far as I know, the only people who have been able to wring any kind of performance from Java, is.......Microsoft.

[old-ager]

"As far as I know, the only people who have been able to wring any kind of performance from Java, is.......Microsoft."

Heh heh. Yes, they did an incredible job very early on. IBM got a pretty good reputation too. Any of course the MS benchmark could be rewritten in J#.

Java partisans are mad at TSS and TMW about this, and accusing them of being, basically, dishonest. But isn't it possible that TMW has decided _objectively_ and honestly (before publishing the benchmark) to go with .NET? We have to assume that they were pro-Java when they started. What changed their minds? Some Javans will say "money" but there are two ways of looking at that too. There's honest money and dishonest money. And that's where the real debate is. The MS-haters claim that MS's money is dishonest. Well, they have never forced ME to buy or do anything!