[mule-user] compare between open source ESB products (JBoss esb, Openesb, Mule ESB and Apache Service Mix))

[Younes Yahyaoui]

Hallo,

i want to compare the following open source esb products: *Mule esb, Sun open esb , JBoss esb and Apache ServiceMix*. Hat someone any idea about a progrmming example or about a roadmap to efficiency compare them und test their performance?

thanks in advance

Younes

---------------------------------------------------------------------
To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

[David Dossot]

Younes,

If the only criteria you're interested in during this evaluation is performance, then... good luck. Comparing ESB performance has been done, discussed, debated and AFAIAC the only conclusion is that, with expert level knowledge in each ESB, you can pretty much end-up with similar results on all of them.

The key words here are: expert level knowledge.

If you perform some quick performance tests based on novice knowledge of the different ESBs you want to compare, your results will say nothing.

Would you consider other aspects in your evaluation? Like: ease of configuration, supported transports, deployment models, existing code reuse, community and professional support...?

Food for thought, I guess.

Cheers,
D.

[andy e]

Younes,

Pick up Open Source in Action from Manning Publishers (manning.com).
It goes through examples using Mule and ServiceMix. And while you're
at it, pick up Mule In Action as well, because it's great. Manning has
lots of 40-50% off coupons every week or two, so you can probably get
them both for $50US.

Andy

[Olivier Roger]

Hello everyone,

I am also interested in open-source ESB comparison.
The criteria David Dossot gave are worth investigating.

I would also add criteria like the project
- technical level required
- long-term viability (e.g. for Open ESB)
- scalability
- availability of management and monitoring tools
- quality/price of the support
- standard compliance

If someone has some information about some of those criteria I am also strongly interested

Olivier

[rotten]

Don't forget Petals also: http://petals.ow2.org/index.html

--
View this message in context: http://old.nabble.com/compare-between-open-source-ESB-products-%28JBoss-esb%2C-Openesb%2C-Mule-ESB-and-Apache-Service-Mix%29%29-tp27546558p27550599.html
Sent from the Mule - User mailing list archive at Nabble.com.

[Asankha C. Perera]

Hi David / Younes

I agree with David's comment that performance comparison tends to depend on the level of knowledge of the person conducting the test

I've tested the performance of many ESBs since June 2007 - including Mule, ServiceMix, Synapse/WSO2 and other proprietary versions, and after the third round of testing [http://wso2.org/library/3740], Dan from Mule ran the same test against Mule with a more 'optimized' configuration that revealed better performance from Mule - although the configuration was a bit functionally different from the scenarios tested against others. [See the above link for a reference to the Mule performance report]

Thus to prevent such an issue, I've just released the results of Round 4 - where I do not publicly release results of any ESB, but make it trivial for any ESB vendor to publicly make available the required configuration - so that end users (like Younes) can run the same tests - on the same hardware etc - leaving everything else constant! The best part is that you can now run this all on Amazon EC2 for less than $2 of computing time.

I hope Mule and other ESB vendors will publicly make available the necessary configurations so that we could all rely on an accurate and fair benchmark. The Round 4 results includes the three test cases conducted earlier by Mule (i.e. Direct Proxy, Content Based Routing [CBR] Proxy, and XSLT Proxy), and adds a new scenario for WS-Security processing.

http://adroitlogic.org/samples-articles-and-tutorials/15-tutorials/48-esb-performance.html

cheers
asankha

>Younes,
>
>If the only criteria you're interested in during this evaluation is performance, then... good luck. Comparing ESB performance has been done, discussed, >debated and AFAIAC the only conclusion is that, with expert level knowledge in each ESB, you can pretty much end-up with similar results on all of them.
>
>The key words here are: expert level knowledge.
>
>If you perform some quick performance tests based on novice knowledge of the different ESBs you want to compare, your results will say nothing.
>
>Would you consider other aspects in your evaluation? Like: ease of configuration, supported transports, deployment models, existing code reuse, >community and professional support...?
>
>Food for thought, I guess.
>
>Cheers,
>D.

On Thu, Feb 11, 2010 at 4:34 AM, Younes Yahyaoui < wrote:

Hallo,

 i want to compare the following open source esb products: *Mule esb, Sun open esb , JBoss esb and Apache ServiceMix*. Hat someone any idea about a progrmming example or about a roadmap to efficiency compare them und test their performance?

thanks in advance

Younes

-- 
Asankha C. Perera
AdroitLogic, http://adroitlogic.org

http://esbmagic.blogspot.com

[Khaled BENDRISS]

Please note  the Mule performance test results at : http://www.mulesoft.com/downloads/Whitepaper_perf_test_results.pdf

Mule 2.0.2 Vs WSO ESB 1.7

The benchmark configurations and scripts that were used are available at:
http://www.mulesoft.com/downloads/mule-benchmark.zip


--
Khaled BEN DRISS

http://net-progress.blogspot.com
http://www.journaldunet.com/developpeur/expert/2500/khaled_ben-driss/
http://twitter.com/netprogress

2010/2/11 Asankha C. Perera <asa...@adroitlogic.com>

[Asankha C. Perera]

Hi Khaled

> Please note the Mule performance test results at :
> http://www.mulesoft.com/downloads/Whitepaper_perf_test_results.pdf

Thanks for the correct link - and I had earlier linked to mulesource
domain by mistake.

> The benchmark configurations and scripts that were used are available at:
> http://www.mulesoft.com/downloads/mule-benchmark.zip

It would be great if someone could include the configuration for the
WS-Security test cases too

cheers
asankha

--
Asankha C. Perera
AdroitLogic, http://adroitlogic.org

http://esbmagic.blogspot.com

---------------------------------------------------------------------

[Younes Yahyaoui]

Hi Asankha C. Perera

you are said: "
trivial for any ESB vendor to publicly make available the required
configuration - so that end users (like Younes) can run the same tests -
on the same hardware etc - leaving everything else constant! The best
part is that you can now run this all on Amazon EC2 for less than $2 of
computing time."

i am intersting to start an ESB als virtual machine in Amazon EC2, e.g Mule ESB or JBoss ESB . Do you Know how to make this possible?

thanks in advance

Younes

[Asankha C. Perera]

Hi Younes

> Hi Asankha C. Perera
>
> you are said: "
> trivial for any ESB vendor to publicly make available the required
> configuration - so that end users (like Younes) can run the same tests -
> on the same hardware etc - leaving everything else constant! The best
> part is that you can now run this all on Amazon EC2 for less than $2 of
> computing time."
>
> i am intersting to start an ESB als virtual machine in Amazon EC2, e.g Mule ESB or JBoss ESB . Do you Know how to make this possible?
>

Setting up any ESB on Amazon EC2 is quite straight forward, and you
could then create an image off it.. but if you require a Mule ESB or
JBoss ESB image off the shelf, it should be from; or, as per the
instructions shared by those vendors or the communities behind them. I
can only make an image of the UltraESB available for you if you want to
perform a comparison

cheers
asankha

--
Asankha C. Perera
AdroitLogic, http://adroitlogic.org

http://esbmagic.blogspot.com

---------------------------------------------------------------------

[Richard L. Burton III]

I can speak on behalf of Mule vs. JBoss ESB. I have experience with ServiceMix + Camel, but not enough that I deem it to be worth me injecting my thoughts into this discussion. The level of knowledge with ServiceMix is rather shallow.

JBoss ESB is honestly a poorly architected ESB. It's based upon a project called Rosetta, that at the time of it's creation was considered usable, much like how Struts 1 was during it's initial release. Mule and others took the lessons learned over the years (Enterprise Integration Patterns EIP is a great read) and applied them. Let me provide some examples to further express this.

Mule has awesome integration with Spring, where JBoss ESB simply has hacked solution with all due respect. In JBoss ESB, your ActionPipelineProcess (This is your 'business' component) must extend AbstractSpringAction which only handles the creation of the ApplicationContext. This makes unit tested a little harder since now you must initialize Spring or write code that's only for unit testing. I can provide some actual code examples if you're interested in hearing more about this.

Additionally, the internals of Rosetta leak into your ActionPipelineProcessor in the form of a ConfigTree. The ConfigTree is an internal representation of the configuration information provided by the jboss-esb.xml file, it's basically a Map in it's simplest form. Instead of JBoss taking advantage of the bean specification, you the developer must pull values from the ConfigTree.

This introduces another interesting problem. They provide tooling support in the form of the JBoss Developer Studio (Basically Eclipse with JBoss Plug-ins). When using custom actions, there's no way to auto detect what configuration values are available and this requires you to look at the Java Docs or documentation. It would have been awesome if they followed the Bean specification in this case. Then the tool can use reflection and tell the developer what's available.

Other fairly common complaints are: Poor Maven support, tight integration with the JBoss Application Server, slow startup times, etc. Simple tasks require more coding than is required and makes simple tasks harder than they should be.

If you're interested in hearing more about my experience with JBoss ESB, I'd be more than happy to extend my contact information and you can freely ask my questions.

Cheers,

Richard L. Burton III

---------------------------------------------------------------------
To unsubscribe from this list, please visit:

http://xircles.codehaus.org/manage_email

--
You received this message because you are subscribed to the Google Groups "mule-user" group.
To post to this group, send email to mule...@googlegroups.com.
To unsubscribe from this group, send email to mule-user+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mule-user?hl=en.

[psmithinoz]

Hi Richard, I would be extremely interested in your comparisons with JBoss ESB. I am of the same opinion as yourself however I'm favouring Open ESB at the moment because of the tooling support. We played with JBoss ESB and found it seriously lacking however we have some "stick in the mud" colleques who refuse to give up their JBoss security blanket. Any information you have would be greatly appreciated.

[harshajith halgaswatta]

Hi Richard,

Thanks for your explanation. it is really worth.I am also in between evaluating Mule ESB vs JBoss ESB. I felt it too that JBoss ESB is still in the stage of incubation, Mule has matured in easy of configuration, spring Integration, third party tools integration etc.

One of my problems is JBoss ESB itself provides a simple management interface, but we do not have anything in Mule community edition. Because we also have a requirement for auditing features.

Appreciate your comments on this.

Thanks
Harsha

[sulemanzia]

where is the love for spring integration ? :P

[Andrew Perepelytsya]

Perhaps he meant integration with spring? It would make little sense to integrate with Spring Integration product, as they overlap. Indirectly it's all possible still, as it's still a spring config.

Andrew