ServiceMix usage

[johne <je...@yahoo.com>]

Is there a listing of users that currently use ServiceMix.   Testimonies? 
Successful deployments in an environment requiring scaling/high performance?

Having this information would help me determine whether to use ServiceMix as
well as help me convey confidence in its use.  I believe this would be
important and useful to others as well.

Thanks,

John

-- 
View this message in context: http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16257034.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.

Re: ServiceMix usage

[Chris Custine <cc...@apache.org>]

There is a list here: http://servicemix.apache.org/users.html but it is
probably extremely out of date an certainly not all encompassing.  Because
of the open source nature, any disclosure of usage of ServiceMix is going to
be entirely voluntary on the part of the user.  There are substantially more
(and larger) users of ServiceMix that are not listed.

Chris


On Mon, Mar 24, 2008 at 11:45 AM, johne <je...@yahoo.com> wrote:

>
> Is there a listing of users that currently use ServiceMix.   Testimonies?
> Successful deployments in an environment requiring scaling/high
> performance?
>
> Having this information would help me determine whether to use ServiceMix
> as
> well as help me convey confidence in its use.  I believe this would be
> important and useful to others as well.
>
> Thanks,
>
> John
>
> --
> View this message in context:
> http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16257034.html
> Sent from the ServiceMix - User mailing list archive at Nabble.com.
>
>

Re: ServiceMix usage

[Terry Cox <te...@meta-concepts.com>]

Unfortunately, most organisations consider their choice of back-end 
integration technologies to be commercially sensitive information for 
obvious reasons. I can certainly assure you that a number of prominent 
international enterprises are using the technology.

At the end of the day, the only way to be sure is to carry out a 
comparative evaluation of various products and do a cost/benefit analysis.

Terry

Re: ServiceMix usage

[Bruce Snyder <br...@gmail.com>]

On Tue, Mar 25, 2008 at 12:27 PM, johne <je...@yahoo.com> wrote:
>
>  Thank you for such a detailed response.  I think if more people saw responses
>  like this, they would be much more likely to jump on board.
>
>  There is no definite, specific application we will be using this in right
>  now, but it is on the table as one possible option to achieve
>  interoperability and scalability improvements among our existing products.
>  Without feedback on use from how other companies are using ServiceMix, we
>  could not even consider its use.  I am trying to determine whether it can
>  speed up development, create a separation of concerns within our framework,
>  be reliable, perform, scale, etc.  I really like its architecture for many
>  types of applications, but I have be able to prove not only can it work but
>  that it has worked.

As has been explained by folks already, many companies won't publicly
speak about their back end systems because they consider that
information to be part of their competitive advantage. Companies are
incredibly protective of such information and actively work to protect
it.

I have been involved with many ServiceMix implementations at various
companies all over the world; unfortunately I'm not able to call them
out by name. These companies come from a wide range of industries
throughout the world, some of which are Fortune 50 companies. Think of
industries such as financial, banking, telecommunications, IT, retail,
aerospace, manufacturing, gaming, travel, e-commerce, government
agencies, etc. These implementations run the gamut of use cases. Some
include an incredible amount of transactions per day while others
don't do a tremendous number of transactions but the value of each
transaction is much higher. There are even multiple use cases where
vendors are using ServiceMix as the base for their own product (in a
sort of OEM style of use). There are a wide range of ServiceMix
implementations out there already and the growth continues to inrease
steadily.

Bruce
-- 
perl -e 'print unpack("u30","D0G)U8V4\@4VYY9&5R\"F)R=6-E+G-N>61E<D\!G;6%I;\"YC;VT*"
);'

Apache ActiveMQ - http://activemq.org/
Apache Camel - http://activemq.org/camel/
Apache ServiceMix - http://servicemix.org/
Apache Geronimo - http://geronimo.apache.org/

Blog: http://bruceblog.org/

Re: ServiceMix usage

[johne <je...@yahoo.com>]

Thank you for such a detailed response.  I think if more people saw responses
like this, they would be much more likely to jump on board.

There is no definite, specific application we will be using this in right
now, but it is on the table as one possible option to achieve
interoperability and scalability improvements among our existing products.  
Without feedback on use from how other companies are using ServiceMix, we
could not even consider its use.  I am trying to determine whether it can
speed up development, create a separation of concerns within our framework,
be reliable, perform, scale, etc.  I really like its architecture for many
types of applications, but I have be able to prove not only can it work but
that it has worked.

-- 
View this message in context: http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16283628.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.

Re: ServiceMix usage

[mattrpav <ma...@airband.com>]

John-

No sweat, the company I was working for was bought and the buying company
decided that they needed to use Tibco, because it was "bigger", more
expensive and took longer to implement projects.  (Yeah, it didn't make any
sense to me either, but they were "striving to be average" so I guess that's
good enough.)

We had a few issues with early versions in the 2.x series, but after 3.0
things really stabilized to what I considered to be "Enterprise Ready". 
When you implement clustering, the underlying technology piece that is used
is ActiveMQ, which has been around a long time and is very stable and
performs at a very high level.   I do not have any performance
considerations whatsoever with ServiceMix 3.x.  The only limitation I ever
experienced was with as our back-end persistence database that was running
on a slightly under powered server.  With a 2 server 2 CPU setup, we could
have easily quadrupled our message handling.   ServiceMix 3.x is very
serviceable, and I imagine most knowledge gained on ServiceMix 3.x will be
applicable for 4.x.  

As far as interfaces go, we used mostly HTTP, and SOAP.  I've personally
worked with the CXF, Camel, WSN, EIP and FTP modules with success.  The
components are all very similar to config/setup, and you always have the
option of going straight Java beans as a fall back.  In a previous life, I
used Axis on Tomcat for SOAP Web Services, but have started using CXF on
ServiceMix instead.  There are a lot of advantages to this, one being that
consuming services between each other is a lot easier.

The learning curve with ServiceMix is the project skeleton-- how to build a
project to deploy on the platform.  Maven makes things easier, and the
ServiceMix examples give you a good baseline to work from.  I hope you don't
think that building projects on ServiceMix is hard, because it is a "large"
project.  Once you get the first one going, it is smooth sailing.

I've used Weblogic Integration Platform, Tibco, and WebMethods.  ServiceMix
performs on par or better than all of them.  ServiceMix's tooling is a bit
behind, but once you get ServiceMix running under Eclipse, and remote
debugging working, you are home free.  (Tip:  Use a Maven Eclipse plugin and
all your library dependency issues will magically disappear).  

Most of the tuning parameters are at the JVM layer, and if you need to tweak
thread pools, you can do that easily through any JMX console.  Here are my
JVM parameters that I add to the bin/servicemix startup script:

-Xmx1024M -Xms512M -XX:MaxPermSize=256m -server

-XX:MaxPermSize is important if you are anticipating a lot of requests via
many parallel threads.  ServiceMix enables JMX by default, so I add a JVM
Service adapter to Hyperic to monitor the memory pool usage.  This helps
identify memory leaks and/or performance tuning issues.

What types of projects are you looking to implement?  Can you share any
details about anticipated load, etc?

Matt Pavlovich  
-- 
View this message in context: http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16267572.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.

Re: ServiceMix usage

[johne <je...@yahoo.com>]

Thank you for the information Matt.  I appreciate your mentioning having had
success.  I understand this is very proprietary types of information and
respect that, but am wondering if you go against multiple interfaces and how
you feel about both any positive and negative experiences of implementation,
performance, etc.  I have been following this project for a long time and
realize it is quite a large undertaking.  I am excited about the 4.0 version
of ServiceMix and knowing others are having success would help our company
come to a decision on whether or not we can use this.

-- 
View this message in context: http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16267146.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.

Re: ServiceMix usage

[mattrpav <ma...@airband.com>]

I've managed implementations for two companies.  The largest was a $1B/yr
annual revenue telecommunications company that used ServiceMix for its
integration layer for internal back office systems integration as well as
e-bonding.  The platform handled 300,000-1Million events/month.  

Running on Linux in a 2 server cluster environment.  2CPUs, 4Gb of RAM each. 
Oracle 9i for persistence.

Matt Pavlovich

-- 
View this message in context: http://www.nabble.com/ServiceMix-usage-tp16257034s12049p16263769.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.