axis-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Pascale, Peter H." <Peter.Pasc...@pearson.com>
Subject RE: Axis server-side performance - concurrency
Date Tue, 08 Apr 2003 20:31:58 GMT
Good questions. The Axis server is running within Tomcat. I don't know
anything about the Axis servlet threading model? What is this? Is it
relevant? 

Peter

-----Original Message-----
From: Clover, James [mailto:James.Clover@disney.com] 
Sent: Tuesday, April 08, 2003 3:31 PM
To: axis-user@ws.apache.org
Subject: RE: Axis server-side performance - concurrency

I don't know a lot about this issue, but I've got some questions that
might shed light:  What container is the Axis server test running on?
How is the Axis servlet threading model configured?  I ask this because
the WebLogic test might be configured to execute the web service
concurrently (a normal thing for "big" application servers), where the
Axis test might not.

James

-----Original Message-----
From: Pascale, Peter H. [mailto:Peter.Pascale@pearson.com] 
Sent: Tuesday, April 08, 2003 1:19 PM
To: 'axis-user@ws.apache.org'
Cc: 'axis-dev@ws.apache.org'
Subject: Axis server-side performance - concurrency


In recent testing we've been able to replicate concurrent performance
characteristics noted on the axis-user list. A few notes about the test:

- The number of concurrent clients is configurable,
- The web service involves non-trivial complex parameter objects that,
while not large, pose a serialization/de-serialization load.
- These results come from testing with a client that uses Axis as its
SOAP engine, against an Axis-implemented service running within tomcat.
- Axis version 1.0

Concurrent Clients		1		5		10
Mean call time (milleseconds)	97.87		806.30	1730.58
Minimum call time			16		31
31
Maximum				2625		18016		34125
Number of Calls			6097		3696		3441
Call per minute			609.70	369.60	344.10
Call per second			10.16		6.16		5.74

Note the sharp rise in mean call with more than one client. Running the
same axis load test client against a service implemented on BEA WebLogic
Server 7.0 on Solaris 2.6 did not produce the concurrency issues.
Therefore I believe the issue is server-side.

My questions:

Is this a known issue? Is there something I'm missing in my service
configuration that would prevent this? Would any further data be of
interest? I may be able to capture more fine-grained detail using a
third party monitoring tool. Before I go off on some testing tangent,
I'm interested in feedback. What do you want to know? Speak know - I'd
like to be of some use to the project!

Thanks.


Peter Pascale



************************************************************************
**** 
This email may contain confidential material. 
If you were not an intended recipient, 
Please notify the sender and delete all copies. 
We may monitor email to and from our network. 
************************************************************************
****

**************************************************************************** 
This email may contain confidential material. 
If you were not an intended recipient, 
Please notify the sender and delete all copies. 
We may monitor email to and from our network. 
****************************************************************************

Mime
View raw message