JMS application is slow

Hi All,
I am facing avery typical scenario.
We have two boxes, one for the DEV and other for the UAT.
Now on the Dev Box (LINUX and MQ) I have a sample program which is able to push N number of messages on the MQ (The communication is via the connection Factory)
However on the UAT box which is a very high end machine we are able to push N/2 messages per second with the same set of all the configuration.

Request you to kindly advise on the same.

OS: Linux 2.6.18-194.32.1.el5 #1 SMP Mon Dec 20 10:52:42 EST 2010 x86_64 x86_64 x86_64 GNU/Linux (Both DEV and UAT)
JAVA : java version "1.6.0_20"
Java(TM) SE Runtime Environment (build 1.6.0_20-b02)
Java HotSpot(TM) 64-Bit Server VM (build 16.3-b01, mixed mode) {For both DEV and UAT}

The only diffrence in the Hardware is the extra layer of VCFS - veritas common file system in UAT Machine

Gagan,
You'll need to do some performance research on the UAT machine. Places to start:

Is the box stressed from running other things? Check CPU, I/O, etc.

What is the bottleneck? Could it be another part of your code and the MQ part is just the symptom?

many thanks for the reply here are the things as you suggested
1. The UAT box is having nothing apart from our application, the CPU and I/O are also fine.
2. The Code is the simplest of the Java Code which just reads the XML and puts it in the MQ.

Is there anything in MQ by which we can check, may be some command or any files (whih MQ generates )

How about network differences ?

Can you please elaborat eon what we need to check on the Network diffrence?

If you posting messages to a remote queue via a local alias then the speed at which the data can be put into the queue might be effected by the speed of the connection between the 2 machines. Or if there destination is different (dev and test) then this might also be a factor. Finally - I recall that queues can be setup differently in MQ - you can specify how resilient they are to be - can't remember the exact term it uses, but for the sake of argument lets suppose the dev environment has been setup so that it's using MQ in a non-fault tolerant manner and testing is done with a queue in safe fault tolerant mode - it could have an impact.

Hi David,
Many thanks for the prompt response.

Let me just elaborate a bit more on the same

1. Machine1. We are able to put X message per second on the QUEUE. This Queue is local to the Queue Manager.
2. Machine2. Same step done on machine 2, queue is also the Local Queue of the QueueManager, the rate is X/2.

We are just PUTTING THE MESSAGE on the Queue, nothing else.

Regarding the Machine1 and Machine2 both are having independent QueueManagers.

Also the Details of the BOX are as follows (Both are same)
1. MQ Version
Name: WebSphere MQ
Version: 7.0.1.3
CMVC level: p701-103-100813
BuildType: IKAP - (Production)

2. Operating System
Linux gcfxstpora1ffu 2.6.18-194.32.1.el5 #1 SMP Mon Dec 20 10:52:42 EST 2010 x86_64 x86_64 x86_64 GNU/Linux

3. Java Version
java version "1.6.0_22"
Java(TM) SE Runtime Environment (build 1.6.0_22-b04)
Java HotSpot(TM) 64-Bit Server VM (build 17.1-b03, mixed mode)

We analyzed the logs and found that we are hhaving the below process extra in Machine2 (Where the speed is slow)

14:57:08.562922 24832.3 SHCON:000003 -----{ xcsConvertString
14:57:08.562934 24832.3 SHCON:000003 fromCCSID:1208 toCCSID:819, in length:4, out length:4 options: 00005001
14:57:08.562937 24832.3 SHCON:000003 ------{ xcsQueryCCSIDType
14:57:08.562940 24832.3 SHCON:000003 ------} xcsQueryCCSIDType rc=OK
14:57:08.562942 24832.3 SHCON:000003 ------{ xcsQueryCCSIDType
14:57:08.562944 24832.3 SHCON:000003 ------} xcsQueryCCSIDType rc=OK
14:57:08.562947 24832.3 SHCON:000003 translated out length:4
14:57:08.562949 24832.3 SHCON:000003 -----} xcsConvertString rc=OK

Any idea, kindly let me know