Question

I have some code which is doing this:

String transactionId = incoming.getJMSCorrelationID();

when I used ActiveMQ for testing I was able to send messages through and pick up this value without problem. However, when I am connected to Websphere the value comes through prefixed with ‘ID:’ I have no idea why this should be the case – the middleware people tell me that my code must be adding it – which it’s not.

It’s a simple case for me to remove it, but I want to know why this prefix appears – any ideas anyone?

Answer 1

I believe the middleware people are wrong:

JMSMessageID from MQMD MessageID

All messages sent from JMS have unique message identifiers assigned by WebSphere® MQ. The value assigned is returned in the MQMD messageId field after the MQPUT call, and is passed back to the application in the JMSMessageID field. The WebSphere MQ messageId is a 24-byte binary value, whereas the JMSMessageID is a string. The JMSMessageID is composed of the binary messageId value converted to a sequence of 48 hexadecimal characters, prefixed with the characters ID:. JMS provides a hint that can be set to disable the production of message identifiers. This hint is ignored, and a unique identifier is assigned in all cases. Any value that is set into the JMSMessageId field before a send() is overwritten.

From: Mapping JMS header fields at send() or publish().

Also I remember noticing that prefix while working with JMS-WebsphereMQ bridge. But actually – why do you care? This is a String id, the specification doesn’t say anything about its structure, except that it has to be unique.

UPDATE: Actually, as mentioned by @skaffman, the ID: prefix is required by JMS specification:

3.4.3. JMSMessageID

All JMSMessageID values must start with the prefix ‘ID:’. Uniqueness of
message ID values across different providers is not required.

This means that WebsphereMQ bridge is adding ID: prefix only to be compliant with JMS spec. And, as also noted by @skaffman, ActiveMQ is not compliant with it…