| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Relationship between expiry and timeout_interval? |
« View previous topic :: View next topic » |
| Author |
Message
|
| Sandman |
 Posted: Mon Jul 14, 2003 8:51 am Post subject: Relationship between expiry and timeout_interval? |
 |
|
Centurion
Joined: 16 Oct 2001
Posts: 134
Location: Lincoln, RI
|
For example, if you have a GUI that asks its server to send/receive a message, the server will stamp the outgoing request message w/ an expiry value, then issue a GET/WAIT w/ a timeout value for the reply.
What value do you set lower and why?
Should the message expire before the GET times out - essentially "guaranteeing" that when the user is notified that "you ain't getting no steenkin' reply", it's actually true that the server won't get one either (and possibly perform some file updates of which the requester will not be made aware)?
Or should the GET timeout before the message expires?
Thank you. |
|
| Back to top |
|
 |
| kirani |
| Posted: Mon Jul 14, 2003 10:32 am Post subject: |
|
|
Jedi Knight
Joined: 05 Sep 2001
Posts: 3779
Location: Torrance, CA, USA
|
Without knowing the details of your transaction and system, I'd speak in general. I'd set the GET timeout value to smaller value (e.g. 30 seconds), and expiry to greater value (e.g. 40 seconds). This is to make sure the messages don't expire before they are read from the queue.
Client (GUI) TimeOut = Time required to transmit message to the server + Time taken by server program to process the request + Time required to transmit the reply message back to the client + some buffer time.
You might want to consider following points when designing such applications,
1. The type of transaction. Inquiry transactions can have smaller expiry time set.
2. Should the server program set new expiry time on reply messages or copy the time from input request message.
3. Set expiry to unlimited, If it's a critical update transaction and you don't want to loose the reply message. This will also require designing for late arrivals.
Others might have different ideas.
_________________
Kiran
IBM Cert. Solution Designer & System Administrator - WBIMB V5
IBM Cert. Solutions Expert - WMQI
IBM Cert. Specialist - WMQI, MQSeries
IBM Cert. Developer - MQSeries
Last edited by kirani on Mon Jul 14, 2003 11:37 am; edited 1 time in total |
|
| Back to top |
|
|
| Sandman |
| Posted: Mon Jul 14, 2003 11:00 am Post subject: |
|
|
Centurion
Joined: 16 Oct 2001
Posts: 134
Location: Lincoln, RI
|
Thanks for the response, kirani.
So you're basically saying, "The timeout should be > the expiry to account for processing before and after the message processing itself", right?
Also, as a rule, our inquiry servers simply copy (what's left of) expiry onto replies they send. We only set unlimited expiry (and persistence=yes) on update requests (and then servers simply copy such attributes into replies).
Thanks again,
Sandman |
|
| Back to top |
|
|
| kirani |
| Posted: Mon Jul 14, 2003 11:39 am Post subject: |
|
|
Jedi Knight
Joined: 05 Sep 2001
Posts: 3779
Location: Torrance, CA, USA
|
| Sandman wrote: |
Thanks for the response, kirani.
So you're basically saying, "The timeout should be > the expiry to account for processing before and after the message processing itself", right?
|
No! Sorry, my mistake! I meant to say the other way. I corrected my earlier post.
_________________
Kiran
IBM Cert. Solution Designer & System Administrator - WBIMB V5
IBM Cert. Solutions Expert - WMQI
IBM Cert. Specialist - WMQI, MQSeries
IBM Cert. Developer - MQSeries
|
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
