| Author |
Message
|
| ican.sbyn |
 Posted: Tue Jun 08, 2010 10:05 am Post subject: http input node - Customising the Teimeout Response |
 |
|
Novice
Joined: 14 Jul 2009
Posts: 19
|
In HTTP Input node, there is a property called "Error Handling"... I can set here the "Maximum client wait time(sec)" and the "Fault Format" (SOAP/HTML). Now if the Max client wait time exceeds the threshold, then the http input node (rather the listener) sends back an automatic response to the sender application in soap/http. Is there any way to catch this one and customise it accordingly (in XML format)?
The requirement is, the response for timeout should not be in SOAP/HTTP format...this should be in custom xml format (may be need to use the compute node)... To summerize, in case of timeout, I should have the control to compute the response to the sender application. Any input will be helpful. |
|
| Back to top |
|
 |
| fatherjack |
| Posted: Tue Jun 08, 2010 2:29 pm Post subject: |
|
|
Knight
Joined: 14 Apr 2010
Posts: 522
Location: Craggy Island
|
In the event of a http timeout to which terminal is the response propgated?
_________________
Never let the facts get in the way of a good theory. |
|
| Back to top |
|
|
| ican.sbyn |
| Posted: Thu Jun 10, 2010 11:29 am Post subject: |
|
|
Novice
Joined: 14 Jul 2009
Posts: 19
|
| It doesnt comeout through any terminal, and thats why the challange is. The listener sends the auto reply to the sender application. |
|
| Back to top |
|
|
| fatherjack |
| Posted: Thu Jun 10, 2010 11:41 am Post subject: |
|
|
Knight
Joined: 14 Apr 2010
Posts: 522
Location: Craggy Island
|
Oops.  Misread your post.
_________________
Never let the facts get in the way of a good theory. |
|
| Back to top |
|
|
| mgk |
| Posted: Thu Jun 10, 2010 1:00 pm Post subject: |
|
|
Padawan
Joined: 31 Jul 2003
Posts: 1642
|
If you were to join the WMB beta program (through your IBM Rep), you would be able to try out new features before they are officially announced.
In this case it sounds like you would find a new timeout terminal on the HTTPInput node which would allow you change the timeout message before it was sent back to the client useful, would this be correct?
Kind Regards,
_________________
MGK
The postings I make on this site are my own and don't necessarily represent IBM's positions, strategies or opinions. |
|
| Back to top |
|
|
| fatherjack |
| Posted: Thu Jun 10, 2010 4:25 pm Post subject: |
|
|
Knight
Joined: 14 Apr 2010
Posts: 522
Location: Craggy Island
|
| mgk wrote: |
| In this case it sounds like you would find a new timeout terminal on the HTTPInput node which would allow you change the timeout message before it was sent back to the client useful, would this be correct? |
Interesting if this is a likely enhancement, but in this, the internet, age, aren't most users used to seeing http timeout messages? So why would you want to present them with something different for this particular timeout.
_________________
Never let the facts get in the way of a good theory. |
|
| Back to top |
|
|
| Vitor |
| Posted: Thu Jun 10, 2010 5:03 pm Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| fatherjack wrote: |
| So why would you want to present them with something different for this particular timeout. |
Because not all users are used to seeing timeout messages, and some clients prefer something a bit more "we're sorry, we seem to be having some difficulty".
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Thu Jun 10, 2010 9:28 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
I agree with Vitor.
In this day and age how many users know what the HTTP error codes mean?
By all means send them back but don't just display them to the user 'As-IS'.
How can they understand the differences between a 404, 403, 405 & a 200 series error?
Displaying these numbers is so 1990's. Something like...
"I'm sorry I haven't a clue as to what is going on" 
Would certainly get a smile from some users in the UK
(hint Radio Comedy series)
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| fatherjack |
| Posted: Thu Jun 10, 2010 11:06 pm Post subject: |
|
|
Knight
Joined: 14 Apr 2010
Posts: 522
Location: Craggy Island
|
| smdavies99 wrote: |
| How can they understand the differences between a 404, 403, 405 & a 200 series error? |
They don't need to. That's the techies job when it gets reported to the helpdesk.
| smdavies99 wrote: |
Displaying these numbers is so 1990's. Something like...
"I'm sorry I haven't a clue as to what is going on" |
It absolutely isn't. It's actually saying "I know exactly what's going on and here is the error". Taking the real error message and converting it into something like "we're sorry, we seem to be having some difficulty" as Vitor suggests surely make problem diagnosis that much more difficult.
_________________
Never let the facts get in the way of a good theory. |
|
| Back to top |
|
|
| SANTYP |
| Posted: Thu Jun 10, 2010 11:56 pm Post subject: Re: http input node - Customising the Teimeout Response |
|
|
Centurion
Joined: 27 Mar 2007
Posts: 142
|
| ican.sbyn wrote: |
In HTTP Input node, there is a property called "Error Handling"... I can set here the "Maximum client wait time(sec)" and the "Fault Format" (SOAP/HTML). Now if the Max client wait time exceeds the threshold, then the http input node (rather the listener) sends back an automatic response to the sender application in soap/http. Is there any way to catch this one and customise it accordingly (in XML format)?
The requirement is, the response for timeout should not be in SOAP/HTTP format...this should be in custom xml format (may be need to use the compute node)... To summerize, in case of timeout, I should have the control to compute the response to the sender application. Any input will be helpful. |
I have a question here ..
when timout occured error message will be propagated through failure terminal.. I think if u coneect a computenode to that terminal u can handle custom message..
suggest me If I am wrong |
|
| Back to top |
|
|
| vmcgloin |
| Posted: Fri Jun 11, 2010 12:08 am Post subject: |
|
|
Knight
Joined: 04 Apr 2002
Posts: 560
Location: Scotland
|
Santyp - I do not think that is the case.
And in the absence of a timeout terminal - which I would definitely like to see by the way - I suggest you set the timeout to be something higher than the client requires and enforce the client timeout in your flow processing i.e. using an MQGet node so that you can control the timeout error message.
Absolutely not ideal design and usually I'm all for removing MQGet nodes, but if this really is a requirement, then this a way you can do it.
Cheers,
Vicky |
|
| Back to top |
|
|
| Esa |
| Posted: Fri Jun 11, 2010 12:28 am Post subject: |
|
|
Grand Master
Joined: 22 May 2008
Posts: 1387
Location: Finland
|
| fatherjack wrote: |
It absolutely isn't. It's actually saying "I know exactly what's going on and here is the error". Taking the real error message and converting it into something like "we're sorry, we seem to be having some difficulty" as Vitor suggests surely make problem diagnosis that much more difficult. |
I don't think it is wise to let clients diagnose your services internal problems. If you return the real error messages you are exposing your implementation details and making it vulnerable. |
|
| Back to top |
|
|
| Esa |
| Posted: Fri Jun 11, 2010 12:50 am Post subject: |
|
|
Grand Master
Joined: 22 May 2008
Posts: 1387
Location: Finland
|
But with the topic at hand -http timeouts - you will have to be quite skilled to be able to expose any details of your endpoint implementation.
With Message Broker the client is seldom a person using a web browser but some sort of software. So I would prefer the current implementation and let the http server send the number as a reply. If you know that your service is consumed by people instead of machines you can take the approach that Vicky described. But the client is software and you replace the standard http timeout reply with a custom message you are actually forcing the client to process the reply on a lever higher that the protocol. Which is in most cases completely unnecessary? |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Jun 11, 2010 6:48 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| smdavies99 wrote: |
"I'm sorry I haven't a clue as to what is going on"
Would certainly get a smile from some users in the UK
(hint Radio Comedy series) |
 
We still mourn Chairman Humph. 
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mqjeff |
| Posted: Fri Jun 11, 2010 11:26 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| Vitor wrote: |
| We still mourn Chairman Humph. |
Are you still allowed to use "we" when talking about Brits?
I suppose you can take the Scotsman out of Scotland.... but never try to take the Scotch out of the Scotsman... |
|
| Back to top |
|
|
|
|
