Author |
Message
|
George Carey |
Posted: Fri May 27, 2011 7:47 am Post subject: IBM Support agrees |
|
|
Knight
Joined: 29 Jan 2007 Posts: 500 Location: DC
|
IBM support agrees and says the same thing.
That makes it official in my book !
Q.E.D.
GTC _________________ "Truth is ... grasping the virtually unconditioned",
Bernard F. Lonergan S.J.
(from book titled "Insight" subtitled "A Study of Human Understanding") |
|
Back to top |
|
 |
PeterPotkay |
Posted: Fri May 27, 2011 8:28 am Post subject: Re: IBM Support agrees |
|
|
 Poobah
Joined: 15 May 2001 Posts: 7723
|
George Carey wrote: |
IBM support agrees and says the same thing |
Agrees with what? _________________ Peter Potkay
Keep Calm and MQ On |
|
Back to top |
|
 |
mqjeff |
Posted: Fri May 27, 2011 9:08 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008 Posts: 17447
|
Bruce - you're still reading more into that statement than exists. It talks about two unique unrelated situations: clients and multi-instance qmgrs. it does not say anything about using SDR channels to non-MI qmgrs.
George - please do clarify exactly what IBM Support has said. |
|
Back to top |
|
 |
ramires |
Posted: Fri May 27, 2011 9:15 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
This is really an interesting discussion! IBM agrees with what? _________________ Obrigado / Thanks you |
|
Back to top |
|
 |
PeterPotkay |
Posted: Fri May 27, 2011 9:48 am Post subject: |
|
|
 Poobah
Joined: 15 May 2001 Posts: 7723
|
ramires wrote: |
but... listening to all the session , they contradict themselves at 35:51 |
Agree! At 36:45, Jonathan Rumsey starts speaking and specifically and explicitly states that the multiple IPs addresses on a SNDR channel's conname needs to point at the same M.I. instance QM - it cannot be the same RCVR channel name on different QMs.
If Level 2 contradicts this in a PMR, I would push back and have them check with Level 3.
I think Question 10 was really asking "If the SENDING queue manager is 7.0.1 and is not a multi instance QM, can you use multiple IPs to aim at a RECEIVING queue manager that is a Multi Instance QM."
Just my 2 cents, I don't know the answer 100% either way. _________________ Peter Potkay
Keep Calm and MQ On |
|
Back to top |
|
 |
bruce2359 |
Posted: Fri May 27, 2011 10:55 am Post subject: Re: cluster over Wan or not |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9482 Location: US: west coast, almost. Otherwise, enroute.
|
George Carey wrote: |
An MQ best practice advisory note some time ago spoke to the wisdon of using sdr/receiver channels over WANs to connect two separately administered entities and not attempt to use cluster channels across WANs ... |
I recall a similar recommendation (admonition, warning) way, way back at the v2.1 - early v5 time-frame. IMS this was mostly due to unreliable WAN networks. While networks still give us grief, they are far more robust today than a decade ago.
The WMQ Queue Manager Clusters manual gives examples of WAN clusters (NEWYORK, PARIS, LONDON). I have to believe that this was no accident (false advertising) on IBMs part.
The answer offered at 36:45 sounded pretty explicit about the intent of multiple connames in SDR channel defs and MI. Clusters were recommended as the alternative. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
George Carey |
Posted: Fri May 27, 2011 11:09 am Post subject: more inclusive |
|
|
Knight
Joined: 29 Jan 2007 Posts: 500 Location: DC
|
Quote: |
"...If Level 2 contradicts this in a PMR, I would push back and have them check with Level 3. ... " |
So you would push back on a more inclusive supported interpretation of the usage of connection lists ?
That is just exactly what I don't want to happen ... Maybe if it is a benefit to the user community you should think about assuring that it is a supported function ...
thanks for the help guys ... but I am done on this post!! _________________ "Truth is ... grasping the virtually unconditioned",
Bernard F. Lonergan S.J.
(from book titled "Insight" subtitled "A Study of Human Understanding") |
|
Back to top |
|
 |
mqjeff |
Posted: Fri May 27, 2011 11:14 am Post subject: Re: more inclusive |
|
|
Grand Master
Joined: 25 Jun 2008 Posts: 17447
|
George Carey wrote: |
but I am done on this post!! |
Hi George - Can you please clarify what IBM support has said? |
|
Back to top |
|
 |
George Carey |
Posted: Fri May 27, 2011 11:24 am Post subject: Not really |
|
|
Knight
Joined: 29 Jan 2007 Posts: 500 Location: DC
|
Not really ... if people are planning on pushing back on it ...
All I will say is ... paraphrasing, the IBM Support interpretation from the same passage in the Infocenter on CONNAME was the same as what Bruce previously concluded that he saw no reason to exclude a sdr/receiver non-MI Qmgr set up as not being supported.
I am seeking an affirmative statement from level 3. _________________ "Truth is ... grasping the virtually unconditioned",
Bernard F. Lonergan S.J.
(from book titled "Insight" subtitled "A Study of Human Understanding") |
|
Back to top |
|
 |
mqjeff |
Posted: Fri May 27, 2011 11:28 am Post subject: Re: Not really |
|
|
Grand Master
Joined: 25 Jun 2008 Posts: 17447
|
George Carey wrote: |
Not really ... if people are planning on pushing back on it ... |
I believe that Peter was suggesting that you push back on L2 if they contradicted the statement from Jonathan Rumsey, not that we would push back on you. |
|
Back to top |
|
 |
ramires |
Posted: Fri May 27, 2011 1:10 pm Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
I believe this is not supported solution. If using persistent messages, and for some reason the sender channel goes indoubt, in the next sender channel restart it needs to resolve that, with the same queue manager.
For non persistent messages it works, but still not supported.
When I find the time I'll try to recreate this, with persistent messages and forcing an indoubt state to see what happens.
I'll post what I found... but no target date
Thanks to all the participants in this thread! |
|
Back to top |
|
 |
George Carey |
Posted: Mon Jun 20, 2011 8:45 am Post subject: Requirement |
|
|
Knight
Joined: 29 Jan 2007 Posts: 500 Location: DC
|
I was able to generate the in-doubt condition on channel fail-over ...
I used the MQ Explorer to resolve the condition and it asked that I manually resolve the in-doubt by either commiting or rolling back last batch and also gave appropriate messages about how messages could be duplicated or lost depending on the resolution choice.
No problem ... I make my choice and in-doubt resolved and everthing continues fine.
I decide there is a nice way to handle this in-doubt situation so that an a-priori choice can be setup with a couple new queue attributes that would by-pass the need to manually resolve the in-doubt.
IBM gave positive feedback on my suggestion and asked that I submit it as a new Requirement ... I forget the process or where a form or url is to do this and can't locate via seaching after a few attempts... thought some one here might quickly know where/how to do this.
GTC _________________ "Truth is ... grasping the virtually unconditioned",
Bernard F. Lonergan S.J.
(from book titled "Insight" subtitled "A Study of Human Understanding") |
|
Back to top |
|
 |
mqjeff |
Posted: Mon Jun 20, 2011 8:49 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008 Posts: 17447
|
|
Back to top |
|
 |
George Carey |
Posted: Mon Jun 20, 2011 9:17 am Post subject: download problem. |
|
|
Knight
Joined: 29 Jan 2007 Posts: 500 Location: DC
|
Thanks .. is this latest/current site for this?
I get an error when trying to view or download the ERCF document.
(the file could not be written to cache)
I will see if some security check is not allowing it as well. _________________ "Truth is ... grasping the virtually unconditioned",
Bernard F. Lonergan S.J.
(from book titled "Insight" subtitled "A Study of Human Understanding") |
|
Back to top |
|
 |
bruce2359 |
Posted: Mon Jun 20, 2011 10:41 am Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9482 Location: US: west coast, almost. Otherwise, enroute.
|
It's a Word .doc document. It downloaded just fine for me. _________________ I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
Back to top |
|
 |
|