Author |
Message
|
wskibum |
Posted: Sat Apr 28, 2012 2:25 pm Post subject: Problems with cluster concept |
|
|
 Apprentice
Joined: 03 Jul 2008 Posts: 38 Location: Northern California
|
Local queue in A clustered
A: >>>>>>>> B:
>
>
>>>>>>>C:
HA
Put to A becomes available for get in B
If B is down becomes available for get in C
I am brand new at clustering, I have searching and reading for days and have run out of time. When I create the cluster it works in reverse.
ANY help would be appreciated |
|
Back to top |
|
 |
exerk |
Posted: Sat Apr 28, 2012 2:51 pm Post subject: Re: Problems with cluster concept |
|
|
 Jedi Council
Joined: 02 Nov 2006 Posts: 6339
|
wskibum wrote: |
...Put to A becomes available for get in B...If B is down becomes available for get in C... |
No. Consider:
QMA - MYQ (cluster queue)
QMB
QMC
A put by an application, connected to QMA, means the message is on MYQ in QMA.
A put by an application, connected to QMB or QMC (or an object that references it), will cause the message to go to MYQ in QMA - the queue managers resolve where the queue is and send the message across the cluster via cluster channels.
If QMA becomes unavailable, e.g. stopped, then any messages on MYQ are unavailable until such time as the queue manager is recovered.
You can only get from queues in a queue manager to which the application is connected directly (bindings mode - cross memory) or to which the application is connected remotely (client mode - network). _________________ It's puzzling, I don't think I've ever seen anything quite like this before...and it's hard to soar like an eagle when you're surrounded by turkeys. |
|
Back to top |
|
 |
wskibum |
Posted: Sat Apr 28, 2012 3:00 pm Post subject: |
|
|
 Apprentice
Joined: 03 Jul 2008 Posts: 38 Location: Northern California
|
So the HA type solution I am trying to put together using clustering is not possible? |
|
Back to top |
|
 |
bruce2359 |
Posted: Sat Apr 28, 2012 3:56 pm Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
|
Back to top |
|
 |
mvic |
Posted: Sat Apr 28, 2012 4:10 pm Post subject: Re: Problems with cluster concept |
|
|
 Jedi
Joined: 09 Mar 2004 Posts: 2080
|
wskibum wrote: |
Local queue in A clustered
A: >>>>>>>> B:
>
>
>>>>>>>C:
HA
Put to A becomes available for get in B
If B is down becomes available for get in C |
Explain that again? |
|
Back to top |
|
 |
bruce2359 |
Posted: Sat Apr 28, 2012 5:09 pm Post subject: |
|
|
 Poobah
Joined: 05 Jan 2008 Posts: 9469 Location: US: west coast, almost. Otherwise, enroute.
|
WMQ message channels are point-to-point. When needed, clustering software will create the appropriate CLUSSDR channel from one qmgr in the cluster to another qmgr in the cluster.
Here's a bumper-sticker to remember: PUT GLOBAL, GET LOCAL.
This means that an application can (may) put a message to any queue anywhere in the network. Conversely, all gets are local (where the consuming application is executing).
If a message is put to a queue on a qmgr, AND that qmgr dies before the message can be consumed, the message is trapped on that qmgrs queue-holding disk until the qmgr becomes available again. _________________ 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 |
|
 |
Vitor |
Posted: Sun Apr 29, 2012 6:28 pm Post subject: |
|
|
 Grand High Poobah
Joined: 11 Nov 2005 Posts: 26093 Location: Texas, USA
|
wskibum wrote: |
So the HA type solution I am trying to put together using clustering is not possible? |
No. WMQ clustering is for workload balancing not HA. It's a shame so many software products use the term "cluster".
Other WMQ attributes (like MI) are used for HA. _________________ Honesty is the best policy.
Insanity is the best defence. |
|
Back to top |
|
 |
|