| Author |
Message
|
| charanR |
 Posted: Wed May 15, 2019 6:54 am Post subject: hub & spoke model using mq clustering documentation |
 |
|
Apprentice
Joined: 05 Oct 2017
Posts: 28
|
|
| Back to top |
|
 |
| Vitor |
| Posted: Wed May 15, 2019 7:12 am Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| charanR wrote: |
I have a question in step 6.
I'm creating CL1.QM1 and CL2.QM1 on gateway qmgr as cluster-receiver but the documentation doesn't say anything about creating a cluster-sender. Is this still true ? |
Apart from these cluster senders you mean?
| Code: |
*... On QM1
DEFINE CHANNEL(CL1.QM2) CHLTYPE(CLUSSDR) CONNAME('localhost(1412)') CLUSTER(CL1) REPLACE
....
DEFINE CHANNEL(CL2.QM3) CHLTYPE(CLUSSDR) CONNAME('localhost(1413)') CLUSTER(CL2) REPLACE
....
|
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| charanR |
| Posted: Wed May 15, 2019 7:34 am Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Apprentice
Joined: 05 Oct 2017
Posts: 28
|
| Vitor wrote: |
| charanR wrote: |
I have a question in step 6.
I'm creating CL1.QM1 and CL2.QM1 on gateway qmgr as cluster-receiver but the documentation doesn't say anything about creating a cluster-sender. Is this still true ? |
Apart from these cluster senders you mean?
| Code: |
*... On QM1
DEFINE CHANNEL(CL1.QM2) CHLTYPE(CLUSSDR) CONNAME('localhost(1412)') CLUSTER(CL1) REPLACE
....
DEFINE CHANNEL(CL2.QM3) CHLTYPE(CLUSSDR) CONNAME('localhost(1413)') CLUSTER(CL2) REPLACE
....
|
|
yes. I should create CL1.QM1 and CL2.QM1 as cluster-senders somewhere correct ? |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed May 15, 2019 7:50 am Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| charanR wrote: |
| yes. I should create CL1.QM1 and CL2.QM1 as cluster-senders somewhere correct ? |
No.
Members of a cluster have one cluster sender channel that points to a full repository. That's the point of a cluster.
Try reading the rest of the cluster documentation rather than just the list of instructions for what you think you need to do.
(and I'd not describe that link as a hub and spoke)
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| charanR |
| Posted: Wed May 15, 2019 8:30 am Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Apprentice
Joined: 05 Oct 2017
Posts: 28
|
| Vitor wrote: |
| charanR wrote: |
| yes. I should create CL1.QM1 and CL2.QM1 as cluster-senders somewhere correct ? |
(and I'd not describe that link as a hub and spoke) |
If this is not the link for a hub and spoke design, do you have a recommended documentation ? |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed May 15, 2019 8:59 am Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| charanR wrote: |
| If this is not the link for a hub and spoke design, do you have a recommended documentation ? |
Pretty much all of the rest of the documentation.
A hub and spoke is a series of queue managers in a point to point configuration. So you designate one queue manager as a hub, then work your way through all the spoke queue managers linking them to the hub.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed May 15, 2019 9:02 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
This is probably a good point to emphasize once again that an MQ cluster is not a hub and spoke configuration but a collective of equals with 2 queue managers nominated as record keepers. Messages pass directly between all members of this collective and do not dogleg through the repositories.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| bruce2359 |
| Posted: Wed May 15, 2019 9:54 am Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9469
Location: US: west coast, almost. Otherwise, enroute.
|
| Vitor wrote: |
| Messages pass directly between all members of this collective and do not dogleg through the repositories. |
"all" is a bit confusing. Messages flow point-to-point from the single qmgr where the msg was created to the single qmgr that owns the destination queue.
_________________
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: Wed May 15, 2019 10:13 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| bruce2359 wrote: |
| Vitor wrote: |
| Messages pass directly between all members of this collective and do not dogleg through the repositories. |
"all" is a bit confusing. Messages flow point-to-point from the single qmgr where the msg was created to the single qmgr that owns the destination queue. |
You've been possessed by the spirit of my old English tutor. Douse your body with holy water, get some sage burning and order some crystals from Amazon. In the event of projectile vomiting call your local church.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Wed May 15, 2019 12:19 pm Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
| Vitor wrote: |
You've been possessed by the spirit of my old English tutor. Douse your body with holy water, get some sage burning and order some crystals from Amazon. In the event of projectile vomiting call your local church. |
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Thu May 16, 2019 3:34 pm Post subject: Re: hub & spoke model using mq clustering documentation |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
Figure 1. Client-server application deployed to hub and spoke architecture using IBM MQ clusters
I wish IBM would remove 'hub and spoke' from that caption. It is the only place it is mentioned on the page, and it is misleading. The correct terminology is 'overlapping clusters', as appears in the page title.
As alluded by a previous post, 'hub and spoke' is a different MQ architecture that uses point-to-point distributed channels, not cluster channels.
_________________
Glenn |
|
| Back to top |
|
|
|
|
