| Author |
Message
|
| Anant.v |
 Posted: Wed Nov 26, 2014 9:22 pm Post subject: Using RHCS for high availability of MQ (v8) |
 |
|
Apprentice
Joined: 26 Nov 2014
Posts: 40
Location: Malaysia
|
Hello All,
First of all, this is the best forum to look for solutions when you run out of knowledge. I am following this forum from the last 1 year or so.
However, this is my first post (As i have always found my questions answered by someone else's post  )
I am trying to configure MQ on RHCS for high availabilty, I went through a lot of docs for this and found that we have to use a "script" which can perform a start/stop/status for the queue manager and take necessary actions to fail it over and start it up on the second node in a HA environment.
My queston is, have any one of you able to confgure it by adding a queue manager as a "resource" to the cluster ? i.e. without using a script to monitor, rather the cluster itself takes care of the queue managers (As is done by MicroSoft Cluster Services)
Any links for this topic are appreciated..
Expecting a reply ! Thanks for the help.. |
|
| Back to top |
|
 |
| fjb_saper |
| Posted: Thu Nov 27, 2014 5:03 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
you will probably need a script pre/and post failover to shut down and start MQ on the node. 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| Anant.v |
| Posted: Thu Nov 27, 2014 9:07 am Post subject: |
|
|
Apprentice
Joined: 26 Nov 2014
Posts: 40
Location: Malaysia
|
Thanks for the quick response !
I have a script , but the script is becoming a single point of failure..
To see the cluster stability, i bugged the script.. which caused the cluster services to fail and eventually i ended up loosing my shared disks !!
Is there any way we can prevent this ? (As in, not making the script a single point of failure). I am just curious to know as we are still in testing phase. |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Thu Nov 27, 2014 11:23 pm Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
Why would the script be a single point of failure? My understanding is that the script is either copied to each node or resides on a shared disk. Your HA is only as good as your HA clustering software. If it fails you up the creek without a paddle...
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| Anant.v |
| Posted: Fri Nov 28, 2014 3:53 am Post subject: |
|
|
Apprentice
Joined: 26 Nov 2014
Posts: 40
Location: Malaysia
|
Yeabsolutely corerct, its copied to both nodes.. but my question was, shouldn't the RHCS cluster be smart enough to just bring down the the queue manager and services, instead of loosing disks ? (which happened in my case)
Anyways, really thanks for your suggestion, i just wanted an expert review (which i got from you) , and parallely we have opened a PMR to consult with IBM. |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Fri Nov 28, 2014 6:26 am Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
|
| Back to top |
|
|
| Anant.v |
| Posted: Fri Nov 28, 2014 10:26 pm Post subject: |
|
|
Apprentice
Joined: 26 Nov 2014
Posts: 40
Location: Malaysia
|
well, thanks Pete,
I have already looked into that. and have a script ready..
i was just looking at doing this thingy without a script.. I guess, after all this discussion, script is the only way to get things working..
Thanks Peter and Saper, for the help ! |
|
| Back to top |
|
|
|
|
