| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
  |
|
| Discounted licensing for multi-instance MB/MQ standby nodes |
« View previous topic :: View next topic » |
| Author |
Message
|
| bsiggers |
 Posted: Wed May 18, 2011 8:46 am Post subject: Discounted licensing for multi-instance MB/MQ standby nodes |
 |
|
Acolyte
Joined: 09 Dec 2010
Posts: 53
Location: Vancouver, BC
|
I happened to see an announcement yesterday in the MQ twitter feed from IBM regarding discounted licensing for the standby nodes for MB7/MQ7 for the multi-instance queue manager/broker features, and thought this might be of interest for some of the people on the forum.
http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?subtype=ca&infotype=an&appname=iSource&supplier=897&letternum=ENUS211-232
Would like to know your thoughts on this - we've had these discussions before re: active/active, active/passive solutions - does this change anyone's mind for broker/MQ solution design?
In my opinion, officially charging full price originally for something that is effectively 90% cold-standby was ridiculous to begin with - but that's just me. |
|
| Back to top |
|
 |
| lancelotlinc |
| Posted: Wed May 18, 2011 8:59 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
With PVU licensing and a smart architect, your TCO is far less using active-active-active with site dispersion and lots of other benefits, rather than an active-passive setup.
With active-passive, you never know if the passive will kick in, and if it will work. No one ever tests disaster preparedness like they should. They sometimes simulate testing, but that is not effective.
With active-active-active, you know all sites are working because they are. If one goes offline, the other two pick up the slack. If system is properly designed to be loosely coupled, there is little impact to SLAs.
TCO is less because buying three RHEL PVU licenses is far less than muscle PVU licenses. And the scalability is infinite.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| mqjeff |
| Posted: Wed May 18, 2011 9:24 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
You still want contingency and rapid recovery in an n-active scenario, so that you don't suffer from failures caused by cascading increases in load.
So you still either have to plan for all n nodes to handle the full load if the other n-1 nodes fail or have plans to recover the rest of the nodes quickly in cases of failure.
So you might still want to make each of your n-nodes be an MI configuration.
Regardless, this still saves money for anyone doing any kind of MI configuration, so it is a good thing martha. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Wed May 18, 2011 9:29 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| mqjeff wrote: |
| it is a good thing martha. |
Hey,, I resemble that remark.
My boss in the USAF used to say : "Read the book Margo~!" RTFM
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| rekarm01 |
| Posted: Sun May 22, 2011 1:42 pm Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 1415
|
| lancelotlinc wrote: |
| With PVU licensing and a smart architect, your TCO is far less using active-active-active with site dispersion and lots of other benefits, rather than an Active-passive setup. |
It's not always either-or. Active-Active is not so good for handling orphaned messages, or preserving message order; (for better or worse, some sites insist on message affinity, no matter what the cost).
Sites can have more complex setups using both Active-Active and Active-passive, such as redundant Active-passive pairs (A1-p1, A2-p2, etc.), or mutually Active-passive pairs where each Active server can also act as a passive server for the other.
DR is a separate matter. Some sites may choose to allocate more hardware resources or bandwidth for their primary sites, than for their DR sites. Even if the setup is identical, the geographic remoteness may introduce undesirable network latencies or other traffic, making DR sites less than ideal for day-to-day operations, but more than adequate for disaster recovery. |
|
| Back to top |
|
|
|
|
|
| |
|
Page 1 of 1 |
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
