This is an archive of the Maximo Yahoo Community. The content of this pages may be a sometimes obsolete so please check post dates.
Thanks to the community owner Christopher Wanko for providing the content.
Howdy Maximo users,
Would you suggest clustering for an environment something like ours ?
Would you also share your experience with clustered environment.
----------------------------------
----------------------------------
We are currently in the process of upgrading from Maximo 6 to Maximo 7.6,
and we are evaluating the pros/cons of Clustering. We have a clustered
environment, and wondering whether we really need one.
>We heavily customized Maximo using Java / Automation scripts / Application
Designer.
>All our crafts use Datasplice, who uses Maximo. And they sync with Maximo
twice a day, during beginning of their shift and later when their shift is
ended.
Our Maximo Statistics:
Active logged in users at any time - 25 to 50
Active Workorders - 9634
Active Assets - 38542
Active PM's - 3614
Active JP's - 4554
Active Escalations - 10
MIF - 20 requests per day
Number of reports executed per day - 10
Server configuration:
Webpshere 8.5.5 with 8GB Ram and 2 CPU
The argument for clustering in your scenario would be to lighten occasional loads on a single JVM. Consider a situation when you have your maximum online users, while many mobile users are synchronizing, and a few people are running a PM WOgen or some crazy ABC report. A single JVM may fall over, but two JVMs have enough headspace to manage that situation (better, not perfectly).
In my experience, having that extra JVM always helps, and never hurts. Your box is a little small -- RAM is cheap so double it if you can -- but it should run fine with two JVMs.
The argument against? Additional configuration complexity. You will need to deploy the EAR twice -- although another reason for clustering is you can split MIF and reporting between two members of the cluster -- and you need to watch ripple starting. It has always worked for me, but it is an extra point of failure.
-C
+ 1 to clustering
We recently upgraded from Maximo 7.1 to Maximo 7.5 and included clustering as part of our upgrade process. I can see a huge difference in the performance during peak times when a lot transactions normally occur. The setup is a little more complicated, but that's just a one time issue compared to the volume of reduced calls I've got about Maximo being slow.