One MWS two IS Clusters

, , ,
1

Folks,

Please help me out with this decision.
Setup →

  1. Two separate IS 7.1.2 clusters, one for real time traffic and one for batch, each with 4 nodes.
  2. Each cluster has its own process engine, doc history, IS internal, XRef database.
  3. Only one Core audit, process audit, analysis, tracker and archive database. These are shared by all 8 nodes in the two clusters.
  4. One broker to which all 8 nodes connect.
  5. MWS is not required to be HA.

The question is about the MWS server. I know that MWS is setup to connect to the process engine, so that our tasks can communicate the task information to the process engine.

As we have two separate clusters of Integration Servers (thus Process Engines), with two different process engine databases, do I need two separate instances of MWS (along with two separate database Instances) to connect to these two PE clusters?

Please feel free to shoot back with any questions you may have.

Thanks,
Vikas.

2

Hi,

You can try defining multiple environment information in MWS. I guess you would need only one.

How about if both the clusters have single PE database??

3

Shahid, Thanks for the reply.

Even if I defined multiple environments, the section (Administrator–>My webMethods–>System Settings) gives me option to connect to only one PE.

As for the second suggestion.
The webMethods install guide mentions
“If you are clustering Process Engines, create a single Process Engine database component
for the cluster to share. If you are not clustering Process Engines, create a separate
Process Engine database component for each Process Engine.”

Now as we have two different clusters, I am assuming our case falls under the later statement.

Honestly, I don’t know what will be the repercussions creating one PE database. Can u throw some light please?

Thanks,
Vikas.