Table of Contents

Search

  1. Preface
  2. Upgrade Overview
  3. Pre-Upgrade Tasks
  4. Database Tasks
  5. Application Server Tasks
  6. Hub Store Upgrade
  7. Hub Server Upgrade (In-place Upgrade)
  8. Process Server Upgrade (In-place Upgrade)
  9. Resource Kit Upgrade (In-place Upgrade)
  10. Post-Upgrade Tasks
  11. Search Configuration Upgrade
  12. ActiveVOS Post-Installation Tasks for the Application Server
  13. ActiveVOS Post-Upgrade Tasks for Business Entity Adapter
  14. ActiveVOS Post-Upgrade Tasks for Subject Areas Adapter
  15. Appendix A: Troubleshooting the Upgrade Process
  16. Appendix B: Frequently Asked Questions
  17. Appendix C: Processing Existing ActiveVOS Tasks
  18. Appendix D: Configuring Metadata Caching

Upgrading from Version 9.0.1 or 9.1.0 for Oracle

Upgrading from Version 9.0.1 or 9.1.0 for Oracle

Logical Grouping of Java Virtual Machine Example

Logical Grouping of Java Virtual Machine Example

By grouping Java Virtual Machines (JVMs), you get a logical group of Hub Servers and Process Servers. When you deploy the Hub Server and Process Server applications in a logical JVM group, communication between the Hub Server and Process Server applications stay within the group. To group JVMs, you assign a group ID to each JVM in the MDM Hub environment.
Process Server grouping is applicable to the cleanse and match process only. A Process Server that is enabled as a Zookeeper server and is enabled for smart search is available to all the groups for indexing and search processing. The logical groups are not applied to the internal server cache of the MDM Hub.
The following table shows an example of logical JVM groups:
JVM Group
JVM
Hub Server
Process Server
Group1
JVM1
Yes
Yes
Group1
JVM4
-
Yes
Group2
JVM2
Yes
Yes
Group3
JVM3
-
Yes
For JVM1, add the following Java option in the startup script:
-Dmdm.node.groupid=Group1
For JVM2, add the following Java option in the startup script:
-Dmdm.node.groupid=Group2
For JVM3, add the following Java option in the startup script:
-Dmdm.node.groupid=Group3
For JVM4, add the following Java option in the startup script:
-Dmdm.node.groupid=Group1
After you configure the JVMs, and deploy the Hub Servers and Process Servers, the groups have the following characteristics:
  • Group1 has two Process Servers, Group2 has one Process Server, and Group3 has one Process Server.
  • All cleanse and batch calls stay in their own group with the exception of smart search. For example, any real-time call on the Hub Server in Group1 affects only the Group1 Process Servers (JVM1 and JVM4).

0 COMMENTS

We’d like to hear from you!