CDD Configuration Procedures : Configuring a Cache OM Cluster — Cluster Tab

Configuring a Cache OM Cluster — Cluster Tab
Use this procedure if you picked Cache OM in the second page of the New Cluster Configuration wizard.
For the wizard procedure, see Adding a Cluster Deployment Descriptor.
To Configure the Cluster Tab for Cache OM
See Chapter 4, Cluster Deployment Descriptor Reference for reference tables mentioned in this section.
1.
In the Cluster tab click General. On the right, specify the following:
See Cluster Tab — General Settings for details.
2.
A tree of option groups appears on the left. Click an option to display an appropriate configuration panel on the right.
3.
In the panel on the left, click Properties. Expand to see the provided coherence group of properties. Add properties as needed and provide values as explained next.
Backup Count
4.
In the following property, define the number of members of the distributed cache service (that is, cache agents) that hold the backup data for each unit of storage in the cache.
   tangosol.coherence.distributed.backupcount
See Backup Count for more details.
Cluster Discovery
5.
You can use either multicast or well-known-address (WKA) discovery, as appropriate. See Cache Cluster Discovery in TIBCO BusinessEvents Architect’s Guide for guidelines about which option is right for your use. Multicast is more commonly used and that is what’s explained in this procedure. To see instructions for well-known address configuration see Configuring Well-Known Address Cluster Member Discovery. Then continue at step 7 after you have finished the WKA configuration.
To Use Multicast Cluster Member Discovery  Multicast is the default discovery method, and defaults for multicast discovery are appropriate for many environments. Specify values if defaults for the following properties are not appropriate in your environment:
   tangosol.coherence.clusteraddress
   tangosol.coherence.clusterport
   tangosol.coherence.ttl
Specifying one or more well-known addresses disables all multicast communication. Remove any well-known address properties, if any.
See Multicast Discovery Properties for details
6.
   tangosol.coherence.localhost
   tangosol.coherence.localport
Other Properties
7.
   tangosol.coherence.override
   tangosol.coherence.cacheconfig
See Other Coherence Properties for details on the above.
Backing Store
8.
Click Backing Store in the tree on the left and complete the fields on the right.
Click Backing Store > Connection and complete the fields on the right to reference the JDBC connection for the backing store, and the pool settings.
See Cluster Tab — Cache OM — Backing Store Settings for details.
Domain Objects
9.
Click Domain Objects > Default and specify the default object management settings for domain objects (entities). See Table 8, Cluster Tab — Cache OM — Domain Object Default Settings for settings.
Also specify the preload fetch size, to preload a subset of the objects into the cache or object table. The setting applies to all preloaded objects and object handles.
10.
a.
In the tree on the left, click Overrides and click Add.
b.
Select the /uri entry that appears on the left, and configure the settings on the right. You can use the defaults set in step 9, or override them
You can also edit existing override entries, and remove entries not needed (by clicking Remove).
11.
Database Concepts
12.
If you use database concepts, available in the TIBCO BusinessEvents Data Modeling add-on product, select Database Concepts on the left and complete the fields in the configuration panel on the right. See TIBCO BusinessEvents Data Modeling Developer’s Guide for details.
Go to Configuring the Collections Tab (All OM Types) to continue configuration.
Configuring Well-Known Address Cluster Member Discovery
If multicast cluster member discovery will not be used in your environment, configure well-known address discovery instead. At least one machine configured as a WKA machine must be running in order for other machines to join the cluster.
To understand how well-known address discovery works see Cluster Member Discovery Using Well-Known-Addresses in TIBCO BusinessEvents Architect’s Guide.
To Configure Well-Known Address Cluster Member Discovery
This is a sub-procedure of To Configure the Cluster Tab for Cache OM.
To configure well-known address discovery, you must configure machine-specific settings at the cluster level, and at the processing unit level. Machines where WKA PUs are to be deployed are called WKA machines in this procedure.
One processing unit deployed to a WKA machine must have the additional WKA configuration. Additional processing units can be deployed to a WKA machine, configured in the usual way, and they will discover and join the cluster in the usual way at runtime.
For details about the properties, see Well-Known Address Properties.
1.
2.
Open the cluster tab Properties sheet and add a pair of WKA properties for each machine you want to configure as a well-known address machine:
   tangosol.coherence.wkan HostIP
   tangosol.coherence.wkan.port Hostport
For example at the cluster level you might have these two WKA machines:
Property: tangosol.coherence.wka1   Value: 10.97.118.151
Property: tangosol.coherence.wka1.port   Value: 8098
Property: tangosol.coherence.wka2   Value: 10.97.118.152
Property: tangosol.coherence.wka2.port   Value: 8098
3.
Open the Processing Units tab properties sheet for a processing unit (PU). Configure one set of WKA properties to match a cluster level set of WKA properties:
   tangosol.coherence.localhost HostIP
   tangosol.coherence.localport Hostport
   be.engine.hostaddress HostIP
and set same the value as in tangosol.coherence.localhost. See Configuring for WKA Cluster Discovery Configuration for more steps you must take.
(For better organization, you can put these properties into a property group, named as desired.)
Make sure the PU level property values match corresponding values set at the cluster level. That is, set the localhost property to the IP of the host where you will deploy the PU, and set the localport property to the port defined in the cluster properties localport property.
Repeat this step until you have configured one PU with matching PU-level properties for each cluster-level set of WKA properties.
Each of these PUs must be deployed to the machine with the specified IP.
For example, if you configured the two well-known addresses shown in step 2, then at the processing unit level you would configure a processing unit to be deployed on each of those machines as follows:
Property: tangosol.coherence.localhost. Value: 10.97.118.151
Property: tangosol.coherence.localport. Value: 8098
Property: tangosol.coherence.localhost. Value: 10.97.118.152
Property: tangosol.coherence.localport. Value: 8098
At deploy time you must select those processing units to deploy on the appropriate (matching) WKA machine. It can be helpful if the name of the PU contains the machine name or other reminder that this PU must be deployed to a specific machine.
See Localhost and Localport Properties
4.
Continue at step 7 after you have finished the WKA configuration to complete cluster-level configuration.