454 WebSphere Application Server V6.1 Problem Determination IBM Redpaper Collection
Figure 17-5 Messaging engine data store schema name
7. Repeat this procedure for each of the messaging engines, making sure all
schema names are unique.
8. Save the configuration changes and restart the cluster.
17.4 Cluster runtime problems
This section discusses problems specific to clustering of messaging engines
where the messaging engines start correctly but there is a problem with message
delivery.
17.4.1 Core group policy is incorrect or no preferred server defined
To exert some control over the location of the messaging engine, you can define
a preferred server in the core group policy. The messaging engine will start on
the preferred server if the cluster member is active.
If a core group policy is defined incorrectly with invalid or non-matching criteria,
the WLM manager randomly assigns the messaging engines to cluster
members, resulting in the messaging engines being started on random cluster
members.
Defining a preferred server
First, define a core group policy that can be associated with a messaging engine.
The default messaging provider includes a DefaultCoreGroup that can be
extended for cluster configurations.
Chapter 17. Clustering problem determination 455
To create a core group policy using the administrative console:
1. Select Servers Core Groups Core group settings.
2. Click on the core group name.
3. In the Configuration tab, select Policies.
4. Click New to define a new policy (Figure 17-6 on page 456):
a. Select One of N as the policy type.
The One of N policy keeps one member of the high availability group
active at all times. It is used by groups that desire singleton failover. In an
HA configuration, there is only one messaging engine.
b. Enter a name for the policy.
c. If you want failback to be active (that is, the messaging engine fails back
when the preferred server comes online after a failover has taken place),
select the failback box, as shown in Figure 17-6 on page 456.
456 WebSphere Application Server V6.1 Problem Determination IBM Redpaper Collection
Figure 17-6 Core group policy for high availability and failback
Next, you define the match criteria that associates a messaging engine with the
policy.
Using the administrative console:
1. Select Servers Core groups Core group settings.
2. Click on the core group.
3. In the configuration tab, select Policies.
4. Click on the policy name.
5. Select Match Criteria.
6. Click New.
Chapter 17. Clustering problem determination 457
7. Enter the name/value pair information. The Name and Value fields together
specify a messaging engine or group of messaging engines, it will be linked to
this policy. Name/value pairs that you can use shown in Table 17-1.
Table 17-1 Match criteria name/value pairs
In this instance, you need to use the WSAF_SIB_MESSAGING_ENGINE
name/value pair, as shown in Figure 17-7.
Figure 17-7 Name / value pair
Finally, specify the preferred server for the policy:
1. Select Servers Core groups Core group settings.
Name Value Messaging engines the
policy will match
type WSAF_SIB Any messaging engine
WSAF_SIB_MESSAGING
_ENGINE
The name of your
messaging engine.
A particular messaging
engine
WSAF_SIB_BUS The name of your bus All messaging engines in a
particular bus
IBM_hc The name of your cluster All messaging engines in a
particular cluster

Get WebSphere Application Server V6.1 Problem Determination: IBM Redpaper Collection now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.