Starting and stopping a cluster
You can start or stop all of the servers in a cluster (cluster members) in one action. When you start a cluster you automatically enable workload management.
Before you begin
- Prerequisites for starting a cluster
- Ensure that the node agents are running.
- Verify that all resources required by applications deployed to the cluster are available.
- Start all prerequisite subsystems.
- Prerequisites for stopping a cluster
- Make sure there is no work in progress; performance monitoring infrastructure counters can indicate whether all queued work is complete.
- Prevent new work from starting by disabling HTTP and IIOP traffic on the cluster members and quiescing the service integration buses.
About this task
If you use a deployment environment
pattern of Remote Messaging or Remote Messaging and Remote
Support, there can be multiple clusters that depend on one another.
If such a case exists, start or stop the infrastructure and the clusters
as follows to avoid potential problems:
- Infrastructure startup sequence:
- Database, Lightweight Directory Access Protocol (LDAP), and web servers
- Deployment manager (if needed)
- Node agents
- Cluster startup sequence:
- Messaging infrastructure cluster
- Application deployment cluster
- Support cluster
- Cluster shutdown sequence:
- Support cluster
- Application deployment cluster
- Messaging infrastructure cluster
Tip: When you are using clusters, the Initial
State property of the Application Server subcomponent ( ) is not intended
to be used to control the state of individual servers in the cluster
at the time the cluster is started or stopped. It is intended only
as a way to control the state of the Application Server subcomponent
of a server. It is best to start and stop the individual members of
a cluster using the Server options of the administrative console or
commands (startServer and stopServer).
To start or stop a cluster, perform the following steps.