To configure Extended Deployment features for WebSphere Commerce
enhancements, you must create a service policy, configure health monitoring,
set dynamic clusters to automatic mode, monitor remote servers, and configure
runtime operations.
关于本任务
This example service policy causes all traffic to the Commerce Accelerator
to be treated as highest priority with a service goal of two seconds. Do the
following to create a service policy:
过程
- Launch the Deployment Manager Administration Console.
- Select Operational Policies > Service Policies.
- Click New to begin the Service Policy creation wizard.
- Choose a name, such as WC_High_Priority.
- Choose a Goal Type of Average Response Time.
- Click Next.
- Type a goal value, such as 2 seconds.
- Select Highest Importance.
- Click Next. A transaction class will be created, WC_High_Priority_TC.
- Click Next. Review the new service policy.
- Click Finish.
- Select Applications > Enterprise Applications.
- Select the WC_demo application.
- Select the Service Policies tab.
- Expand Work Classes for HTTP Requests.
- Create a new work class.
- Name it Accelerator High Priority.
- Click Next
- Select the CommerceAccelerator.war.
- Add a wildcard custom URI Pattern (*).
- Click Next.
- Review the work class and select Finish.
- Expand Work Classes for HTTP Requests.
- Expand Accelerator High Priority.
- Select the default transaction class to be WC_High_Priority_TC .
- Click OK and save the changes.
- Do the following to configure health monitoring:
- This example health policy monitors for excessive memory usage. In the
event that excessive memory usage is detected, a task is generated warning
the administrator, with the option of restarting the node with the problem.
- Launch the Deployment Manager Administration Console.
- Select Operational Policies > Health Policies.
- Click New to begin the Health Policy creation wizard.
- Type a name, such as WC_Excessive_Memory.
- Select the health condition Memory Condition: excessive memory usage.
- Click Next.
- Leave the default conditions of 85% JVM heapsize with a 5 minute time
period.
- The default reaction is a supervised scenario in which a runtime task
is generated warning the administrator.
- Click Next.
- Select the member type as Dynamic Cluster.
- Select the dynamic cluster to which the WC_demo application belongs.
- Click Next.
- Review the health policy and click Finish. Save changes.
- Dynamic Clusters allow server resources to be started and stopped
as demand increases and decreases. If there are more applications than just
WebSphere Commerce installed to the cell, enabling automatic mode automatically
shifts server resources to the applications in higher demand by starting and
stopping server resources appropriately. Do the following to use dynamic clusters
in automatic mode:
- Launch the Deployment Manager Administration Console.
- Select Servers > Dynamic Clusters.
- Select the dynamic cluster that WebSphere Commerce is installed to.
- Configure the dynamic cluster properties, for example:
- Keep 1 instance started at all times
- Do not limit the number of instances that can start
- Do not allow more than 1 instance to start on the same node
- Return to the Dynamic Cluster collection panel.
- Select the Dynamic Cluster that WebSphere Commerce is installed to.
- Set the mode to automatic.
Application Servers within the dynamic cluster will
now be started and stopped as demand shifts.
- Do the following to monitor remote servers:
- WebSphere Extended Deployment provides a remote monitoring agent for mixed
server environments to provide health monitoring of external nodes such as
Web servers and databases.
- The remote agent sends average CPU utilization, node speed, number of
CPUs, and several other attributes and values to the plug-in that polls the
remote agent for CPU information.
- Consider installing the agent to both the Web server and database nodes
to monitor them from the deployment manager.
- WebSphere Extended Deployment provides charting and mapping capabilities
in order to monitor the topology and health of the cell. Do the following
to configure runtime operations:
- Launch the Deployment Manager Administration Console.
- Select Runtime Operations > Runtime Map.
- The runtime map displays an overall picture of the
cell hierarchy. Components which are experiencing health problems are highlighted.
- Select Runtime Operations > Topology. Various charting scopes and
data are available to plot.
- To save statistical data, enable the Visualization Data Server. Select System
Administration > Visualization Data Server . Visualization data can be
stored for a specified amount of time.