The long-running scheduler is a
Java 2 Platform Enterprise Edition (J2EE) application. The WebSphere Extended
Deployment Operations Optimization component is required to host the job scheduler
in a dynamic cluster. To make the long-running scheduler highly
available, deploy it to a dynamic cluster within your WebSphere Extended Deployment
environment. Only unclustered application servers or static clusters are supported
if you deploy Compute Grid without the Operations Optimization component.
Before you begin
Determine
which nodes in your cell host the
long-running scheduler.
The administrative console, application clients, J2EE applications and Web
service requests access the
long-running scheduler.
The f
long-running scheduler must access the
long-running scheduler database, and the nodes
that host execution environments.
Privileges for the long-running scheduler differ,
depending on the various roles. Roles include monitor, operator, configurator,
and administrator. If you are a user with either a monitor or an operator
role, you can only view the long-running scheduler information.
If you have the role of configurator or administrator, you have all the configuration
privileges for the long-running scheduler.
Procedure
- In the administrative console, select System administration
> Node groups.
- Click New.
- Type an appropriate name for the node group.
For example, LongRunningSchedulerNG .
- Click OK.
- Select the node group that you created from the table.
- Select Node group members, then click Add.
- Check the boxes next to the nodes that host the long-running scheduler,
then click Add.
- In the administrative console, select Servers > Dynamic Clusters.
- Click New.
- Type an appropriate name for the dynamic
cluster. For example, LongRunningSchedulerDC.
- In the Map to node group menu, select the node group that
you created.
- In the Default application server template menu, select defaultXD.
- Make sure Keep one instance started at all times is selected.
- Select Limit the number of instances that can start, then
type 1 in
the text field. For a high availability environment, use a dynamic
cluster with the minimum and maximum, both set to 1, automatic mode, and an on demand router
in front of the dynamic cluster.
- Click Next, then Finish.
- If you want WebSphere Extended Deployment to automatically manage
the location of your long-running scheduler,
select Automatic from the mode menu, check the box for the dynamic
cluster that you created, and click Set mode.
Results
The following example shows the steps you need to perform only for
static or dynamic cluster members added after the job scheduler has been assigned
to a host. When the job scheduler is assigned a host, WebSphere Extended Deployment
automatically performs the following steps for cluster members already in
a cluster.
Example
- In the administrative console, select Servers > Application servers.
- Click the name of the application server in the dynamic cluster. The application
server's name will be in the form dynamicclustername_nodename.
- Select Container Services > Startup beans service.
- Check the box next to Enable service at server startup and click OK.
- Repeat for other application servers in the dynamic cluster.
- Save your changes.
- In the administrative console, select Environment > Virtual Hosts.
- Click grid_host.
- Click Host Aliases.
- Click New.
- Under Host Name, type a virtual host name, or *.
- Under Port, type the WC_gridhost port.
- Click Apply.
- If security is enabled, repeat these steps to create
host aliases based on the WC_grid_host_secure port.
- Restart server.