Overview of the planning process

This topic describes the steps for planning a PowerHA® SystemMirror® cluster.

Step 1: Planning for highly available applications

In this step, you plan the core of the cluster: the applications to be made highly available, the types of resources they require, the number of nodes, shared IP addresses, and a mode for sharing disks (nonconcurrent or concurrent access). Your goal is to develop a high-level view of the system that serves as a starting point for the cluster design. After making these initial decisions, start to draw a diagram of the cluster. Initial cluster planning describes this step of the planning process.

Step 2: Planning cluster topology

In this step, you decide on names for the cluster and the nodes. Optionally, you also decide on names for sites and decide which nodes belong to which site. Initial cluster planning describes this step of the planning process.

Step 3: Planning sites

In this step, you decide whether the site will use a stretched cluster or a linked cluster. A stretched cluster contains nodes from sites at the same geographical locations. Stretched clusters must share a repository disk. A linked cluster contains nodes from sites that are located at different geographical locations. Linked clusters use a separate repository disk.

Step 4: Planning cluster network connectivity

In this step, you plan the networks that connect the nodes in your system. You first examine issues relating to TCP/IP and point-to-point networks in a PowerHA SystemMirror environment. Planning cluster network connectivity describes this step of the planning process.

Step 5: Planning shared disk devices

In this step, you plan the shared disk devices for the cluster. You decide which disk storage technologies you will use in your cluster, and examine issues relating to those technologies in the PowerHA SystemMirror environment. Planning shared disk and tape devices describes this step of the planning process.

Step 6: Planning shared LVM components

In this step, you plan the shared volume groups for the cluster. You first examine issues relating to LVM components in a PowerHA SystemMirror environment. Planning shared LVM components describes this step of the planning process.

Step 7: Planning resource groups

Planning resource groups incorporates all of the information you have generated in the previous steps. In addition, you need to decide whether to use dependent resource groups or particular runtime policies for keeping certain related resource groups on the same node or on different nodes. Planning resource groups describes this step of the planning process.

Step 8: Planning cluster event processing

In this step, you plan the event processing for your cluster. Planning for cluster events describes this step of the planning process.

Step 9: Planning PowerHA SystemMirror clients

In this step, you examine issues relating to PowerHA SystemMirror clients. Planning for PowerHA SystemMirror clients describes this step of the planning process.