Implementation of centralized management of high availability clusters by IBM System Director

The application environment is a SAN architecture with IBM Power7 's minicomputer and IBM DS5020 storage as its core. The article is based on the graphical interface as a datum point, the interception of a large number of operation results screenshots and a text for each picture, I hope that through the text of the picture plus writing mode to the reader to bring a more intuitive understanding, and to bring you help and inspiration.

This section provides a description of some of the terms and abbreviations that appear in this article. To avoid lengthy articles, there will be no repetitive explanations.

IBM Director Server

IBM Director Server is a powerful cross-platform management tool. It simplifies the management of complex physical and http://www.aliyun.com/zixun/aggregation/12423.html "> Virtual Systems". Through the browser for graphical interface operation, unified management of user environment all the various devices. This article is used to manage all parts of IBM PowerHA systemmirror using IBM Director Server.

LPAR

Logical partitioning is to divide a single server into multiple logical servers and run separate applications with one another.

VIOS

Virtual I/O Server (VM) is used to centrally manage logical partitions on physical hosts.

LUNs

The logical unit number, which is both a virtual disk. This article can be seen as a disk.

Shared storage and shared workloads

shared storage, shared disks. In this article, shared storage is interconnected with VIOS physical machines, and the LUN is assigned to VIOS. The disks on these storage are then assigned to the LPAR by VIOS.

Cluster

Clusters, with a community environment consisting of more than two physical or virtual hosts. All hosts within a cluster can communicate with each other and can perform complementary operations.

Node

node, to support the host cluster exists, can be physical host can also be a virtual host, the LPAR in this article is a virtual host.

Service IP

Service IP address. is the IP address of a highly available cluster to provide services externally.

Heartbeat

Heartbeat, when a node in a cluster crashes, the crash message tells other surviving nodes through the heartbeat. This allows Active nodes to take over the application services of the crash node. There are three types of heartbeat: Disk Heartbeat, network heartbeat and serial heartbeat. This article uses a disk heartbeat.

Recourse

Resources, refers to the hardware, software, or files on the host. For PowerHA, it can also be a combination of software and hardware.

Recourse Group

A resource group, a combination of multiple resources. In HA environment, it includes network resources, heartbeat resources, storage resources, service resources, monitoring resources and so on.

IPAT

IP address takeover node takes over and hosts service IP addresses on one node to other nodes.

Figure 1. Topological diagram of experimental environment

From the topology above, we can see that there are two clusters in this experiment, and each cluster contains two nodes. All nodes are lpars on the same physical machine, and these lpars are managed by the same VIOS. VIOS is connected to an external storage, IBM Storage 5020. This storage provides the shared disks required by each cluster. Director Server is an LPAR on another physical machine that is also installed on an AIX operating system.