Windows Cluster Development-write Microsoft Cluster Server (MSC) Resource DLL: 2) Introduction

Introduction

The Microsoft Cluster Server (MSC) connects multiple Windows NT-based operating systems to present them to network customers as a single, high-availability system. From the perspective of the system administrator, MSC provides additional management convenience and scalability, And the MSC architecture is scalable for future versions and can be perceived by clusters (cluster-aware) the application provides a standard infrastructure.

This article aims to describe the well-defined applications from a high-level perspective of the process in order to make full use of the advantages of MSC. It also includes how to make full use of the advantages of MSC, compiling resource DLL, debugging applications and services, and installing them in a cluster environment.

Note:This document assumes that you have successfully installed the cluster software in the cluster environment, and have a Microsoft platform software development kit (platform software development kit), and the compiling environment can work properly. If you encounter problems in the development environment, see the SDK documentation.

Clusters and high availability

By targeting client requests based on resource availability and server load, MSC allows applications and services to run on Windows NT servers in a more effective manner. (In the first version of MSC, load balancing is performed manually; in the future, automatic load balancing will be provided) if a system or node in the cluster fails or fails due to a hardware or software fault, its working set will be taken over by other systems in the cluster until the system that fails returns to normal.

Note:Designed with high availability rather than true fault tolerance ). The word "Fault Tolerance" is generally used to describe technologies that provide higher levels of recovery and recovery capabilities. Typical Fault-Tolerant servers usually use advanced hardware and data redundancy and are combined by specially crafted software to provide near-real-time recovery for any single-point hardware or software failure. The cost of this solution is much higher than that of the cluster because redundant hardware must be purchased for fault recovery, which is usually idle. Using standard and low-cost hardware, MSC provides a high-availability solution that maximizes the use of computing resources.

Non-shared node Mode

Version 1.0 is a two-node cluster based on the non-shared node mode. The non-shared node Mode means that when several nodes in the cluster may access a certain device or resource, the resource can only be owned and managed by one system at any time. (In the MSC cluster, resources are defined as any switch between online and offline states, managed by the cluster, and only owned by one node at a certain time, hardware or logical components that can be transferred between nodes)

Each node has its own memory, system disk, operating system, and a subset of cluster resources. If the node fails, other nodes will gain ownership of the resources of the failed node (this process is called failover ). MSC registers the network address of the new nodes for these resources to route client requests to the nodes that currently have these resources and are available. When the failed system returns to the online status later, you can set up the MSC to properly resend resources and client requests (this process is called fault recovery failback ).

Note:When the node fails, all clients will lose the connection. To achieve truly transparent failover, the client application must be able to establish a new connection when the node fails.

Shows the general settings of the MSC cluster:

Figure 1. standard two-node MSC Configuration

The following sections describe the structure of the MSC.