Design And Implementation Of High Availability System

High availability (HA) is a balance between availability and cost. Based on the analysis of high availability key techniques of computer system, a research on the architecture of a high availability system based on shared-storage is done. A prototype of two nodes, service-group oriented high availability system based on shared-storage is designed and its management module is implemented. It will have good scalability. In order to improve system availability and applicability, combining system design and implementation, integrating applications of relevant technology and research, we design a high-availability system model. Through system-level redundancy and high availability systems management software schedule processing, our high-availability system can handle fault and ensure data security in lower costs to achicve very high system availability. We integrate all kinds of technology to improve system availability, and the system has good applicability.This thesis discusses the concepts related to system availability. In order to improve system services time and reduce the system recovery time. We designed a Hot-Standby system from system architecture. Once the system fails, we immediately detect the fault, and finish the master/slave switch on the server. This is transparent to the users and system availability and applicability are also improved. In addition, we designed a management software to protect system availability. We designed a heartbeat module, a monitoring module, a data backup module and an error recovery module. We focused on the designing of the heartbeat module and the information exchange between monitoring data modules. This ensures software monitoring capabilities and improves system response when failure occurred, the system can switch to another available server immediately. Finally, we have tested the system. The result shows that our system meets the high availability parameters required. By using the relevant technology, we improve the system availability and applicability.