Application Of Oracle RAC In The Social Security Data Center

In the background of "universal insurance", the social security number of coverage is increasing, the service contents are growing and the service area is expanding. The shortage problem of service sources management, means and tools is increasingly evident. The universal insurance puts forward higher requirements for the data center’s support ability. In order to ensure the smooth progress of application service of Aksu Prefecture Labor and Social Security Administration, and to realize the information and data sharing between application businesses, a provincial and municipal connected high-availability, shared data management center platform has been built.A solution of Oracle RAC is presented in this paper, using the failover and load balancing features of the Oracle 11 g R2, a social security database clustering has been established to meet the high availability requirements of users. According to the user needs, by comparing the Oracle RAC, host HA, Data Guard and other technical solutions, the Oracle RAC solutions have been proposed. From the actual situation of the data center, the planning work of hardware selection of Oracle 11 g R2 database, network planning, document storage, storage equipment selection is carried out, and the system installation and deployment is completed. This paper introduces the configuration method of fault transfer, through simulation of the service based failover transfer in the database downtime; the test is performed realizing the automatic failover without user intervention, to meet the high availability requirements. By configuring the load balancing based on the client-side and server-side, a large number of connection requests of users are simulated, and the load balancing of system is tested. Oracle RAC is an instance level redundancy protection without protection on the storage-level and network-level faults, which is disadvantage of Oracle RAC technology schemes.After the online operation of the social security center data system, the operation of the system is relatively stable with high performance. In addition to planned downtime for maintenance, there is no unplanned downtime failure. By viewing the server CPU utilization, system log and process state, the load balance is basically achieved improving the performance of the system. Not only it meets the needs of customers, but also creates better economic and social benefits. This technical solution not only can meet the construction requirements of current system, but also retains a very flexible scalability. It also can increase the number of nodes with the increasing of business, which provides the protection for the future expansion of the system platform.