在虚拟化的平台上运行关键应用U2VL.pptVIP

谢谢！ 期待与您共同思考和探索！ Print Only (The text is same as previous presentation slide) Now that the data is freed from the monolithic prison of the RDBMS, we can move each group of data into the same virtual machine (and in many cases the same operating system process) as the application tier. Now each virtual machine will take a portion of the Web, Application and Data Management load. These virtual machines are easier to deploy because they no longer have a configuration dependency on a remote database instance. We can add or remove instances of these virtual machines as required in order to respond to changes in the application load and the data will rebalance itself across the available members of the cluster. New members of the data virtualization layer will discover each other automatically without changing configuration files or restarting the system. When any member fails or shuts down, the other members will already be prepared with a copy of the data to take over the load. In order to support legacy relational and OLAP applications, the data management layer will keep other databases up to date with changes. This update channel can be either synchronous or asynchronous with the application updates, but using the asynchronous option will allow for lower latency and higher throughput for the application. Print Only (The text is same as previous presentation slide) Now that the data is freed from the monolithic prison of the RDBMS, we can move each group of data into the same virtual machine (and in many cases the same operating system process) as the application tier. Now each virtual machine will take a portion of the Web, Application and Data Management load. These virtual machines are easier to deploy because they no longer have a configuration dependency on a remote database instance. We can add or remove instances of these virtual machines as required in order to respond to changes in the application load and the data will rebalance itself across the