DUAL SITE CLUSTER OF APPLICATION SERVERS ON HYPER-V

Business critical applications require a solution that is not dependent on information equipment placed at just one location. To achieve downtime below 52.5 minutes annually, corresponding to availability of about 99.99% it is necessary to provide services at more than one location. Dual site cluster is presented that uses virtualization platform at each location and makes a favorable price / performance ratio while ensuring an extremely high availability. On each virtualization platform, several application servers can be installed, ensuring both great user capacity and granularity of solution, all with the using very favorable per CPU licensing. Separate NLB clusters of application servers installed on each virtualization platform are integrated into a dual site cluster using the highly available geotargeting redirections. Overprovisioning in the event of locations failures is achieved using cloud bursting to a public cloud. Details of implementation are given.