User:Gauravd05/Sandbox

Gaurav's Sandbox

Overview of ExpressCluster
ExpressCluster (sold as CLUSTERPRO in Japan) is NEC’s enterprise level solution for high-availability cluster and disaster recovery on the Windows and Linux platforms. It is widely used in Japan and is gaining popularity worldwide. The solution was first launched in 1996, since then it has undergone many changes and enhancements and gained a leading position in Japanese market.

It rapidly established itself in Japan's high availability and disaster recovery market as a market leader. It is a leader in the Japanese market since 2004. With the success of CLUSTERPRO in Japan, NEC launched CLUSTERPRO in global market under the brand name ExpressCluster in 2004.

Failover
ExpressCluster supports multiple configurations and can include up to 32 nodes in a cluster, each of which is either active or standby server. Active server is the server which actually executes the application in normal scenario. Standby server is a server which is always ready to take over the responsibility of executing an application in case active server/application fails. This switching of application from active server to standby server in case of failure is known as failover. Similarly switching of application from standby server to active server (when active server again recovers) is known as failback. Failover can also be executed manually for scheduled maintenance of active server.

ExpressCluster supports both automatic and manual failover and failback. In case of automatic failover, as soon as the active sever fails, ExpressCluster automatically switches the running applications and data to standby server with near zero downtime.

Support for various Server Applications
ExpressCluster can failover any windows or Linux based application, but it has some inbuilt support for key business application. ExpressCluster have additional agent software’s to detect minutely the failure in following key applications.

• Microsoft® Exchange Standard & Enterprise editions

• Microsoft SQL Server Standard & Enterprise editions

• Oracle® DB 9i/10g Standard & Enterprise editions

• IBM® DB2® Standard & Enterprise editions                     ...and more

With the help of these agents ExpressCluster can not only detect the existence of application on system but can also go at deeper level to know the exact status of an application. For instance in case an application is hung due to some reason, ExpressCluster can detect such a situation using agents and can failover the hung application along with other related cluster resources to standby server.

Clustering in Virtual Environment
Virtualization is popular technologies now days due to various factors including cost and infrastructure utilization. ExpressCluster also provides support for virtual environments like VMware and Hyper-V etc. ExpressCluster can failover at both Guest OS level and Host OS level and supports combination of Virtual and Physical failovers. E.g., it can failover whole virtual machine (when running in host machine environment) and also specific applications on virtual machine (when running in virtual machine environment).

Failover Groups
Using ExpressCluster, network administrator can create a failover group to represent a group of applications and related resources that need to be switched to standby server in case of application, OS or hardware failure. On failover, all applications and related resources which are part of failover group are automatically switched to standby server with near zero downtime. Resources which can be a part of failover group are necessary for proper working and accessibility of application on standby server. e.g. floating IP, mirror disk, shared disk, folder sharing permissions, virtual IP and virtual host name etc.

Monitoring Targets
Most of the clustering solution available today deployed the mechanism of knowing process existence to detect any failure and many of those do not detect OS or hardware level failures, but ExpressCluster have inbuilt deeper level of failure detection mechanism. ExpressCluster can quickly detect application hung-ups, application termination, internal OS issues and even hardware faults. It always keeps its communication open with application, OS and hardware layers to know their latest health status. In case of any issue being detected, it immediately switches the failover group to standby server enabling service continuity at all times.

Failover Cluster Topology
ExpressCluster can be configured in different cluster topologies.

Active - Standby – This topology represents a group of two servers. One of these two servers is assigned an active role and another server acts as standby server. In normal condition, all failover group applications execute on active server. Whenever any failure occurs either in application, operating system or hardware of an active server, ExpressCluster quickly and accurately recognizes an error and switches the failover groups to their respective secondary servers (standby servers).

Active - Active - This topology also represents a server group of two servers. In this case both servers are active but in relation to different failover groups (different instance of same application or different applications). At the same time they also act as standby server for another failover group running on another server. This topology enables administrators to utilize both servers simultaneously.

M active - N standby [0 <= M <= 31, 0 <= N <= 31, M + N <= 32] (for shared disk clustering) - This topology is an extension of active-standby topology providing network administrator more flexibility in designing cluster.

ExpressCluster also supports multiple other variations of the configurations mentioned depending on an organization’s need.