Configuring High Availability
< Go Back
This video describes how to configure the high-availability features of Windows HPC Server 2008, which takes advantage of the enhanced failover clustering capability of Windows Server 2008 to provide job resiliency when a cluster head node fails.
The video begins with a high-level view of Windows HPC Server 2008 and a detailed description of the Microsoft HPC Pack 2008 and its four functions: networking and MPI, job scheduling, storage, and system management (which includes high availability for the head node). The video then describes the components of the high-availability solution: two head nodes and shared storage. If the primary head node fails, the failover clustering feature of Windows Server 2008 ensures that functions are failed over automatically to the secondary head node. Compute nodes can continue running jobs, and clients can submit jobs after a couple of seconds. Users can also fail over manually if needed; for example, in situations where the head node needs to be upgraded or patched.
The video then provides a step-by-step description of how to set up the cluster for high availability in Windows HPC Server 2008: configure a 2-node failover cluster (both head nodes must use Windows Server 2008 Enterprise or higher), set up and configure SQL Server 2005 SP2 Standard on the 2-node cluster, and run the HPC Pack Setup on each of the two head nodes in the failover cluster. Each step is described in detail, and considerations unique to the Windows HPC Server 2008 setup are noted throughout the process.
This video describes how to configure the high-availability features of Windows HPC Server 2008, which takes advantage of the enhanced failover clustering capability of Windows Server 2008 to provide job resiliency when a cluster head node fails.
The video begins with a high-level view of Windows HPC Server 2008 and a detailed description of the Microsoft HPC Pack 2008 and its four functions: networking and MPI, job scheduling, storage, and system management (which includes high availability for the head node). The video then describes the components of the high-availability solution: two head nodes and shared storage. If the primary head node fails, the failover clustering feature of Windows Server 2008 ensures that functions are failed over automatically to the secondary head node. Compute nodes can continue running jobs, and clients can submit jobs after a couple of seconds. Users can also fail over manually if needed; for example, in situations where the head node needs to be upgraded or patched.
The video then provides a step-by-step description of how to set up the cluster for high availability in Windows HPC Server 2008: configure a 2-node failover cluster (both head nodes must use Windows Server 2008 Enterprise or higher), set up and configure SQL Server 2005 SP2 Standard on the 2-node cluster, and run the HPC Pack Setup on each of the two head nodes in the failover cluster. Each step is described in detail, and considerations unique to the Windows HPC Server 2008 setup are noted throughout the process.
