Configure to failover clustering Network | What is the failover clustering ? Configuration failover cluster concept

What is the Failover Clustering?

A failover cluster is a set of servers that work together to increase the availability of services and applications. The clustered servers (called nodes) are connected through a network connection (physical or virtual) and by software. If one of the nodes fails, another node begins to provide services (a process known as failover). Failover clusters can be used for a wide range of network services including database applications such as Exchange Server or SQL Server, file servers, or network services such as Dynamic Host Control Protocol (DHCP) services.

The most common failover cluster is the active-passive cluster. In an active-passive cluster, both servers are configured to work as one, but only one at a time. The active node provides the network services whereas the passive node waits for something to happen to the active node where it cannot provide network services. If the active node goes down, the passive node becomes the active node and resumes providing the network services. When the failed node is restored, it becomes the passive node.

Active-passive clusters provide high availability but do not provide scalability. The only exception to the scalability issue is if you implement file servers on Cluster Shared Volumes (CSV). If you need higher performance, you can typically reduce the load or use more powerful hardware for the individual nodes.

Clustered services:-

Clustered services are services or applications that are made highly available by installing them on a failover cluster. Clustered services are active on one node, but can be moved to another node.

Nodes: Servers that make up the cluster and that run the Cluster service. They host the resources and applications associated to cluster.

 

• Network:  A common network that connects the cluster nodes. Three types of networks can be used in a cluster: public, private, and public-and-private.

 

• Cluster storage:  A storage system that is shared between cluster nodes and usually connects using fiber channel or iSCSI.

 

• Clients:  Computers (or users) that use the Cluster service.

 

• Cluster service:  The service that runs on Windows servers that manages and coordinates cluster resources to provide high availability.

CONFIGURING CLUSTER NETWORKING:-

When you connect the cluster to the network, you should consider using the following networks:

 

• Private network:  Used by cluster nodes to communicate with each other.

 

• Public network:  Used by the clients to access the cluster and its shared resources.

 

• Public-and-private network: Used to communicate with external storage systems.

Similar to a Network Load Balancing (NLB) cluster, a failover cluster has full connectivity and communication with the other nodes in the cluster using a private network. In addition, the cluster is aware when a node is added or removed from the cluster. Cluster nodes are kept aware of the status of the other nodes and services through the use of

 

Heartbeats:  Heartbeats transmit and receive using UDP port 3343 unicast (legacy clusters used UDP broadcast).

CONFIGURING CLUSTER STORAGE:-

Most failover clusters use shared storage to provide consistent data to all cluster nodes. There are three shared-storage options for a failover cluster:

• Shared serial attached SCSI (SAS):  A low-cost option that requires the cluster nodes to be physically close to the drives. Typically, the SAS has a limited number of connections for cluster nodes.

• Internet SCSI (iSCSI): A type of storage area network (SAN) that transmits SCSI commands over an IP network. iSCSI drives can be connected to a Windows server using the iSCSI initiator, which is included with Windows Server 2016.

• Fiber Channel: A form of SAN technology that is based on fibre optics. It typically offers better performance than iSCSI SANs but is more expensive because it requires specialized equipment.

• Fiber Channel over Ethernet (FCoE):  A form of SAN technology that uses the Fibre Channel protocol using 10 Gigabit Ethernet networks or higher on a converged network.