|
Individual servers in a server cluster are called nodes.
A server cluster can consist of up to eight nodes.
The individual physical or logical components managed by a cluster are called resources. Resources are combined to form groups, which are the basic administrative unit of a server cluster.
Applications operating on a server cluster move between nodes in a process called failover. What happens during a failover is configurable and depends on the server cluster design.
The resource that is used to control the server cluster is the quorum. Every server cluster has a quorum resource whose form is determined by the cluster design model adopted.
Server clusters require very specific up-front planning for successful implementation.
NLB is used to increase availability for applications that service TCP/IP traffic only. Other protocols are not supported.
Individual servers in an NLB cluster are called hosts.
NLB clusters determine their operating state through a process called convergence.
NLB clusters can be administered from with a graphical tool (NLB Manager) or a command-line tool (NLB.exe). The graphical tool is more secure.
An NLB cluster does not require multiple network adapters in each host, although this is recommended.
NLB provides no additional security features and requires more stringent security practices than do stand-alone servers.
|