vertical cluster

One system can be supported by more than one node, forming a cluster. A vertical cluster consists of two or more application server instances on a single server.

For example, two copies of IBM WebSphere Application Server, each running a copy of Pega 7 Platform, can be installed on a single IBM server. A single PegaRULES database supporting both installations can reside on this server on another server. Interactive users connect to one or the other node based on distinct TCP/IP ports.

Vertical clustering provides both throughput and uptime benefits:

• Each application server uses a dedicated Java Virtual Machine, providing potentially better utilization of real memory and greater throughput than can be achieved by a single JVM.
• The application servers can be started and stopped independently, so changes that require stops do not require user downtime. When one is failed or temporarily stopped, more CPU resources are available to the other, so the true throughput loss may be less than 50 percent.

You can use horizontal clustering in combination with vertical clustering.

Related terms

Open topic with navigation