Clustered Systems in Operating System:
A clustered system uses multiple CPUs to accomplish a task. It is different from parallel system in that clustered system consists of two or more individual systems coupled together. The clustered computers share storage and are closely linked via LAN networking.
The clustering is usually performed to provide high availability. A layer of cluster software rims on cluster nodes. Each node can monitor one or more nodes over the LAN. The monitored machine can fail in some cases. The monitoring machine can take ownership of its storage. The monitoring machine can also restart applications that were running on the failed machine. The failed machine can remain down but the users will see a brief of the service.
Types of Clustered System:
The clustered system can be of the following forms:
Asymmetric Clustering: In this form, one machine is in hot standby mode and other machine is running the application. The hot standby machine performs nothing. It only monitors the server. It becomes the active server if the server fails.
Symmetric Clustering: In this mode, two or more machines run the applications. They also monitor each other at the same time. This mode is more efficient because it uses all available machines. It can be used only if multiple applications are available to be executed.
Some other forms of clusters are parallel clusters and clustering over WAN. Parallel clusters allow multiple hosts to access the same data on a shared storage. Most operating systems do not provide the support for simultaneous data access by multiple hosts. That is why parallel clusters are usually accomplished by special versions of software and special releases of applications. An example of such software is Oracle Parallel Server. It is a version of Oracle database that is designed to run on parallel clusters. Each machine runs Oracle and a layer of software tracks access to the shared disk. Each machine also has full access to all data in the database.
Most clusters do not support shared access to data on the disk. The distributed file systems must provide access control and locking to the files to prevent conflicting operations. This type of service is commonly known as distributed lock manager (DLM).
