Cloud Computing A Series: Grid Computing
Grid Computing: Compute as a resource
Grid Computing a predecessor to the modern-day cloud computing, basically revolves around the concept, “Compute as a resource” what does this mean. Servers, datacentres contain hundreds if not thousands of hardware and require the specific software complacency for the given work to be executed for a particular organization. So the idea came up of, what if compute can be offered as a resource where all the hardware required would be centralized, and could be easily accessible through a middleman such as a broker that would process the requests and perform the appropriate actions.
Grid computing was meant to be used by individuals who gain access to computing devices without knowing where the resource is located or what hardware it's running on.
Particular emphasis was given to heterogeneous infrastructure(the infrastructure that more than one type of O.S’s and memory)
Resource agreements had to be made between any connection on what resources could be shared between various organizations.
Sharing was highly controlled with resource providers and consumers grouped into virtual organizations * primarily based on sharing conditions.
*Virtual Organizations
Virtual Organization form the basic unit for enabling access to shared resources with specific resource sharing policies applicable for users.
Features of a Grid
- Coordinates resources that are not subjected to centralized control
- Using standard, open, general-purpose protocols and interfaces
- To deliver the non-trivial quality of service
Benifits of Grid Computing
- Exploit under used resources
- Resource Load Balancing
- Virtualize resources across an enterprise
- Data Grids, Compute grids
- Enable Collaboration for virtual Organizations
The official definition of Grid Computing is “A computational grid is a hardware and software infrastructure that provides dependable, consistent, and inexpensive access to high-end computation”.
The Grid Layered Architecture
The grid can be very much compared to an electrical power grid, which can be used to better explain the architecture. The grid resources act as the power stations that contain the power, i.e the power lines perhaps, the resource broker mentioned above acts like the power station i.e the centre that controls access to electricity in your home, the user can be described as a toaster, that “requests ” electricity from the central power grid.
The Architecture of Grid Computing is as follows:
- Application Layer: Applications running on the grid (Toaster)
- Collective: the computing resource required (Ac or DC)
- Resource: The resources required to perform grid computing (Solar panels)
- Connectivity: enables communication between the fabric layer and user (Power lines / Power station )
- Fabric Layer : Controls things locally (Switch)
Grids are typically managed by middleware(resouce broker): A special type of middleware that enables sharing and manage grid compnents based on user requirements and resource attributes.
