In Unit 2, my understanding of grid computing expanded greatly. Before this unit, I thought grid computing only involved connecting many computers to increase processing power. However, I learned that it is actually a distributed computing system that enables the sharing of different resources such as CPU power, storage, bandwidth, software, and specialized hardware across multiple locations and organizations.
One important concept I learned is how grid systems handle applications and jobs. Large applications can be divided into smaller tasks that run simultaneously on different machines. This parallel processing improves performance and reduces execution time, although efficiency may sometimes be affected when tasks depend on shared resources.
I also gained knowledge about data grids, where data is stored across multiple systems instead of a single central location. Techniques like data striping improve speed and access, but they require proper management to ensure reliability and data security.
Scheduling mechanisms were another key learning area. I understood how schedulers allocate resources using approaches such as normal scheduling, reservation, and scavenging. These methods help balance system efficiency, fairness, and predictability depending on workload demands.
Additionally, I learned that grid systems can operate as intragrids within one organization or intergrids across multiple organizations, which require stronger coordination and security. The unit also explained why grid computing is important, especially for solving large scientific problems and utilizing idle computing resources efficiently.
Overall, this unit helped me realize that grid computing is not just about powerful machines but about effective coordination, resource sharing, and management of distributed systems to solve complex problems efficiently.