Unit 2: Grid Terms and Concepts
COE4261 – Grid Computing
In Unit 2, my understanding of grid computing really changed. Before this unit, I used to think that grid computing was simply about connecting many computers together to increase processing power. However, after studying the concepts in detail, I realized that it is much broader than that. Grid computing involves sharing different types of resources such as CPU power, storage, bandwidth, software licenses, and even specific hardware and policies across distributed systems.
One concept that helped me understand the topic better was how applications and jobs are handled in a grid environment. I learned that a large application can be divided into smaller jobs that run in parallel on different machines. This improves performance and reduces execution time for complex tasks. At the same time, I understood that scalability is not always perfect because some jobs may depend on others or compete for shared resources.
I was also interested in learning about data grids. Instead of storing data in one central location, data can be distributed across multiple machines. Techniques like data striping improve performance by splitting data into parts and storing them separately. However, this also requires strong management to ensure reliability and prevent data loss.
Another important concept for me was scheduling. I learned how schedulers allocate jobs to available resources based on policies and priorities. Understanding the difference between scavenging, reservation, and normal scheduling helped me see how grids balance efficiency and predictability depending on system needs.
Overall, this unit helped me see that grid computing is not only about powerful machines but about coordination, management, and smart resource allocation. It has strengthened my understanding of distributed systems and how large-scale computing environments are organized.