In this unit, I understood that grid computing is much more than just connecting many computers together. At first, I thought it was mainly about combining processing power, but now I see that it involves many types of resources like CPU, storage, communication bandwidth, software licenses, and even special hardware or policies.
One thing that really helped me understand the topic is how applications and jobs work in a grid. I learned that an application can be divided into smaller jobs, and those jobs can run in parallel on different machines. This makes processing faster, especially for large or complex tasks. However, I also understood that it is not always perfectly scalable because some jobs depend on others or compete for resources.
I also found the idea of data grids interesting. Storage is not just in one place; it can be spread across many machines. Techniques like data striping help improve speed by dividing data across different storage devices. But at the same time, this requires proper management to avoid data loss or system failures.
Another important concept for me was scheduling. The scheduler decides which machine should run which job, depending on availability, policies, and priorities. I now understand the difference between scavenging (using idle machines), reservation (booking resources in advance), and normal scheduling. Each method has its advantages depending on whether we want efficiency or predictability.
I also realized that security and management are very important, especially when grids grow larger or cross organizational boundaries. Authentication, monitoring, and load measurement tools help maintain control and reliability.
Overall, this unit helped me clearly understand how grid computing systems are organized and managed. It showed me that performance is not just about powerful machines, but also about smart scheduling, resource management, communication efficiency, and proper policies working together.