Unit 2: Grid Terms and Concepts
In this unit, I learned that grid computing is a system where multiple machines, or nodes, share resources such as computing power, storage, communication, software, and specialized devices. These resources can be heterogeneous, geographically dispersed, and managed as a unified virtual system.
I understood that computation in a grid can be used by running applications remotely, splitting workloads for parallel execution, or performing repeated executions across many machines. The concept of scalability is key to understanding how efficiently a grid uses multiple processors.
Storage resources can be combined into a data grid using distributed file systems, enabling techniques like data striping for higher performance, replication for reliability, and unified namespaces for easier access. Communication capacity, both internal and external, is critical to efficiently move jobs and data.
The unit also highlighted how jobs, applications, and subjobs are managed. Schedulers, reservation systems, and scavenging help optimize the use of resources while respecting policies and priorities. Grid management software monitors usage, enforces security, and can even automatically recover from failures using autonomic computing.
Finally, I learned about the difference between intragrids (within a single organization) and intergrids (across organizations), including the increasing importance of security, policies, and distributed management as grids grow in scale and complexity.
Overall, this unit gave me a solid understanding of how grid components work together and how grid computing enables efficient, scalable, and collaborative use of distributed resources.