Unit 2: Grid Terms and Concepts.
Unit 2 taught me the basics of the concepts of grid computing. This unit made me value more the ability of distributed systems to combine different resources, including computation, storage, communication bandwidth, software, and specific hardware, into a single infrastructure. The understanding of a grid as a synchronized assembly of heterogeneous machines indicated the difficulty of efficient utilization and administration of shared resources.
Among the most important lessons, I learnt the importance of scalability. The parallel processing and workload distribution discussion revealed that, to effectively utilize the performance of a grid, it is not just necessary to have multiple processors at the ready, but also to have the level at which application can be subdivided into autonomous tasks. I also was aware that the perfect scalability of jobs, delays in communication, and contention between resources could be constrained by interdependencies between jobs, and as such; scheduling emerged to be an important part of grid systems.
The unit also reiterated the importance of distributed storage systems where it described the use of Andrew File System and General Parallel File System technologies that allow unified name space, data striping, and enhance reliability. Learning about the role of data grids in increasing capacity and performance level enlarged my view on managing large amounts of data.
Moreover, the differences among scheduling, reservation, and scavenging helped to understand that grids achieve efficiency-predictability balance. Of particular interest to me was the way in which policy constraints, software licensing requirements, and security requirements can affect the job allocation process. The difference between intragrids and intergrids was also another factor that strengthened the growing significance of governance and security as grids increase across organizational lines.
On the whole, this unit enhanced my conceptual knowledge of the interactions between the philosophical (technical, administrative), and policy factors in order to maintain the efficient functioning of the grid. It has increased my capacity to critically investigate distributed computing settings and comprehend real world grid situations both technically and organizationally.