Unit 2 provides a full understanding of grid computing through the description of its terminology, resources, components, and functioning.
This unit starts with the introduction on the nature of resources found in a grid. The most critical resource is computation, where several computers share processor time to perform tasks either remotely, in parallel, or repeatedly on various nodes. Scalability is defined as the efficiency of additional processors in improving computation performance.
The unit continues by explaining storage resources, which comprise memory and secondary storage. Grid computing enables the integration of distributed storage resources into a single system (data grid), enhancing capacity, speed, reliability, and data sharing through concepts such as unified file systems, data striping, and journaling.
Communication resources are also critical, where fast networks enable computers to share jobs and data. Bandwidth, internal and external connectivity, and redundant communication paths influence grid performance and reliability.
The unit also covers software and licenses, where grids enable the use of specialized or costly software by routing jobs to computers with the software installed. License management systems are used to control simultaneous usage to cut costs.
It also discusses special equipment, architectures, and policies, pointing out that the machines in a grid may vary in hardware, operating systems, or capabilities. Grid administrators can set policies and artificial resource attributes to manage job execution.
In the jobs and applications section, the unit explains that applications are made up of several jobs, which may have dependencies and workflows. Jobs are distributed to the grid and run in parallel whenever possible, with results gathered to generate final outputs.
The terms scheduling, reservation, and scavenging are used to explain resource allocation. Schedulers allocate jobs based on availability and policies. Scavenging involves using idle machines for grid tasks, while reservation allocates resources for high-priority tasks.
The unit also discusses grid software components, which handle communication, security, scheduling, and resource management.
Finally, it discusses the various grid configurations, including intragrid and intergrid, which involve grids within one organization and grids between different organizations, respectively. The unit points out the increasing complexity, policy needs, and security needs as grids grow in size.
Unit 2 provides a solid foundation in grid computing by discussing resources, management techniques, job processing, and system architecture, allowing learners to understand how distributed systems work together efficiently.