Scribd
Upload
20 views3 pages

Working Principle of Grid Scheduling

Uploaded by

Varshini
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
20 views3 pages

Working Principle of Grid Scheduling

Uploaded by

Varshini
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Grid scheduling in grid computing is the process of efficiently allocating resources (e.g.

, CPU,
memory, storage) across a distributed and heterogeneous set of nodes within a grid to complete
tasks or jobs. Grid scheduling involves deciding which task should run on which resource at a
particular time, taking into account various factors like availability, priority, job requirements, and
resource characteristics.

Working Principle of Grid Scheduling

The grid scheduling process typically follows a series of steps that allow the system to optimize
resource usage, minimize delays, and ensure efficient execution of jobs. Here's an outline of the
working principle:

1. Job Submission and Requirements Specification

 A user submits a job (or task) that needs to be executed in the grid environment.

 The job usually comes with specific requirements such as:

o Computational resources (e.g., CPU power, memory).

o Storage requirements (e.g., data access, file storage).

o Time constraints (e.g., deadlines, maximum execution time).

o Priority (e.g., high-priority jobs, low-priority jobs).

The user might also provide preferences regarding the type of resources or geographical location
where the task should be executed.

2. Job/Task Classification

 The submitted job is classified based on:

o Job type (e.g., parallel, sequential, data-intensive).

o Priority (e.g., normal, urgent).

o Resource requirements (e.g., specialized hardware, memory, or disk space).

Jobs may be categorized to ensure they are directed to appropriate resources that meet the specific
requirements.

3. Resource Discovery and Matching

 The grid scheduler queries the resource management system to discover available resources
(nodes) in the grid. These resources can include clusters, supercomputers, virtual machines,
and even idle desktops.

 The scheduler then matches the job requirements with the available resources based on:

o Resource availability: Checking if sufficient computational power, storage, or other


resources are available.

o Resource capability: Ensuring the resource can meet the specific needs of the job
(e.g., specialized hardware, environment, or software dependencies).
o Location: Depending on the job, location-based constraints may be considered, such
as proximity to data for faster access.

4. Job Scheduling and Execution Allocation

 Once the suitable resources are identified, the grid scheduler assigns the task to a resource
or set of resources.

 The allocation is performed based on scheduling policies, which could include:

o Static Scheduling: Pre-determined policies, such as round-robin or time-sharing.

o Dynamic Scheduling: Adaptively assigning tasks based on the current grid conditions
(e.g., load balancing, available resources).

o Deadline-based Scheduling: Tasks that have deadlines are prioritized.

o Fairness-based Scheduling: Ensuring that all jobs get a fair share of resources.

5. Execution Monitoring

 Once a job starts execution on a resource, the system continuously monitors its progress to
ensure that it runs efficiently.

 If the job is not performing optimally (e.g., due to resource contention, overload, or failure),
the grid scheduler may intervene by:

o Re-scheduling the job to another node.

o Migrating the job (moving it to a different resource if the original resource becomes
unavailable or too slow).

o Preempting lower-priority tasks to make room for higher-priority ones.

6. Data Management and Communication

 Grid scheduling often involves coordinating the transfer of data between different nodes in
the grid. The scheduler must ensure:

o Data locality: Jobs requiring large datasets should ideally be executed on nodes close
to where the data is stored, minimizing data transfer time.

o Data consistency: Ensuring that the right version of the data is used and that the job
receives the data it requires.

o Communication optimization: Minimizing network latency and bandwidth usage


during job execution.

7. Completion and Feedback

 Once the job finishes execution, the system retrieves the results and stores them in the
appropriate location.

 The grid scheduler updates its state to reflect resource availability and usage.

 The user is notified about the job's completion and its output.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy