You are on page 1of 12

Advance Distributed

Computing
Bareera Anam 327976
Junaid Sajid 330746
Muhammad Asim 328415
OUTLINE

❏ Distributed Systems
❏ Distributed Scheduling
❏ Load
❏ Problem Statement
❏ Architecture
❏ Task Generator
❏ Task Scheduler
❏ Computing Nodes
❏ Scheduling Algorithm
❏ Conclusion
Distributed Systems

A system with multiple components located on different


machines that communicate and coordinate actions in order
to appear as a single coherent system to the end-user.
Distributed Scheduling

❏ A distributed system may have a mix of heavily and a lightly loading


system.Hence, migrating a task to share or balance load can
help.Resource management component of a system which moves
jobs around the processors to balance load and maximize overall
performance
❏ As, if scheduling a process on in machine 1 or machine 2 and when the
process is finished, it needs to schedule a job on machine 3.
Load

❏ Load on a system or node can correspond to the queue length of


task or processes that need to be processed.
❏ Queue length is proportional to task response time.
❏ Distributed load is transferring tasks among the nodes.
❏ If nodes are highly nodded then it effects performance.
❏ If task transfer is failed, then use timeout.
Problem Statement

❏ We have to design a distributed data centre where we have three


types of loads

1- Task Generator

2- Load Balancer

3- Computing Nodes
❏ We have 10000 tasks dataset and we have to perform these tasks and
schedule these with a scheduling algorithm and load the balance
among these.
Architechture
Task Generator

❏ To generate tasks at the random interval of time.


❏ Sends generated task to scheduler.
❏ Each task contain Task_ID and Task_MIPS.
❏ Task_ID is unique GUID and Task_MIPS is random value
between 1000 and 5000 for each task.
Task Scheduler

❏ It takes task from task generator and perform


scheduling algorithms on each task and send to
the next computing node and for the processing
score of each task.
Computing Nodes

❏ The task scheduler forwards the tasks to nodes for the


processing of the tasks where task initially stored in a
queue.First Come First Serve used to assign computing
resources to the task.
Scheduling Algorithms

❏ First Come First Serve


❏ Shortest Job First
Conclusion

❏ An efficient approach in generating multiple tasks at a time and


then handle these tasks efficiently and flexibly with the help of
scheduling and generating algorithm.
❏ Aim is to provide an efficient working with the minimal response
time.
❏ Proposed algorithm distributed load uniformly as we can see from
the results with improve response time as compared to the existing
algorithms.

You might also like