Distributed Process Scheduling : A Summary

Download Report

Transcript Distributed Process Scheduling : A Summary

Distributed Process Scheduling :
A Summary
By
Pragati Sahu
System Performance Model

Precedence process Model
Applied for concurrent process.

Communication process Model
Applied for process that coexist and communicate asynchronously.

Disjoint Process Model
Process that run independently.
Speedup Factor
S= F(Algorithm,System,Schedule)
Static Process Scheduling

Mapping of process to processor is
determined before the execution process.
Precedence Process Model

Communication Process Model

Example
Example
Static Scheduling Challenges


Prior knowledge of execution time and
communication behavior of the process is
required.
Once a process is assigned to a processor it
remains there until completion of execution.
Dynamic Load Sharing and Balance

Sender initiated Algorithm

Transfer of process require 3 basic decisions. i.e.
Transfer Policy, Selection Policy and location policy.

Receiver initiated Algorithm

Receiver pulls process to be executed to its site.
Uses similar transfer policy i.e. activates pull when
queue size is below threshold.
More Stable than the sender.


Distributed Process Implementation
The three significant application scenario :
 Remote Service
The message is interpreted as a request for a known service at
remote site

Remote Execution
The messages contain a program to be executed at the remote site.

Process Migration
The messages representing process are migrated to the remote site
for continuing execution.
Real Time Scheduling
Rate Monotonic




Optimal static-priority scheduling
It assigns priority according to period
A task with a shorter period has a higher priority
Executes a job with the shortest period
Deadline Monotonic


Optimal static-priority scheduling
It is harder to analyze as no formula based on the load
that guarantee feasible schedule.
Real Time Scheduling
Earliest Deadline First



Optimal dynamic priority scheduling
A task with a shorter deadline has a higher priority
Executes a job with the earliest deadline
Recent Research Paper

Liu Dun-nan, Jiang Xin-fan, Hu Bin-qi ,hang Si-yuan, Real-time
scheduling feedback fuzzy control system based on area control
error and power generation error in :9th International Conference on
Fuzzy Systems and Knowledge Discovery (FSKD),2012.

Weijing Song,Shasha Yue, Lizhe Wang, Wanfeng Zhang, Dingsheng
Liu, Task Scheduling of Massive Spatial Data Processing across
Distributed Data Centers: What's New?, in: 17th International
Conference on Parallel and Distributed Systems (ICPADS) ,2011.
Future Work


Enhancements in real time scheduling for
Cloud and Big Data.
Energy efficient scheduling techniques for
vast datacenters i.e. Big Data.
Thank You !!!