Surplus Fair Scheduling

1
Surplus Fair Scheduling

• Ricardo A. Baratto
• COMS E6118 Spring 2004

2
Proportional-Share CPU Scheduling for
Symmetric Multiprocessors
3
What? Why?

• Fair resource allocation
• Proportional to application's share
• WFQ Demers et al., Parekh et al., Stride
  Waldspurger et al.
• SFQ Goyal et al.
• SMART Nieh et al.

4
The Problem

• Unfeasible Weight Assignment Starvation
• Upper Bound on Assignment
• Frequent Arrival Departures Unfairness
• Burst Behavior

5
Surplus Fair Scheduling

1. Weight Readjustment
2. Generalized Multiprocessor Sharing(Surplus
   Approximation)

6
Weight Readjustment

• Readjust threads violating feasibility
  constraint
• Optimal
• Number of threads
• Weight change
• Efficient
• O(p)

7
GMS and Surplus Approximation

• Generalization of GPS Parekh et al.
• fi Readjusted weight
• Ai(t1, t2) CPU service received
• Si Weighted CPU service Si q/fi
• ai Ai(0, t) - AiGMS(0, t) (fi . Si) - (fi
  . v)

8
SFS Algorithm

1. update Fi and compute Si
2. compute v min(Si)
3. compute ai
4. schedule ti with min(ai)
5. If t changes readjust weights

9
Implementation and Complexity

• Three ordered queues
• w Weight readjustment
• S Virtual time
• a Next to schedule
• Add thread Insert Readjust O(tp)
• Delete thread Delete Readjust O(p)
• Schedule Update surplus Sort O(t log t)
  every p schedules

10
Evaluation Weight Readjustment
11
Evaluation Short Jobs
12
Evaluation Application Isolation