Download Chapter 9

Announcement
• Midterm Exam is on Monday June 8th.
• Office hours on Friday 2-3 pm.
• Exam Material will be posted on Piazza.
Operating
Systems:
Internals
and
Design
Principles
Chapter 9
Uniprocessor Scheduling
Seventh Edition
By William Stallings
Modified by Rana Forsati for CSE 410
Learning Objective
• Identify different goals of process scheduling algorithms
• Identify different types of scheduling and their role
• Given a set of processes, evaluate them to identify the
next process to run
• Identify effect of different scheduling algorithms on CPU
bound/I/O bound processes
Scheduling and Process State Transitions
Types of Scheduling
■
Broken down into three separate functions:
long term
scheduling
medium term
scheduling
short term
scheduling
Long-Term Scheduling
■
Determines when to create a new process and which
programs are admitted next to the system for processing
■
Once admitted a user program becomes a process
and is added to the queue for short term scheduler
■
Controls the degree of multiprogramming
■
The more processes that are created, the smaller the
percentage of time that each process can be executed
Medium-Term Scheduling
■
Part of the swapping function
■
Swapping-in decisions are based on the need to
manage the degree of multiprogramming
■ Considers the memory requirements of
the swapped-out processes
Short-Term Scheduling
■
Known as the dispatcher
■
The main objective of short-term scheduling is to allocate
CPU time to processes
■
Executes most frequently [minimum overhead?]
■
Invoked when an event occurs that may lead to the
blocking of the current process or that may provide an
opportunity to preempt a currently running process in favor
of another
- Clock interrupts
– I/O interrupts
– Operating system calls
Short-term success criteria?
Short Term
Scheduling Algorithms
9
Aim of Short-term-Scheduling
■
Assign processes to be executed by the processor in a way that meets system
(performance) objectives:
■
Response time: amount of time it takes from when a request was submitted
until the first response is produced, not output (for time-sharing
environment)
■
■
Throughput: # of processes that complete their execution per time unit
■
■
Minimize response time
Maximize throughput
Processor efficiency: keep the CPU as busy as possible
■
Maximize CPU utilization
Short-Term Scheduling
Dispatcher
OR
short-term scheduler
■
Short-Term Scheduling: Selects from among the processes in memory that are ready to execute,
and allocates the CPU to one of them
Scheduling Objectives
The scheduling function should:
– Share time fairly among processes
– Prevent starvation of processes
– Have low overhead
– Prioritize processes when necessary [flexible]
Scheduling Policies
• First-Come-First-Served (FCFS)
• Round-Robin (RR)
• Shortest-Process-Next (SPN)
• Shortest-Remaining Time (SRT)
• Highest Response Ratio Next (HRRN)
Selection Function
■
Determines which process, among ready processes, is
selected next for execution
■
This function may be based on priority, resource
requirements, or the execution characteristics of the
process:
▪ time spent in system so far, waiting
▪ time spent in execution so far
▪ total service time required by the process [generally,
this quantity must be estimated or supplied by the user]
Decision Mode
▪ Two categories:
▪Nonpreemptive
✴Once a process is in the running state, it will continue to execute
until
• it terminates OR
• it blocks itself for I/O or some OS service
▪Preemptive
✴ Currently running process may be interrupted and moved to the
Ready state by the operating system.
• a new process arrives
• interrupt occurs
• periodically based on a clock interrupt
Process Scheduling Example
Whatʼs the execution order?
First-Come-First-Served
(FCFS)
• Each process joins the Ready queue
• When the current process ceases to execute, the oldest process in
the Ready queue is selected
A
B
C
D
E
What are the potential consequences of this algorithm?
First-Come-First-Served
(FCFS)
• Favors long processes: performs much better for long
processes than short ones
• A short process may have to wait a very long
time before it can execute
• Favors CPU-bound processes (mostly uses the
processor)
• I/O processes have to wait until CPU-bound
process completes
First-Come-First-Served
(FCFS)
First-Come-First-Served
(FCFS)
Process
Service Time (ms)
P1
24
P2
3
P3
3
• Suppose that the processes arrive in the order: P1 , P2 , P3
P1
P2
P3
0
24
27
• Waiting time for P1 = 0; P2 = 24; P3 = 27
• Average waiting time: (0 + 24 + 27)/3 = 17 ms
30
First-Come-First-Served
(FCFS)
Suppose that the processes arrive in the order
P2 , P3 , P1
P2
0
P3
3
P1
6
• Waiting time for P1 = 6; P2 = 0; P3 = 3
• Average waiting time: (6 + 0 + 3)/3 = 3 ms
• Much better than previous case
• Convoy effect: short process behind long process
30
Round Robin
• Clock interrupt is generated at periodic intervals
• When an interrupt occurs, the currently running process
is placed in the ready queue
• Next ready job is selected
• This technique is also known as time slicing ,because each
process is given a slice of time before being preempted.
• Principal design issue is the length of the time quantum, or
slice, to be used [too large —> ?, too short —> ?]
Round Robin
• Each process gets a small unit of CPU time known as time
slicing (time quantum).
• After this time has elapsed, the process is preempted and added
to the end of the Ready queue.
Process
P1
P2
P3
P1
0
P2
4
Service Time
24
3
3
P3
7
P1
10
P1
14
P1
18 22
P1
26
P1
30
• An amount of time is
determined that allows
each process to use the
processor for that length
of time q=4
Round Robin
q=4
A
B
C
D
E
Whatʼs the execution order?
Round Robin, q = 4
Round Robin, q =1
Round Robin
Example
Finish time of each process?
a) Round Robin q=30
b) Round Robin q=10
Example
Solution
A
B
C
D
E
Shortest Process Next
(SPN)
•
Another approach to reducing the bias in favor of long processes
inherent in FCFS
• Associate with each process the length of its next CPU service.
Use these lengths to schedule the process with the shortest time
■
Process with the shortest expected processing time is selected next
■
A short process jumps ahead of longer processes
■
Nonpreemptive policy
Shortest Process Next
(SPN)- Issues
■
The difficulty is knowing or estimating the length of the required
processing time of each process.
■
A risk with SPN is the possibility of starvation for longer
processes, as long as there is a steady supply of shorter processes.
■
It still is not desirable for a time-sharing or transaction processing
environment because of the lack of preemption.
Shortest Process Next
(SPN)
Process Arrival Time
P1
0.0
Service Time
6
P2
0.0
8
P3
0.0
7
P4
0.0
3
• SPN:
P4
0
P3
P1
3
9
P2
16
• Average waiting time = (3 + 16 + 9 + 0) / 4 = 7 ms
24
Shortest Process Next
(SPN)
A
B
C
D
E
Whatʼs the execution order?
Shortest Process Next
(SPN)
Shortest Process Next
(SPN)
Shortest Process Next
(SPN)- Service Estimation
■
One difficulty is the need to know, or at least estimate, the
required processing time of each process (Predictability of longer
processes is reduced)
■
Possibility of starvation for longer processes
■
If estimated time for process not correct, the operating system may
abort it
• Can be done by using the length of previous CPU services, using
exponential averaging
Shortest Remaining Time (SRT)
■
Preemptive version of SPN
■
Scheduler always chooses the process that has
the shortest expected remaining processing
time
• While a process A is running, if a new process
B comes whose length is shorter than the
remaining time of process A, then B preempts
A and B is started to run.
■
Must estimate processing time
A
B
C
D
E
Whatʼs the execution order?
Shortest Remaining Time (SRT)
■
Should give superior turnaround time performance to SPN because a short job is given
immediate preference to a running longer job
■
Risk of starvation of longer processes
Shortest Remaining Time (SRT)
•
Process
P1
Arrival Time
0.0
Burst Time
8
P2
1.0
4
P3
2.0
9
P4
3.0
5
SRT:
P1
0
•
P2
1
P4
5
P1
10
P3
17
Average waiting time = (9 + 0 + 2 + 15) / 4 = 6.5 ms
26
Highest Response Ratio Next
(HRRN)
■
Chooses next process with the
greatest value of response ratio
■
Attractive because it accounts for
the age of the process
■
While shorter jobs are favored,
aging without service increases
the ratio so that a longer process
will eventually get past competing
shorter jobs
Highest Response Ratio Next
(HRRN)
The algorithms with CPU-I/O cycle
• The previous examples included processes with just a single CPU
service.
Each process execution is a CPU-I/O cycle
– First, it executes on CPU
– Then, it waits for I/O
– After completion of I/O, executes on CPU again and so on.
– It ends with execution on CPU.
• Example
– <3, 2, 1, 4, 5>
• CPU for 3 time units then I/O for 2, …
• We will make the following assumptions while considering these
algorithms with I/O services as well.
• In this class, we will assume that all I/O is concurrent.
The algorithms with CPU-I/O cycle
FCFS:
– A process that finishes its I/O enters at the end of the ready queue. No other
change required.
Round Robin:
– Same as above
– We will assume the process enters the ready queue as soon as its I/O is finished.
– If a process finishes its CPU service, another process can start executing
immediately.
– No history is maintained
• E.g., if the time quantum is 10 and the process CPU service is 4 then the process
will release the CPU after time 4.
– After this process releases the CPU, another process can start right away and will
get time quantum of 10
– When the first process comes back again, it will get the entire time quantum of
10.
The algorithms with CPU-I/O cycle
• SPN and SRT
– When a process is being considered, we will focus on the
next CPU service to decide who runs first
• HRRN
– We will consider total service time and total waiting time
in deciding the process to be chosen
First-Come-First-Served
(FCFS)
Consider a system where there are 3 processes, A, B, and C. Their details are as follows
A: Arrival time = 0, CPU-I/O cycle = <4,3,2,6,6>
B: Arrival time = 2, CPU-I/O cycle = <2,5,4,5,4>
C: Arrival time = 4, CPU I/O cycle = <5,2,3,3,5>
Show the execution with FCFS:
Feedback
Scheduling
• Penalize jobs that have
been running longer
• Donʼt know remaining
time process needs to
• execute
• Preemptive with time
quantum
• Demoted to the next
lower-priority queue
• With each queue
(except the lowest
priority queue), FCFS
• Lowest-priority queue:
RR
Traditional UNIX Scheduling
• Multilevel feedback using round robin within each of
the priority queues
• Priorities are recomputed once per second
• Base priority divides all processes into fixed bands of
priority levels
• Adjustment factor used to keep process in its assigned
band (decreasing order of priority):
– Swapper
– Block I/O device control
– File manipulation
– Character I/O device control
– User processes
Summary
■
The operating system must make three types of scheduling decisions with respect to the
execution of processes:
■ Long-term – determines when new processes are admitted to the system
■ Medium-term – part of the swapping function and determines when a program is
brought into main memory so that it may be executed
■ Short-term – determines which ready process will be executed next by the
processor
■
From a user’s point of view, response time is generally the most important characteristic
of a system; from a system point of view, throughput or processor utilization is
important
■
Algorithms:
■
FCFS, Round Robin, SPN, SRT, HRRN, Feedback