Transcript Module 6: CPU Scheduling

Chapter 5: CPU Scheduling
Operating System Concepts – 8th Edition,
Silberschatz, Galvin and Gagne ©2009
Chapter 5: CPU Scheduling
 Basic Concepts
 Scheduling Criteria
 Scheduling Algorithms
 Thread Scheduling
 Multiple-Processor Scheduling
 Operating Systems Examples
 Algorithm Evaluation
Operating System Concepts – 8th Edition
5.2
Silberschatz, Galvin and Gagne ©2009
Objectives
 To introduce CPU scheduling, which is the basis for multiprogrammed
operating systems
 To describe various CPU-scheduling algorithms
 To discuss evaluation criteria for selecting a CPU-scheduling algorithm for a
particular system
Operating System Concepts – 8th Edition
5.3
Silberschatz, Galvin and Gagne ©2009
Basic Concepts
 Maximum CPU utilization obtained with multiprogramming
 CPU–I/O Burst Cycle – Process execution consists of a cycle of
CPU execution and I/O wait

While a process waits for I/O, CPU sits idle if no multiprogramming
 Instead the OS can give CPU to another process
 CPU burst distribution
Operating System Concepts – 8th Edition
5.4
Silberschatz, Galvin and Gagne ©2009
Alternating Sequence of CPU And I/O Bursts
Operating System Concepts – 8th Edition
5.5
Silberschatz, Galvin and Gagne ©2009
Histogram of CPU-burst Times
Operating System Concepts – 8th Edition
5.6
Silberschatz, Galvin and Gagne ©2009
CPU Scheduler
 Short-term Scheduler
 Selects from among the processes in memory that are ready to execute,
and allocates the CPU to one of them
 CPU scheduling decisions may take place when a process:
1. Switches from running to waiting state
2. Switches from running to ready state
3. Switches from waiting to ready
4. Terminates
 Scheduling under 1 and 4 is nonpreemptive/cooperative
 All other scheduling is preemptive
Operating System Concepts – 8th Edition
5.7
Silberschatz, Galvin and Gagne ©2009
CPU Scheduler
 Nonpreemptive: Once the process is allocated the CPU, it keeps it until
termination/wait.
eg. Windows 3.x/95
 No special hardware (like timers) needed.
 Preemptive scheduling – running process can be removed for another
 Issues: Shared data consistency – Synchronization (Ch. 6)
 What happens when the kernel is in a system call and the process asking
for that call is preempted? UNIX – context switches can only happen after
system calls. Other solutions – Sec 5.5, 19.5
 Typically we cannot disable interrupts
Operating System Concepts – 8th Edition
5.8
Silberschatz, Galvin and Gagne ©2009
Dispatcher
 Dispatcher module gives control of the CPU to the process
selected by the short-term scheduler; this involves:

switching context

switching to user mode

jumping to the proper location in the user program to restart
that program
 Dispatch latency – time it takes for the dispatcher to stop one
process and start another running
Operating System Concepts – 8th Edition
5.9
Silberschatz, Galvin and Gagne ©2009
Scheduling Criteria
 CPU utilization – keep the CPU as busy as possible
 Throughput – # of processes that complete their execution per
time unit
 Turnaround time – amount of time to execute a particular process
 Waiting time – amount of time a process has been waiting in the
ready queue
 Response time – amount of time it takes from when a request was
submitted until the first response is produced, not output (for timesharing environment)
Operating System Concepts – 8th Edition
5.10
Silberschatz, Galvin and Gagne ©2009
Scheduling Algorithm Optimization Criteria
 Max CPU utilization
 Max throughput
 Min turnaround time
 Min waiting time
 Min response time
Operating System Concepts – 8th Edition
5.11
Silberschatz, Galvin and Gagne ©2009
Scheduling Algorithms
 First-Come, First-Served Scheduling
 Shortest-Job-First Scheduling
 Priority Scheduling
 Round-Robin Scheduling
 Multilevel Queue Scheduling
 Multilevel Feedback Queue Scheduling
Operating System Concepts – 8th Edition
5.12
Silberschatz, Galvin and Gagne ©2009
First-Come, First-Served (FCFS) Scheduling
Process
Burst Time
P1
24
P2
3
P3
3
 Suppose that the processes arrive in the order: P1 , P2 , P3
The Gantt Chart for the schedule is:
P1
P2
0
24
P3
27
30
 Waiting time for P1 = 0; P2 = 24; P3 = 27
 Average waiting time: (0 + 24 + 27)/3 = 17
Operating System Concepts – 8th Edition
5.13
Silberschatz, Galvin and Gagne ©2009
FCFS Scheduling (Cont)
Suppose that the processes arrive in the order
P2 , P3 , P1
 The Gantt chart for the schedule is:
P2
P3
P1
0
3
6
 Waiting time for P1 = 6; P2 = 0; P3 = 3
30
 Average waiting time: (6 + 0 + 3)/3 = 3
 Much better than previous case
 Convoy effect short process behind long process – open CPU bound
process followed by multiple I/O processes.
 Nonpreemptive
Operating System Concepts – 8th Edition
5.14
Silberschatz, Galvin and Gagne ©2009
Shortest-Job-First (SJF) Scheduling
 Associate with each process the length of its next CPU burst. Use these
lengths to schedule the process with the shortest time
 If burst times are the same – break ties using FCFS
 SJF is provably optimal – gives minimum average waiting time for a given
set of processes

Reasoning – move the short process before a long one. This decreases
the waiting time of the short process more than it increases the waiting
time of the long one. Hence the average decreases.

The difficulty is knowing the length of the next CPU request
Operating System Concepts – 8th Edition
5.15
Silberschatz, Galvin and Gagne ©2009
Example of SJF
Process
Arrival Time
Burst Time
P1
0.0
6
P2
2.0
8
P3
4.0
7
P4
5.0
3
 SJF scheduling chart
P4
0
P3
P1
3
9
P2
16
24
 Average waiting time = (3 + 16 + 9 + 0) / 4 = 7
Operating System Concepts – 8th Edition
5.16
Silberschatz, Galvin and Gagne ©2009
Determining Length of Next CPU Burst
 Can only estimate the length
 Can be done by using the length of previous CPU bursts, using exponential
averaging
1. t n  actual length of n th CPU burst
2.  n 1  predicted value for the next CPU burst
3.  , 0    1
4. Define :
Operating System Concepts – 8th Edition
 n1   tn  1    n .
5.17
Silberschatz, Galvin and Gagne ©2009
Prediction of the Length of the Next CPU Burst
Operating System Concepts – 8th Edition
5.18
Silberschatz, Galvin and Gagne ©2009
Examples of Exponential Averaging
  =0
n+1 = n
 Recent history does not count

  =1
n+1 =  tn
 Only the actual last CPU burst counts
 If we expand the formula, we get:
n+1 =  tn+(1 - ) tn -1 + …

+(1 -  )j  tn -j + …
+(1 -  )n +1 0
 Since both  and (1 - ) are less than or equal to 1, each successive term
has less weight than its predecessor
 Can be preemptive or nonpreemptive
Operating System Concepts – 8th Edition
5.19
Silberschatz, Galvin and Gagne ©2009
Priority Scheduling
 A priority number (integer) is associated with each process
 The CPU is allocated to the process with the highest priority (smallest
integer  highest priority)

Preemptive

nonpreemptive
 SJF is a priority scheduling where priority is the predicted next CPU burst
time
 Problem  Starvation – low priority processes may never execute
 Solution  Aging – as time progresses increase the priority of the process
Operating System Concepts – 8th Edition
5.20
Silberschatz, Galvin and Gagne ©2009
Round Robin (RR)
 Each process gets a small unit of CPU time (time quantum),
usually 10-100 milliseconds. After this time has elapsed, the
process is preempted and added to the end of the ready queue.
 If there are n processes in the ready queue and the time
quantum is q, then each process gets 1/n of the CPU time in
chunks of at most q time units at once. No process waits more
than (n-1)q time units.
 Performance

q large  FIFO

q small  q must be large with respect to context switch,
otherwise overhead is too high. Also called processor
sharing – appears like each process has a processor
Operating System Concepts – 8th Edition
5.21
Silberschatz, Galvin and Gagne ©2009
Example of RR with Time Quantum = 4
Process
P1
P2
P3
Burst Time
24
3
3
 The Gantt chart is:
P1
0
P2
4
P3
7
P1
10
P1
14
P1
18 22
P1
26
P1
30
 Typically, higher average turnaround than SJF, but better response
 Avg Wait=17/3=5.66
Operating System Concepts – 8th Edition
5.22
Silberschatz, Galvin and Gagne ©2009
Time Quantum and Context Switch Time
Operating System Concepts – 8th Edition
5.23
Silberschatz, Galvin and Gagne ©2009
Turnaround Time Varies With The Time Quantum
Operating System Concepts – 8th Edition
5.24
Silberschatz, Galvin and Gagne ©2009
Multilevel Queue
 Ready queue is partitioned into separate queues:
foreground (interactive)
background (batch)
 Each queue has its own scheduling algorithm

foreground – RR

background – FCFS
 Scheduling must be done between the queues

Fixed priority scheduling; (i.e., serve all from foreground then from
background). Possibility of starvation.

Time slice – each queue gets a certain amount of CPU time which it can
schedule amongst its processes; i.e., 80% to foreground in RR

20% to background in FCFS
Operating System Concepts – 8th Edition
5.25
Silberschatz, Galvin and Gagne ©2009
Multilevel Queue Scheduling
Operating System Concepts – 8th Edition
5.26
Silberschatz, Galvin and Gagne ©2009
Multilevel Feedback Queue
 A process can move between the various queues; aging can be
implemented this way
 Multilevel-feedback-queue scheduler defined by the following
parameters:

number of queues

scheduling algorithms for each queue

method used to determine when to upgrade a process

method used to determine when to demote a process

method used to determine which queue a process will enter
when that process needs service
Operating System Concepts – 8th Edition
5.27
Silberschatz, Galvin and Gagne ©2009
Example of Multilevel Feedback Queue
 Three queues:

Q0 – RR with time quantum 8 milliseconds

Q1 – RR time quantum 16 milliseconds

Q2 – FCFS
 Scheduling

A new job enters queue Q0 which is served FCFS. When it gains CPU,
job receives 8 milliseconds. If it does not finish in 8 milliseconds, job is
moved to queue Q1.

At Q1 job is again served FCFS and receives 16 additional milliseconds.
If it still does not complete, it is preempted and moved to queue Q2.
Operating System Concepts – 8th Edition
5.28
Silberschatz, Galvin and Gagne ©2009
Multilevel Feedback Queues
Operating System Concepts – 8th Edition
5.29
Silberschatz, Galvin and Gagne ©2009
Thread Scheduling
 Distinction between user-level and kernel-level threads
 OS only schedules kernel-level threads. User-level threads are
scheduled through a direct or indirect (LWP) mapping
 Many-to-one and many-to-many models, thread library schedules
user-level threads to run on LWP

Known as process-contention scope (PCS) since scheduling
competition is within the process
 Kernel thread scheduled onto available CPU is system-contention
scope (SCS) – competition among all threads in system
 Typically – PCS is priority based. Programmer can set user-level
thread priorities
Operating System Concepts – 8th Edition
5.30
Silberschatz, Galvin and Gagne ©2009
Pthread Scheduling
 API allows specifying either PCS or SCS during thread creation

PTHREAD_SCOPE_PROCESS schedules threads using PCS
scheduling

PTHREAD_SCOPE_SYSTEM schedules threads using SCS
scheduling.
 2 methods to get and set the scope

pthread_attr_setscope(pthread_attr_t *attr, int scope)

pthread_attr_getscope(pthread_attr_t *attr, int *scope)
 attr – pointer to the attribute set
 scope – PSC/SCS
Operating System Concepts – 8th Edition
5.31
Silberschatz, Galvin and Gagne ©2009
Pthread Scheduling API
#include <pthread.h>
#include <stdio.h>
#define NUM THREADS 5
int main(int argc, char *argv[])
{
int i;
pthread_t tid[NUM THREADS];
pthread_attr_t attr;
/* get the default attributes */
pthread_attr init(&attr);
/* set the scheduling algorithm to PROCESS or SYSTEM */
pthread_attr setscope(&attr, PTHREAD_SCOPE_SYSTEM);
/* set the scheduling policy - FIFO, RT, or OTHER */
pthread_attr setschedpolicy(&attr, SCHED_OTHER);
/* create the threads */
for (i = 0; i < NUM THREADS; i++)
pthread create(&tid[i],&attr,runner,NULL);
Operating System Concepts – 8th Edition
5.32
Silberschatz, Galvin and Gagne ©2009
Pthread Scheduling API
/* now join on each thread */
for (i = 0; i < NUM THREADS; i++)
pthread join(tid[i], NULL);
}
/* Each thread will begin control in this function */
void *runner(void *param)
{
printf("I am a thread\n");
pthread exit(0);
}
Operating System Concepts – 8th Edition
5.33
Silberschatz, Galvin and Gagne ©2009
Multiple-Processor Scheduling
 CPU scheduling more complex when multiple CPUs are
available
 ASSUMPTION - Homogeneous processors within a
multiprocessor
 Asymmetric multiprocessing – only one processor
accesses the system data structures, alleviating the need
for data sharing
Problems – Bottleneck, Single point of failure
 Symmetric multiprocessing (SMP) – each processor
is self-scheduling, all processes in common ready queue,
or each has its own private queue of ready processes\
 Most common – Windows XP, 2000, Linux, OS X
 Remainder of the discussion applies to SMP.
Operating System Concepts – 8th Edition
5.34
Silberschatz, Galvin and Gagne ©2009
Multiprocessor Scheduling
 Processor affinity – process has affinity for processor on which it is
currently running
 Reason – caching. As a process runs, the cache gets populated and it is
increasingly likely that the requests will be satisfied from the cache.

soft affinity
OS tries to keep the process running on the same processor, but this is
not binding.
Migration possible.

hard affinity
Supported in Linux. Allows a process to specify this.
 Solaris supports the creation of processor sets. Also soft but somewhat
more restricted.
Operating System Concepts – 8th Edition
5.35
Silberschatz, Galvin and Gagne ©2009
Multiprocessor Scheduling
 Affinity may be decided by the architecture of the main-memory.
 NUMA – Non Uniform Memory Access
 CPU has faster access to some memory.
 Multiprocessors systems where each CPU has a memory board.
 It can also access memory on other CPU’s but there is a delay
 OS design influenced by the architecture and optimized for performance
Operating System Concepts – 8th Edition
5.36
Silberschatz, Galvin and Gagne ©2009
NUMA and CPU Scheduling
Operating System Concepts – 8th Edition
5.37
Silberschatz, Galvin and Gagne ©2009
Load Balancing
 Goal: Keep workload evenly distributed between processors
 Usually only necessary if each processor has its own individual queue
 If there is a common queue, a processor can just pick a job from here when
free
 Push migration – specific task to check load on each processor and
redistribute if needed.
 Pull migration – idle processor pulls task form a busy one.
 Usually both are implemented in parallel. Eg. Linux scheduler
 Note that this conteracts with affinity.
Operating System Concepts – 8th Edition
5.38
Silberschatz, Galvin and Gagne ©2009
Multicore Processors
 Recent trend to place multiple processor cores on same physical chip
 Each core has its own register set
 Faster and consume less power
 Multiple threads per core also growing

Takes advantage of memory stall to make progress on another thread
while memory retrieve happens
 Memory stalls – can be upto 50% of the time
 Solution – Multithreaded Multicores
Operating System Concepts – 8th Edition
5.39
Silberschatz, Galvin and Gagne ©2009
Multithreaded Multicore System
Operating System Concepts – 8th Edition
5.40
Silberschatz, Galvin and Gagne ©2009
Multithreaded Multicore
 Coarse-grained
More cost with switching between threads
 Fine-grained
Much finer level of granularity in switching between threads – logic for
thread switching included in architecture
 2 levels of scheduling are happening here.
Operating System Concepts – 8th Edition
5.41
Silberschatz, Galvin and Gagne ©2009
Operating System Examples
 Solaris scheduling
 Windows XP scheduling
 Linux scheduling
Operating System Concepts – 8th Edition
5.42
Silberschatz, Galvin and Gagne ©2009
Solaris Scheduling
 Priority based thread scheduling with 6 classes:

Time sharing

Interactive

Real Time

System

Fair Share

Fixed Priority
 Default class for a process is Time Sharing
 TS – dynamic priorities and slice lengths using a multilevel queue
Eg. Shown for different priorities
Operating System Concepts – 8th Edition
5.43
Silberschatz, Galvin and Gagne ©2009
Solaris Dispatch Table
Operating System Concepts – 8th Edition
5.44
Silberschatz, Galvin and Gagne ©2009
Solaris Dispatch Table
 Priority – Higher number, Higher Priority
 Quantum length inversely proportional to Priority
 Time Quantum expired – new priority for thread that has used its entire
quantum without blocking (CPU intensive threads)
 Return from sleep – Priority of a thread returning from a sleep state eg.
Waiting for I/O. When I/O is available its priority is boosted
Operating System Concepts – 8th Edition
5.45
Silberschatz, Galvin and Gagne ©2009
Solaris Scheduling
Operating System Concepts – 8th Edition
5.46
Silberschatz, Galvin and Gagne ©2009
Windows XP Scheduling
 32-level priority scheme to determine order of execution
 Split into two classes
 Variable class – 1-15
 Real-time class – 16-31
 Several Priority classes in the API, followed by relative priority within a class
 Variable priorities
 Typically priority of the foreground process is increased – usually by 3
Operating System Concepts – 8th Edition
5.47
Silberschatz, Galvin and Gagne ©2009
Windows XP Priorities
Operating System Concepts – 8th Edition
5.48
Silberschatz, Galvin and Gagne ©2009
Linux Scheduling
 Constant order O(1) scheduling time
 Regardless of number of tasks
 Two priority ranges: time-sharing(nice) and real-time
 Real-time range from 0 to 99 and nice value from 100 to 140
 Lower values -> Higher priorities
 Unlike Solaris, Higher priority is given Larger Time slice
Operating System Concepts – 8th Edition
5.49
Silberschatz, Galvin and Gagne ©2009
Priorities and Time-slice length
Operating System Concepts – 8th Edition
5.50
Silberschatz, Galvin and Gagne ©2009
List of Tasks Indexed According to Priorities
Operating System Concepts – 8th Edition
5.51
Silberschatz, Galvin and Gagne ©2009
Algorithm Evaluation
 Deterministic modeling – takes a particular
predetermined workload and defines the performance of
each algorithm for that workload eg. Calculating the
average wait time for each model
 Queueing models
 Implementation
Operating System Concepts – 8th Edition
5.52
Silberschatz, Galvin and Gagne ©2009
Analytical Evaluation
Operating System Concepts – 8th Edition
Processes
Burst Time
P1
10
P2
29
P3
3
P4
7
P5
12
5.53
Silberschatz, Galvin and Gagne ©2009
FCFS Wait time=(0+10+39+42+49)/5=28ms
SJF Wait time=(10+32+0+3+20)/5=13 ms
RR Wait time=(0+32+20+23+40)/5=23 ms
Operating System Concepts – 8th Edition
5.54
Silberschatz, Galvin and Gagne ©2009
Queuing Network Analysis
 Processes are dynamic and cannot be estimated
 But CPU and I/O burst distributions can be
 Formula estimating the probability of a particular burst
 Similarly arrival times can be shown by a distribution
 Given these two distributions, possible to compute avg throughput,
utilization, waiting time etc.
 Let n : avg. queue length
 W : avg wait time in the queue
  : Avg Arrival Rate

n=xW
Little’s Formula
 If the system is steady, number entering must be equal to number leaving
Operating System Concepts – 8th Edition
5.55
Silberschatz, Galvin and Gagne ©2009
Evaluation of CPU schedulers by Simulation
Operating System Concepts – 8th Edition
5.56
Silberschatz, Galvin and Gagne ©2009
End of Chapter 5
Operating System Concepts – 8th Edition,
Silberschatz, Galvin and Gagne ©2009