Ch05 CPU Scheduling
Transcript of Ch05 CPU Scheduling
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 1/45
Course 5: CPU Scheduling
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 2/45
5.2 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Course 4: Review
Thread – single, sequential stream of execution
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 3/45
5.3 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Course 4: Review (Cont.)
Responsiveness – e.g. web browser, word processor etc
Resource Sharing – e.g. IPC issues
Economy (vs. process model) – e.g. Solaris – thread creation – 30times faster, thread context switch 5 times faster
Scalability – e.g. utilization of MP architectures
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 4/45
5.4 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Course 5: CPU Scheduling
Basic Concepts
Scheduling Algorithms
Multiple-Processor Scheduling
Operating Systems Examples
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 5/45
5.5 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Scheduling Basic
Concepts
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 6/45
5.6 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Basic Concepts
Process scheduling aka kernel-
level thread scheduling inthreaded systems
Scheduling may apply to all thecomputer resources
CPU scheduling:
Maximum CPU utilizationobtained withmultiprogramming
CPU –I/O Burst Cycle –
Process execution consists ofa cycle of CPU execution andI/O wait
CPU burst distribution
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 7/455.7 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Histogram of CPU-burst Times
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 8/455.8 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
CPU Scheduler
Selects from among the processes in memory that are ready toexecute (PCB queues), 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 (interrupt)
3. Switches from waiting to ready
4. Terminates
Scheduling under 1 and 4 is nonpreemptive/cooperative
All other scheduling is preemptive
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 9/455.9 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Dispatcher
Dispatcher module gives control of the CPU to the processselected 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 oneprocess and start another running
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 10/455.10 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Scheduling Criteria
CPU utilization – keep the CPU as busy as possible [40lightly loaded, 90 heavily used]
Throughput – # of processes that complete their executionper time unit
Turnaround time – amount of time to execute a particularprocess
Waiting time – amount of time a process has been waitingin the ready queue
Response time – amount of time it takes from when arequest was submitted until the first response is produced,not output (for time-sharing environment)
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 11/455.11 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Optimization Criteria
General
Max CPU utilization
Max throughput
Min turnaround time
Min waiting time Min response time
Contextual:
Interactive systems – minimize variance
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 12/455.12 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Scheduling Algorithms
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 13/455.13 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
First-Come, First-Served (FCFS) Scheduling
Process Burst TimeP 1 24
P 2 3
P 3 3
Suppose that the processes arrive in the order: P 1 , P 2 , P 3
The Gantt Chart for the schedule is:
Waiting time for P 1 = 0; P 2 = 24; P 3 = 27
Average waiting time: (0 + 24 + 27)/3 = 17
P1 P2 P3
24 27 300
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 14/455.14 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
FCFS Scheduling (Cont.)
Suppose that the processes arrive in the order
P 2 , P 3 , P 1
The Gantt chart for the schedule is:
Waiting time for P 1 = 6; P 2 = 0; P 3 = 3
Average waiting time: (6 + 0 + 3)/3 = 3
Much better than previous case
Convoy effect short process behind long process
FCFS – not optimal for time-sharing systems
P1P3P2
63 300
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 15/455.15 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Shortest-Job-First (SJF) Scheduling
Associate with each process the length of its next CPU burst. Usethese lengths to schedule the process with the shortest time
Two schemes:
nonpreemptive – once CPU given to the process it cannot bepreempted until completes its CPU burst
preemptive –
if a new process arrives with CPU burst lengthless than remaining time of current executing process,preempt. This scheme is know as theShortest-Remaining-Time-First (SRTF)
SJF is optimal – gives minimum average waiting time for a givenset of processes
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 16/455.16 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Process Arrival Time Burst Time
P 1 0.0 7
P 2 2.0 4
P 3 4.0 1
P 4 5.0 4
SJF (non-preemptive)
Average waiting time = (0 + 6 + 3 + 7)/4 = 4
Example of Non-Preemptive SJF
P1 P3 P2
73 160
P4
8 12
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 17/455.17 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Example of Preemptive SJF
Process Arrival Time Burst Time
P 1 0.0 7
P 2 2.0 4
P 3 4.0 1
P 4 5.0 4
SJF (preemptive)
Average waiting time = (9 + 1 + 0 +2)/4 = 3
P1 P3P2
4211
0
P4
5 7
P2 P1
16
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 18/455.18 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Determining Length of Next CPU Burst
Can only estimate the length
Can be done by using the length of previous CPU bursts, usingexponential averaging
:Define 4.
10, 3.burst CPUnextthefor valuepredicted 2.
burst CPUof length actual 1.
1n
th
n nt
.11 nnn
t
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 19/455.19 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Prediction of the Length of the Next CPU Burst
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 20/455.20 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Examples of Exponential Averaging
=0
n+1 = n
Recent history does not count
=1
n+1 = t n
Only the actual last CPU burst counts If we expand the formula, we get:
n+1 = tn+(1 - ) t n -1 + …
+( 1 - ) j t n - j + …
+( 1 - )n +10
Since both and (1 - ) are less than or equal to 1, eachsuccessive term has less weight than its predecessor
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 21/455.21 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Priority Scheduling
A priority number (integer) is associated with each process
Priority may be calculated considering: time limits, memoryrequirements, no. of open files, avg. I/O burst/avg. CPU burst
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 CPUburst time
Problem Starvation – low priority processes may never execute
Solution Aging – as time progresses increase the priority of theprocess
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 22/455.22 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Round Robin (RR)
Each process gets a small unit of CPU time (time quantum),usually 10-100 milliseconds. After this time has elapsed, theprocess is preempted and added to the end of the ready queue.
If there are n processes in the ready queue and the timequantum is q, then each process gets 1/ n of the CPU time inchunks 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
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 23/45
5.23 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Example of RR with Time Quantum = 20
Process Burst Time
P 1 53
P 2 17
P 3 68
P 4 24
The Gantt chart is:
Typically, higher average turnaround than SJF, but better response
P1 P2 P3 P4 P1 P3 P4 P1 P3 P3
0 20 37 57 77 97 117 121 134 154 162
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 24/45
5.24 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Time Quantum and Context Switch Time
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 25/45
5.25 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Turnaround Time Varies With The Time Quantum
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 26/45
5.26 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
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 thenfrom 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% toforeground in RR
20% to background in FCFS
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 27/45
5.27 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multilevel Queue Scheduling
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 28/45
5.28 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multilevel Feedback Queue
A process can move between the various queues; aging can beimplemented this way
Multilevel-feedback-queue scheduler defined by the followingparameters:
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 enterwhen that process needs service
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 29/45
5.29 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
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 RR. When it gainsCPU, job receives 8 milliseconds. If it does not finish in 8milliseconds, job is moved to queue Q1.
At Q1 job is again served RR and receives 16 additionalmilliseconds. If it still does not complete, it is preempted and
moved to queue Q2.
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 30/45
5.30 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multilevel Feedback Queues
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 31/45
5.31 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multiple-Processor Scheduling
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 32/45
5.32 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multiple-Processor Scheduling
CPU scheduling more complex when multiple CPUs areavailable
Asymmetric multiprocessing – only one processoraccesses the system data structures, alleviating the needfor data sharing
Symmetric multiprocessing - each processor is self-scheduled
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 33/45
5.33 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Processor affinity
Most recent data in cache If a process is switched from one processor to
another, there is a cost of invalidating andrepopulating caches
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 34/45
5.34 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Load balancing
Workload evenly distributed across all processors
Necessary on systems with multiple queues
Affects processor affinity rule
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 35/45
5.35 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Multicore processors
Multiple processors on the same chip
Memory stall = time waiting for data to be available whenaccessing memory
In order to optimize memory stall => multithreaded cores,each hardware thread appearing as a logical processor
Two levels of scheduling – OS, hardware
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 36/45
5.36 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Operating System Examples
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 37/45
5.37 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Operating System Examples
Solaris scheduling Windows XP scheduling
Linux scheduling
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 38/45
5.38 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Solaris Scheduling
6 priority classes, default TS
Inverse relation btw priorities andtime slices
Interactive processes – higherpriority, CPU-bound processeslower priority
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 39/45
5.39 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Solaris Dispatch Table
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 40/45
5.40 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Windows XP Priorities
Preemptive
6 priority classes / relative priority in each class
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 41/45
5.41 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Linux Scheduling
Two algorithms: time-sharing and real-time Time-sharing
Prioritized credit-based – process with most credits isscheduled next
Credit subtracted when timer interrupt occurs
When credit = 0, another process chosen When all processes have credit = 0, recrediting occurs
Based on factors including priority and history
Real-time
Soft real-time
Posix.1b compliant – two classes FCFS and RR
Highest priority process always runs first
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 42/45
5.42 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
The Relationship Between Priorities and Time-slice length
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 43/45
5.43 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
Algorithm Evaluation
Deterministic modeling –
takes a particularpredetermined workload and defines the performance ofeach algorithm for that workload
Queueing models
Implementation
8/10/2019 Ch05 CPU Scheduling
http://slidepdf.com/reader/full/ch05-cpu-scheduling 44/45
5.44 Silberschatz, Galvin and GagneOperating System Concepts – 8th Edition
5.15