Shital K Dhamal 6 1/0 management 4 611, 10 devices, Organization ofthe 10 Function, Disk Organzation, 10 Management and Disk Scheduling: FCFS, SSTF, SCAN, CSCAN, LOOK, C-LOOK. 1/0 hardware One of the important jobs of an Operating System is to manage various 1/0 devices including mouse, keyboards, touch pad, disk drives, display adapters, USB devices, Bit-mapped screen, LED, Analog-to-digital converter, On/off'switch, network connections, audio VO, printers et. An I/O system is required to take an application /O request and send it to the physical device, then take whatever response comes back from the device and send it to the application. /O devices can be divided into two categories — + Block devices ~ A block device is one with which the driver communicates by sending entire blocks of data, For example, Hard disks, USB cameras, Disk-On-Key ete. + Character devices — A character device is one with which the driver communicates by sending and receiving single characters (bytes, octets). For example, serial ports, parallel ports, sounds cards ete Device Controllers «Device drivers are software modules that can be plugged info an OS to handle a particular device. Operating System takes help from device drivers to handle all YO devices + The Device Controller works like an interface between a device and a device driver. VO units (Keyboard, mouse, printer, ete.) typically consist of a mechanical component and an electronic component where electronic component is called the device controller: + There is always a device controller and a device driver for each device to commmnicate with the Operating Systems. A device controller may be able to handle multiple devices. As an interface its main task is to convert serial bit stream to block of bytes, perform error correction as necessary.Shital K Dhamal + Any device connected to the computer is connected by a plug and socket, and the socket is connected to a device controller. Following is a model for comecting the CPU, memory, controllers, and VO devices where CPU and device controllers all use a common bus for communication. Memory Monitor Keyboard USB Drive Disk Drive cu Memory Video Keyboard uss Disk | | Controller Synchronous vs asynchronous JO + Synchronous I/O ~ In this scheme CPU execution waits while 1/0 proceeds + Asynchronous 1/0 ~ 1/0 proceeds concurrently with CPU execution Q Explain Infermupt-driven IO and discuss the advantages of Interrupt-driven IO over programmed 10. Mode of Transfer: Data transfer between the central processing unit(CPU) and VO devices can be handled in generally three types of modes which are given below: 1. Programmed 1/0 (Special Instruction 1/0) 2. Interrupt- initiated 1/0 or Memory mapped 1/0 3. Direct memory access (DMA). 1. Programmed /O This V/O technique is the simplest to exchange data between extemal devices and processors In this technique, the processor or Central Processing Unit (CPU) rums or executes a program giving direct control of VO operations.Shital K Dhamal Processor issues a command to the YO module and waits for the operation to complete. Also, the processor keeps checking the /O module status until it finds the completion of the operation. The processor's time is wasted, in case the processor is faster than the /O module. Its module is considered to be a slow module. This uses CPU instructions that are specifically made for controlling VO devices. These instructions typically allow data to be sent to an VO device or read from an VO device. Egs Prinf(), scant) 2. Interrupt initiated /O or Memory Mapped I/O- Itis similar to the programmed-driven VO technique. The processor does uot wait until the /O operation is completed. The processor performs other tasks while the /O operation is being performed. When the VO operation is completed, the VO module interrupts the processor letting the processor know the operation is completed. Its module is faster than the programmed /O module. The processor starts the /O device and instructs it to generate and send an interrupt signal when the operation is finished. This technique requires an interrupt for each character that is written orread. + The same address space is shared by memory and /O devices. + The device is connected directly to certain main memory locations so that 1] device can transfer blocks of data to/from memory (rather than real memory) without going trough CPU. + To transfer data from the device to an actual memory, the CPU has to read the data from the memory-mapped device registers (buffer) aud write it to the actual memory Memory Mapped !oShital K Dhamal 3 Direct Memory Access Slow devices like keyboards will generate an interrupt to the main CPU after each byte is transfered. If a fast device such as a disk generated an interupt for each byte, the operating system would spend most of its time handling these interrupts. So, a typical computer uses direct memory access (DMA) hardware to reduce this overhead. Direct Memory Access (DMA) means CPU grants /0 module authority to read from or write to memory without involvement. DMA module itself controls the exchange of data between the main memory and the I/O device. CPU is only involved at the beginning and end of the transfer and interrupted only afler the entire block has been transferred. Direct Memory Access needs special hardware called DMA controller (DMAC) that manages the data transfers and arbitrates access to the system bus. The controllers are programmed with, source and destination pointers (where to read/write the data), counters to track the number of transferred bytes, and settings, which include /O and memory types, interrupts, and states for the CPU eyeles. What is Direct Memory Access DATA DMA Controller | DMA (pmac) tte haa MemoryShital K Dhamal Data Bus Device Controller USB Drive Disk Printer DMA Contoller Disk Contoller Taney (Tay Count) ) Count (owe | Display Mestage - Data ‘Byte_count = Othen there Isno more data “Transfers completed wil repeat the steps What is Direct Memory Access ek nm EShital K Dhamal The operating system uses the DMA hardware as follows — 1 2. 3. 4. 5 Device driver is instructed to transfer disk data to a buffer address X. Device driver then instructs the disk controller to transfer data to the buffer. Disk controller starts DMA transfer. . Disk controller sends each byte to DMA controller. DMA controller transfers bytes to the buffer, increases the memory address, and decreases the counter C until C becomes zero. When C becomes zero, DMA interrupts CPU to signal transfer completion. Hard Disk Organization The disk is divided into tracks. Each track is further divided into sectors. The point to be noted here is that outer tracks are bigger in size than inner tracks but they contain the same number of sectors and have equal storage capacity. Read-Write(R-W) head moves over the rotating hard disk. It is this Read-Write head that performs all the read and write operations on the disk To perform a read or write operation ona memory location, we need to place the R-W head over that position. Some important terms must be noted here:Shital K Dhamal Platters -> Surface -> Track -> Fix number of Sectors Data is stored in sectors = Disk Size = Platters * Surface * Tracks * Sectors * Data Disk Size = 8* 2 * 256 * 512 * 512 KB Seek Time :-Time taken to reach up to desired track nesd-Witevead G02! = To minimize Seek Time ‘etuators Atm Movement -Left orRisht Disk Scheduling Algorithi rd Disk Architecture ee Disk Scheduling As we know, a process needs two type of time, CPU time and IO time. For VO, it requests the Operating system to access the disk. However, the operating system must be fair enough to satisfy each request and at the same time, operating system must maintain the efficiency and speed of process execution. The technique that operating system uses to determine the request which is to be satisfied next is called disk scheduling. Important terms related to disk scheduling. Seek Time Seek time is the time taken to locate the disk arm to a specified track where the read/write request will be satisfied. Rotational Latency Its the time taken by the desired sector to rotate itself to the position from where it can access the R/W heads.Shital K Dhamal ‘Transfer Time Its the time taken to transfer the data, Disk Access Time Disk access time is given as, Disk Access Time = Rotational Latency + Seek Time + Transfer Time Disk Response Time Itis the average of time spent by each request waiting for the IO operation. Purpose of Disk Scheduling The main purpose of disk scheduling algorithm is to select a disk request from the queue of 10 requests and decide the schedule when this request will be processed. Goal of Disk Scheduling Algorithm © Faimess © High throughout © Minimal traveling head time Disk Scheduling Algorithms The list of various disk scheduling algorithms is given below. Each algorithm carries some advantages and disadvantages. The limitation of each algorithm leads to the evolution of a new algorithm. © FCFS scheduling algorithm © SSTF (shortest seek time first) algorithm © SCAN scheduling © C-SCAN scheduling © LOOK Scheduling © C-LOOK schedulingShital K Dhamal FCFS Scheduling Algorithm Itis the simplest Disk Scheduling algorithm. It services the IO requests in the order in which they arrive. There is no starvation in this algorithm, every request is serviced. 1. First Come First Serve Disk Scheduling Algorithm A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of F/W head =50. Calculate total number of track movement by 8{W head. 1. First Come First Serve Disk Scheduling Algorithm Request queue contains track no 82, 170, 43, 140, 24, Current Position of R/W head =! 190 respectively. 0 6 24 43 50 82 140 190__199Shital K Dhamal 1. First Come First Serve Disk Scheduling Algorithm ° 16 24 43. 50 82 140 170190 199 Seek Time = (82—50) + (170-82) + (170-43) + (140-43) + (140-16) + (190-16) = 642 Seek Time = (170-50) + (170-43)+ (140-43)+ (140-16) + (190-16) = 642 Consider the following disk request sequence for a disk with 100 tracks 45, 21, 67, 90, 4, 50, 89, 52, 61, 87, 25 Head pointer starting at 50 and moving in left direction . Find the number of head movements in cylinders Sa eis o 4 212545 50 52 61 67 87 89 90 99 4, 21, 25, 45, 50, 52, 61, 67, 87, 89, 90 10Shital K Dhamal 45, 21, 67, 90, 4, 50, 89, 52, 61, 87, 25 Ft ome Fst Sev Dk ened arn Seek Time = (50-45)+(45-21)+(67-21) +(90-67)+(90-4)+ (89-4) (50-4)+(89-50)+(61-52)+(87-61)+(87-25) = 403 Seek Time = (50-21) + (90-21) + (90-4) + (89-4) + (89-52) + (87-52) + (87-25) 403 First Come First Serve (FCFS) Advantages - 1. First Come First Serve algorithm has a very simple logic, it executes the process requests one by one in the sequence they arrive. 2. Thus, First Come First Serve is very simple and easy to understand and implement. aShital K Dhamal 3. InFCFS eventually, every process gets a chance to execute, so no starvation occurs Disadvantages — 1. This scheduling algorithm is non-pre-emptive, which means the process can’t be stopped in the middle of execution and will run its full course 2. FCFS being a non-pre-emptive scheduling algorithm, the short processes which are at the back of the queue have to wait for the long process at the front to finish 3. The throughput of FCES is not very efficient, 4. FCFS is implemented only on small systems where input-output efficiency is not paramount SSTF Scheduling Algorithm The shortest seek time first (SSTF) algorithm selects the disk VO request which requires the least disk arm movement from its current position regardless of the direction. It reduces the total seek time as compared to FCFS It allows the head to move to the closest track in the service queue. Shortest Seek Time First Algorithm A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at SO Calculate total number of track movement by R/W head using SSTF 12Shital K Dhamal Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Oo 16 24 43 50 82 140 170 __190 Seek Time = (50 — 16) + (190-16) 34+174 + =208 Shortest Seek Time First Alggrithm, Shortest Seek Time First Disk Scheduling (SSTF) Consider the following disk request sequence for a disk with 100 tracks. 45, 21, 67, 90, 4, 89, 52, 61, 87, 25 Head pointer starting at 50. Find the number of head movements in cylinders using SSTF scheduling. 199 13Shital K Dhamal 45, 21, 67, 90, 4, 89, 52, 61, 87, 25 O 4 21 25 45 50 52 61 67 87 89 90 100 i 4 it | i Seek Time = (50-45) + (90-45) + (90-4) _ 5 +45 +86 =136 Shortest Seek Time First (SSTF) : Advantages — 1. The total seek time is reduced compared to First Come First Serve. throughput SSTF improves and ine 3. Less average waiting time and response time in SSTP. Disadvantages — 1, In STF there is an overhead of finding out the closest request. Starvation may occur for requests far from the head In SSTF high variance is present in response time and waiting time. 4. Frequent switching of the Head’s direction slows the algorithm. Scan Algorithm Itis also called an Elevator Algorithm. In this algorithm, the disk ann moves in a particular direction til the end, satisfying all the requests coming in its path, and then it tums back and moves in the reverse direction satisfying requests coming in its path It works in the way an elevator works, the elevator moves in a direction completely till the last floor of that direction and then tums back. 14Shital K Dhamal SCAN Algorithm (Elevator Algorithm) A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at 50 and moving in right direction(towards larger value) 1. Calculate total number of track movement by R/W head using SCAN. 2. If read/write takes Ins to move from one track to another then total time taken - Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. 0 16 24 43 50 82 140 170 190 199 | Seek Time = (82 — 50) + (140-82)+ (170-140)+ (190-170)+ (199-190) +(199-43) +(43-24) +(24-16) =32458+30+20+9+156+19+8 = 332 46 ha 15Shital K Dhamal Ifread/write takes Ins to move from Seek Time = (199-50) + (199 -16) one track to another then total time taken —- wi =149 +183 = 332 = 332 * Ins = 332 ns SCAN Algorithm (Elevator Algorithm) A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at 50 and moving in left direction(towards smaller value) Calculate total number of track movernent by R/W head using SCAN. If read/write takes Ins to move from one track to another then total time taken - 16Shital K Dhamal Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. O16 24 43 50 82 140 170 190 199 Seek Time = (50-0) +(190 -0) 0+ 190 40 SCAN Algorithm Consider the following disk request sequence for a disk with 100 tracks. 98, 137, 122, 183, 14, 133, 65, 78 Head pointer starting at 54 and moving in left direction. Find the number of head movements in cylinders using CSCAN scheduling. wvShital K Dhamal Consider the following disk request sequence for a disk with 200 tracks. 98, 137, 122, 183, 14, 133, 65, 78 0 14 54 65 78 98 122 133 137 183 199 [_ yf ‘ f | | | | poet H+} | | Seek Time = (54-0)+(183-0) =54+183 =237 Msmern ome SCAN Algorithm * SCAN: Advantages — 1. Scan scheduling algorithm is simple and easy to understand and implement. 2. Starvation is avoided in SCAN algorithm. 3. Low variance Occurs in waiting time aud respouse time. Disadvantages — 1, Long waiting time occurs for the cylinders which are just visited by the head 2. In SCAN the head moves till the end of the disk despite the absence of requests, to be serviced. C-SCAN algorithm In C-SCAN algorithm, the arm of the disk moves in a particular direction servicing requests until it reaches the last cylinder, then it jumps to the last cylinder of the opposite direction. without servicing any request then it turns back and start moving in that direction servicing the remaining requests. 18Shital K Dhamal CSCAN Algorithm Consider the following disk request sequence for a disk with 100 tracks. 98, 137, 122, 183, 14, 133, 65, 78 Head pointer starting at 54 and moving in left direction. Find the number of head movements in cylinders using CSCAN scheduling. Consider the following disk request sequence for a disk with 200 tracks. 98, 137, 122, 183, 14, 133, 65, 78 o 14 54 65 78 98 122 133_137_ 183 199 Seek Time = (54-0) +(199 -0) + (199 - 65) =544199+134 =387 J CSCAN Algorithm « 19Shital K Dhamal Circular SCAN (C- SCAN) Algorithm A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at 50 and moving in right direction(towards larger value) Calculate total number of track movement by R/W head using C - SCAN. Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. 0 | | | { 1624 4350 82 140 170 190 199 Seek Time = (199-50) + (199-0) +(43-0) =391 C-SCAN: Advantages — 1. C-SCAN Algorithm is the successor and the improved version of the SCAN scheduling Algorithm. ‘The Head move from one end to the other of the disk while serving all the requests in between. 3. The waiting time for the cylinders which were just visited by the head is reduced in C-SCAN compared to the SCAN Algorithm. 20Shital K Dhamal 4. Uniform waiting time is provided 5. Better response time is provided Disadvantages — 1. More seek movements are caused in C-SCAN compared to SCAN Algorithm. 2. InC-SCAN even if there ate no requests lef to be serviced the Head will still travel to the end of the disk unlike SCAN algorithm. Look Scheduling LOOK is the advanced version of SCAN (elevator) disk scheduling algorithm which gives slightly better seek time than any other algorithm in the hierarchy (FCFS->SRIF->SCAN- >C-SCAN->LOOK). The LOOK algorithm services request similarly as SCAN algorithm meanwhile it also “looks” ahead as if there are more tracks that are needed to be serviced in the same direction. If there are no pending requests in the moving direction the head reverses the direction and start servicing requests in the opposite direction. ‘The main reason behind the better performance of LOOK algorithm in comparison to SCAN is becaust this algorithm the head is not allowed to move till the end of the disk. Consider the following disk request sequence for a disk with 200 tracks 98, 137, 122, 183, 14, 133, 65, 78 Head pointer starting at 54 and moving in left direction. Find the number of head movements in cylinders using LOOK scheduling. aShital K Dhamal 98, 137, 122, 183, 14, 133, 65, 78 O14 54 65 78 98 122 133 134 183 Seek Time = (54-14) + (183-14) =40+169 =209 LOOK Disk Scheduling A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at 50 and moving in the right direction. Calculate total number of track movement by R/W head using LOOK 199 22Shital K Dhamal Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. O16 24 43.50 82 140 170 190 199 O16 24 43 50 82 140 170 190 _199 Seek Time = (190- 50) + (190-16) =140+174 =314 LOOK: Advantages — 1. If there are no requests left to be services the Head will not move to the end of the disk unlike SCAN algorithm 2. Better performance is provided compared to SCAN Algorithm, 3. Starvation is avoided in LOOK scheduling algorithun, 4. Low variance is provided in waiting time and response time. 23Shital K Dhamal Disadvantages — 1. Overhead of finding the end requests is present 2. Cylinders which are just visited by Head have to wait for long time. C- LOOK Disk Scheduling + C-LOOK is an enhanced version of both SCAN as well as LOOK disk scheduling algorithms. + This algorithm also uses the idea of wrapping the tracks as a circular eylinder as C- SCAN algorithm but the seek time is better than C-SCAN algorithm. In this algorithm, the head services requests only in one direction (either left or right) until all the requests in this direction are not serviced and. + then jumps back to the farthest request on the other direction and service the remaining requests which gives a better uniform servicing as well as avoids wasting seek time for going till the end of the disk. Circular LOOK A disk contains 200 tracks (0 to 199). Request queue contains track no 82, 170, 43, 140, 24, 16, 190 respectively. Current Position of R/W head =50. Head pointer starting at 50, Direction is towards large value, Calculate total number of track movement by R/W head using CLOOK 24Shital K Dhamal Circular LOOK 0 24 43 50 82 140 170 190 199 SeekTime= (190-50) + (190 - 16) +(43 - 16) =140+174+27 =341 CLOOK Disk Scheduling Algorithm Consider the following disk request sequence for a disk with 100 tracks. 98, 137, 122, 183, 14, 133, 65, 78 Head pointer starting at 54 and moving in left direction. Find the number of head movements in cylinders using CLOOK scheduling. 25Shital K Dhamal 98, 137, 122, 183, 14, 133, 65, 78 Refer net Oo. 14 54 3199 1 5 78 98 122 133 134 1 f i Seek Time = (54—14)+ (183-14) +(183- 65) 40 +169+118 327 C-LOOK : Advantages — 1. InC-LOOK the head does not have to move till the end of the disk if there are no requests to be serviced. 2. There is less waiting time for the cylinders which are just visited by the head in C-LOOK. C-LOOK provides better performance when compared to LOOK Algorithm. Starvation is avoided in C-LOOK. Low variance is provided in waiting time and response time. Disadvantages - In C-LOOK an overhead of finding the end requests is present. Summary of Disk Scheduling Algorithms — retatwaae~21 o 4 5 10 as 20 26Shital K Dhamal Summary of Disk Scheduling Algorithmg’s!ss¢724 ‘5,15. Readwrite head is at 20, ‘Moving towards larger value a.scan o 8 5 10. as 20 a.cscant o 4 5 10, as 20 Summary of Disk Scheduling Algorithms fiises:2, 4,5,15 Read-write head is at 10, moving towards larger value o 4 5 10 45 20 o 4 s 10 a5. 20 Buffering in Operating System ‘The buffering in the operating system is an area in the main memory that is used for storing data in a buffer or cache temporatily, this buffered data then can be accessed more quickly as compared to the original source of the data, Buffering creates a synchronization between two devices having different processing speeds. For example, if a hard dise (supplier of data) has high speed and a printer (accepter of data) has low speed, then buffering is required. ‘Types of Buffering There are three main types of buffering in the operating system, such as: aShital K Dhamal Types of Buffer tre] Taig 1, Single Buffer In Single Buffering, only one buffer is used to transfer the data between two devices. The producer produces one block of data into the buffer. After that, the consumer consumes the buffer. Only when the buffer is empty, the processor again produces the data, OPERATING SYSTEM USER PROCESS I/O DEVICE Single Buffer: SINGLE I/O BUFFERING 2. Double Buffer In Double Buffering, two buffers are used in place of one. In this buffering, the producer produces items in one buffer while the consumer consumes items from another buffer simultaneously. So, the producer not needs to wait for filling the buffer. Double buffering is also known as buffer swapping OPERATING SYSTEM USER PROCESS 1/0 DEVICE DOUBLE 1/0 BUFFERING 28Shital K Dhamal 3. Circular Buffer When more than two buffers are used, then itis called circular buffering. It is used to solve the issues associated with the double buffering technique. Sometimes, the double buffering becomes insufficient, when the process performs rapid bursts of /O. In the circular buffer, each individual buffer acts a unit. 1/0 DEVICE CIRCULAR 1/0 BUFFERING Advantages of Buffering + Buffering reduces the mumber of /O operations required to access data. + Buffering reduces the amount of time that processes have to wait for the data + Buffering improves the performance of /O operations as it allows data to be read or written in large blocks instead of 1 byte or 1 character at a time. + Buffering can improve the overall performance of the system by redueing the number of system calls and context switches required for /O operations. Limitations of Buffering + Buffers of large sizes consume significant amount of memory that can degrade the system performance. 29Shital K Dhamal University Questions Module 6 — I/O Management Sno | Question Year Marks 1 Suppose that a disk drive has 5000 cylinders, numbered 010 | May 2019 | 10 4999, The drive is currently serving a request at cylinder 143, | May 2017 and the previous request was at cylinder 125. The queue of | y00 5917 pending requests in FIFO is ordered as 80, 1470, 913, 1777, 948, 1022. 1750,130. What is the total distance that the disk arm moves for following by applying the following algorithms? 1. FCFS 2. SSTF 3. LOOK 4. SCAN, 2 Explain disk scheduling algorithms/methods with examples | May 2018 | 10 Dec 2019 May 2022 Dec 2023 3 Explain Interrupt [0 and discuss the advantages of Dec 2019 | 10 Interrupt-driven IO over programmed IO. Explain in detail Interrupt Driven IO, Programmed IO, and DMA 4 ‘Compare disk scheduling algorithms Dec 2018 | 10 5 Give details of VO Buffering techniques May 2022 | 5 6 ‘Write short note on DMA May 2017 | 5 30