Presentation is loading. Please wait.

Presentation is loading. Please wait.

File StructureSNU-OOPSLA Lab1 Chap3. Secondary Storage and System Software File Structures by Folk, Zoellick, and Ricarrdi.

Similar presentations


Presentation on theme: "File StructureSNU-OOPSLA Lab1 Chap3. Secondary Storage and System Software File Structures by Folk, Zoellick, and Ricarrdi."— Presentation transcript:

1 File StructureSNU-OOPSLA Lab1 Chap3. Secondary Storage and System Software File Structures by Folk, Zoellick, and Ricarrdi

2 File StructureSNU-OOPSLA Lab2 Chapter Objectives u Describe the organization of typical disk drives, including basic units of organization and their relationships u Identity and describes the factors affecting disk access time, and describe methods for estimating access times and space requirements u Describe magnetic tapes, identify some tape applications, and investigate the implications of block size on space requirements and transmission speeds u Identify fundamental differences between media and criteria that can be used to match the right medium to an application u Describe in general terms the events that occur when data is transmitted between a program and a secondary storage device u Introduce concepts and techniques of buffer management u Illustrate many of the concepts introduced in the chapter, especially system software concepts, in the context of UNIX

3 File StructureSNU-OOPSLA Lab3Contents 3.1 Disks 3.2 Magnetic Tape 3.3 Disk versus Tape 3.4 Introduction to CD-ROM 3.5 Physical Organization of CD-ROM 3.6 CD-ROM Strengths and Weaknesses 3.7 Storage as a Hierarchy 3.8 A Journey of a Byte 3.9 Buffer Management 3.10 I/O in UNIX

4 4 Disks u Random Access Device( the disk shape provide this feature while other geometric shapes not provide it) u Disk device - DASD(Direct Access Storage Devices) u Magnetic disk( store the binary data by using the two magnetize nodes N and S). u hard disk - most common u floppy disk - inexpensive, little capacity u magnetic tape - serial access u Iomega Zip (100 M byte) (The Zip drive is a medium-capacity removable floppy), Jaz (1 G byte) (The Jaz drives a removable hard disk storage system introduced by the Iomega company in 1995.) u Nonmagnetic disk u optical discs - increasingly important for secondary storage u e.g. CD, DVD 3.1 Disks

5 u Compact disc (CD) is a digital optical disc data storage format. u DVD (sometimes explained as "digital video disc" or "digital versatile disc" is a digital optical disc storage format. u Blu-ray Disc (BD) is a digital optical disc data storage format designed to supersede the DVD format, in that it is capable of storing high-definition video resolution) Blu-ray Discs contain 25 GB per layer. File StructureSNU-OOPSLA Lab5

6 Electronic Disks u Electronic disk is a data storage device functionally similar to a hard disk drive but using flash memory. u Flash memory is an electronic non- volatile computer storage medium that can be electrically erased and reprogrammed. File StructureSNU-OOPSLA Lab6

7 7 Organization of Disks(1) u Components of Disk u Tracks (Number of adjacent sectors at the same diameter) u Sectors Ü the smallest addressable portion of a disk(smallest block that can be read or write) u Disk pack Ü a collection of a lot of platters u Cylinder (vertical collection of tracks) u Arms are moving together so we can read or write vertical sector in parallel. u Seeking (Finding the requested sectors that represent our file) u Read/write (r/w )arm movement set the position of requested sectors. u the slowest part of reading data from disk because it is mechanical motion. 3.1 Disks

8 File StructureSNU-OOPSLA Lab8 Organization of Disks(2) TrackSector Gaps Surface of disk showing tracks and sectors 3.1 Disks

9 File StructureSNU-OOPSLA Lab9 Boom Read/write heads Spindle Platters Schematic illustration of disk drive 3.1 Disks Organization of Disks(3)

10 File StructureSNU-OOPSLA Lab10 Estimating Capacities(1) u Cylinder --> Track --> Sector --> Byte u Track capacity = # of sectors per track * bytes per sector u Cylinder capacity = # of tracks per cylinder * track capacity u Drive capacity = # of cylinders * cylinder capacity ** Inner track, Outer track? Same capacity? 3.1 Disks

11 u Inner track: the track more near to the disk center. u Outer track: the track more far from the disk center. u Is the size of the inner and outer tracks equal (same capacity)? u The size of the inner outer tracks logically same (have the same number of sectors) but physically no they have different sizes so that we have gaps. File StructureSNU-OOPSLA Lab11

12 Example: Hard disk with 5 platters each surface contains 20 tracks and each track contains 50 sectors of size 512 bytes; find the disk size? File StructureSNU-OOPSLA Lab12

13 File StructureSNU-OOPSLA Lab13 Estimating Capacities(2) u Ex) file with 20,000 fixed-length data records on 300MB disk u # of bytes per sector = 512 u # of sectors per track = 40 u # of tracks per cylinder = 11 u # of cylinders = 1,331 u How many cylinders does the file require if each data record requires 256 bytes? 3.1 Disks

14 File StructureSNU-OOPSLA Lab14 Organizing Tracks by Sector(1) u Two ways to organize data on a disk u by sector & by user-defined block u Physical placement of sectors u logical view and physical view u adequate way to view a file logically, but always not good way to store sectors physically u interleaving — problems occur due to the gap between disk revolution speed and disk controller speed — high performance disks (with high speed disk controller) now offer 1:1 interleaving 3.1 Disks

15 File StructureSNU-OOPSLA Lab15 1 2 3 4 21 22 23 32 33 34 If interleaving factor is 3..... Disk 3.1 Disks Interleaving Organizing Tracks by Sector(2)

16 File StructureSNU-OOPSLA Lab16 Organizing Tracks by Sector(3) u Clusters (for improving file access speed) u One file consists of one or more clusters u Cluster is the smallest unit of space One cluster is a fixed number of contiguous sectors. u Contiguous sectors per cluster and Mapping table u No additional seek (arm movement) for one cluster u Very useful when access granularity is medium u large cluster vs. small cluster u FAT (File Allocation Table) u contains a list of all the clusters in file u each cluster entry --> physical location of the cluster 3.1 Disks

17 File StructureSNU-OOPSLA Lab17 FAT(File Allocation Table) FAT(File Allocation Table) - A table containing mappings to the physical locations of all the clusters in all files on disk storage Cluster number Cluster location 123...123... File allocation table(FAT) The part of the FAT pertaining to our file...... 2 1 3.1 Disks Organizing Tracks by Sector(3)

18 File StructureSNU-OOPSLA Lab18 Organizing Tracks by Sector(4) u Extents (further attempt for file access speed) u Contiguous clusters per extent and Mapping table u File may have several extents u Very useful when accessing the whole file sequentially u Many extents of a file increase amount of seeking 3.1 Disks file extents extent

19 File StructureSNU-OOPSLA Lab19 Organizing Tracks by Sector(5) u Fragmentation u the space that goes unused within a cluster, block, track, or other unit of physical storage u internal fragmentation: loss of space within a sector or cluster u trade-offs in use of large cluster sizes u ex) size of sector is 512 bytes and size of all records in file is 300 bytes u we may enforce one record / sector u we may allow span sectors 3.1 Disks

20 File StructureSNU-OOPSLA Lab20 Organizing Tracks by Block(1) u User-defined size of blocks can be fixed or variable u No sector spanning ==> No internal fragmentation u Blocking factor u the number of records that are to be stored in each block u In block-addressing, each block of data is accompanied by one or more subblocks containing extra information about the data block u Note: the “block” has a different meaning in the context of the UNIX I/O system 3.1 Disks

21 File StructureSNU-OOPSLA Lab21 Sector organization vs. Block organization 1 1 1 1 1 1 2 2 22 2 3 3 3 44 4 4 5 5 5 Sector 1Sector 2Sector 3Sector 4Sector 5 (a) sector organization (b) block organization 3.1 Disks 1 1 1 1 1 1 2 2 22 2 3 3 3 44 4 4 5 5 5

22 File StructureSNU-OOPSLA Lab22 Organizing Tracks by Block(2) u Subblocks u count subblocks u contain the num of bytes in the accompanying data block u key subblocks u contain key for last record in data block Count subblock Count subblock Key subblock Data subblock Count subblock Data subblock Count subblock Key subblock Data subblock Count subblock Key subblock Data subblock 3.1 Disks

23 File StructureSNU-OOPSLA Lab23 Nondata Overhead(1) u Stored during preformatting u Sector-addressable disks (physical) u sector/track address, condition, gaps, and synchronization marks u of no concern to programmer u Block-organized disks (user-defined) u some of nondata for programmer u more nondata provided with blocks than with sectors 3.1 Disks

24 File StructureSNU-OOPSLA Lab24 Nondata Overhead(2) u Ex) block-addressable disks u 20,000 bytes/track u subblock, interblock gap equivalent to 300 bytes/block u If we want to store a file containing 100-byte records, how many records can be stored per track if blocking factor is 10 or 60 ? è if blocking factor is 10, then 20000 / ( (10*100)+300) = 15 blocks ==> 150 records è if blocking factor is 60, 180 records(=3 blocks) 3.1 Disks

25 File StructureSNU-OOPSLA Lab25 Nondata Overhead(3) u Block-organized scheme u larger blocking factor (many records in a block) u less amount of nondata block u more efficient use of storage u pros. u flexibility -> savings in time & efficiency u cons. u internal fragmentation problem u necessary for programmer and OS to do extra data organization u sector interleaving cannot be used to improve performance 3.1 Disks

26 File StructureSNU-OOPSLA Lab26 Cost of Disk Access u Disk Access Time è Seek time + Rotational delay + Transfer time u Seek time u time to move arm to locate cylinder u more costly in multi-user environment u average seek time for a particular file operation u Rotational delay u time taken for disk to rotate correct sector under read/write arm u Transfer time (for actual data) number of bytes transferred number of bytes on a track rotation time 3.1 Disks

27 File StructureSNU-OOPSLA Lab27 Timing Computation u Seagate Cheetah 9 Gbyte drive u Minimum seek (track-to-track): 0.78msec u Maximum seek: 19 msec u Average seek: 8 msec u Rotation delay: 3 msec u Read one track: 6 msec u Read one cluster: 0.28 msec u Read 8704K size file u sequentially access the whole file: 1.7 sec u randomly access the whole file: 9.25 sec

28 File StructureSNU-OOPSLA Lab28 Effect of Block Size on Performance (In UNIX) u UC Berkeley CSRG team u Many experiments in BSD UNIX System u Trade-off between block size, fragmentation, access time u Doubling block size to 1,024 bytes improves performance by more than a factor of 2 u In 512-byte block size: the slow throughput & low internal fragmentation (6.9%) u In 4,096-byte block: the fastest throughput & internal fragmentation (45.6%) ==> Invent the cluster concept! 3.1 Disks

29 File StructureSNU-OOPSLA Lab29 Disk as Bottleneck u Disk performance is increasing, but still slow! u Even high-performance network is faster than disk (5 M/ sec HD is slower than 100M bps LAN) u Disk bound jobs u CPU and network must wait u Solution techniques u multiprogramming u striping - parallel I/Os u avoid accessing disk u RAM disk, disk caches - locality 3.1 Disks

30 File StructureSNU-OOPSLA Lab30 Magnetic Tape u A set of parallel tracks u 9 tracks - parity bit u Frame u one-bit-wide slice of tape u Interblock gaps u permit stopping and starting 3.2 Magnetic Tape

31 File StructureSNU-OOPSLA Lab31 Nine-track tape TrackFrame 011010010011010010 Gap Data block Gap 3.2 Magnetic Tape

32 File StructureSNU-OOPSLA Lab32 Estimating Tape Length(1) u There is an interblock gap for each data block u Space requirement s è s = n * ( b + g ) u b is the physical length of a data block u g is the length of an interblock gap u n is the number of data blocks u Tape density u Tape speed u Size of interblock gap 3.2 Magnetic Tape

33 File StructureSNU-OOPSLA Lab33 Estimating Tape Length(2) u Ex) one million 100-byte records u 6,250 BPI tape u 0.3 inches of interblock gap u How much tape is needed? u when blocking factor is between 1 and 50 u Nominal recording density u Effective recording density: u num of byte per block / num of inches for block 3.2 Magnetic Tape

34 File StructureSNU-OOPSLA Lab34 Estimating Data Transmission Times u Factors of data transmission rate u interblock gaps u effective recording density u nominal recording density u speed of r/w head u time to start/stop the tape 3.2 Magnetic Tape

35 File StructureSNU-OOPSLA Lab35 Disk VS. Tape u Disk u Random access u Immediate access u Expensive seek in sequential processing u Tape u Sequential access u Long-term storage u No seek in sequential processing uDecrease in cost of disk and RAM uMore RAM space is available in I/O buffers, so disk I/O decreases uTertiary storage for backup: CD-ROM, tape... 3.3 Disk versus Tape

36 File StructureSNU-OOPSLA Lab36 Introduction to CD-ROM 3.4 Introduction to CD-ROM u CD-ROM: Compact Disc Read-Only Memory u Can hold over 600MB(200,000 pages) u Easy to replicate u Useful for publishing or distributing medium u But, not storing and retrieving data u CD-ROM is a child of CD audio u CD audio provides u High storage capacity u Moderate data transfer rate u But, against high seek performance  Poor seek performance

37 File StructureSNU-OOPSLA Lab37 History of CD-ROM: Videodisc u Videodisc technology developed in late 1960's and early 1970's u The goal was to store movie u A number of methods for storing video signals 1. Use a needle to respond mechanically to grooves in a disc like a vinyl LP record 2. Use optical storage u Many companies were fighting which approach should become a standard  Videodisk  LaserVision  CD audio  CD-ROM 3.4 Introduction to CD-ROM

38 File StructureSNU-OOPSLA Lab38 History of CD-ROM: LaserVision u LaserVision u Emerged as the winner of standard battles about standard u Support CLV(Constant Linear Velocity) and CAV(Constant Angular Velocity) format u Fast seek performance by using CAV format u Data are stored in analog form u Why did they fail? u Earlier disputes over the physical format of the video disc, many incompatible encoding scheme and error correction techniques were used by many firms  No standard format  the market never grew 3.4 Introduction to CD-ROM

39 File StructureSNU-OOPSLA Lab39 History of CD-ROM: CD-ROM u CD-ROM u Philips and Sony developed CD-ROM in 1984 in order to store music on a disc u Use a digital data format u The development of CD-ROM as a licensing system results in widely acceptance in the industry u Promised to provide a standard physical format u Any CD-ROM drive can read any sector which they want u Computer applications store data in a file not in terms of sector, thus, file system standard should be needed u In late 1985, videodisc/digital data industry moved into CD-ROM industry u In early summer of 1986, an official standard for organizing files was worked out 3.4 Introduction to CD-ROM

40 File StructureSNU-OOPSLA Lab40 3.5 Physical Organization of CD-ROM Physical Organization of Master Disk u Master Disc u Formed by using the digital data, 0 or 1 u Made of glass and coated that is changed by the laser beam u Two part of CD-ROM u Pit u The areas that is hit by the laser beam u Scatter the light u Land u Smooth, unchanged areas between pits u Reflect the light laser beam pit land light

41 File StructureSNU-OOPSLA Lab41 Encoding Scheme of CD-ROM u Encoding scheme u The alternating pattern of high- and low-density reflected light is the signal u 1 : transition from pit to land and back again u 0s : the amount of time between transitions u Constraint u The limits of the resolution of the optical pickup (generic hardware limit), there must be at least two 0’s between any pair of 1’s (no two adjacent 1s) u We cannot represent all bit patterns, thus, we need translation scheme u We need at least 14 bits to represent 8 bits under this constraint u EFM(eight to fourteen modulation) coding is used 1 10 10000100000000 EFM 00000001 2 3.5 Physical Organization of CD-ROM

42 File StructureSNU-OOPSLA Lab42 Format of CD-ROM u CD audio chose CLV format instead of CAV format because u CD audio requires large storage space u CD audio is played from the beginning to the end sequentially u Format of CD-ROM u CLV(Constant Linear Velocity) u A single spiral pattern u Same amount of space for each sector u Capability for writing all of sectors at the maximum density u Rotational speed is slower in reading outer edge than in inner edge u Finding the correct speed though trial and error u Characteristics u Poor seek performance u No straightforward way to jump to a specified location 3.5 Physical Organization of CD-ROM

43 File StructureSNU-OOPSLA Lab43 Constant Angular Velocity Disk u Magnetic disk usually uses CAV(Constant Angular Velocity) u Concentric tracks and pie-shaped sectors u Data density is higher in inner edge than in outer edge u Storage waste: total storage is less than a half of CLV u Spin the disc at the same speed for all positions  Easy to find a specific location on a disk  good seek performance CAVCLV 3.5 Physical Organization of CD-ROM

44 File StructureSNU-OOPSLA Lab44 Addressing of CD-ROM u Addressing u Magnetic disk: cylinder/track/sector approach u CD-ROM: a sector-addressing scheme u Track density varies thus, each second of playing time on a CD is divided into 75 sectors u 75 sectors/sec, 2 Kbytes/sector u At least one-hour of playing time u Maximum capacity can be calculated: 600 Mbytes 60 min * 60 sec/min * 75 sectors/sec = 270,000 sectors u We address a given sector by referring minutes, second, and sector of play u 16:22:34 means 34th sector in the 22nd second in the 16th minutes of play 3.5 Physical Organization of CD-ROM

45 File StructureSNU-OOPSLA Lab45 Fundamental Design of the CD disc u Initially designed for delivering digital audio information u Store audio data in digital form u Wave patterns should be converted into digital form u Measure of the height of the sound: 65,536 different gradation(16 bits) u Sampling rate: 44.1 kHz, because of 2 times of 20,000 Hz upto which people can listen u 16 bits sample, 44,100 times per second, and two channel for stereo sound, we should store 176,400 bytes per seconds u Storage capacity of CD is 75 sectors per seconds, we have 2,352 bytes per sector u CD-ROM divides this raw sector as shown in the following figure 12 bytes synch 4 bytes sector ID 2,048 bytes user data 4 bytes error detection 8 bytes null 276 bytes error correction 3.5 Physical Organization of CD-ROM

46 File StructureSNU-OOPSLA Lab46 3.6 CD-ROM Strengths and Weaknesses File Structure Problem of CD-ROM u The most important problem in using CD-ROM as a storage media u Strongness and weakness of CD-ROM u Strong aspects of CD-ROM u Data transfer rate: 75 sectors/sec u Storage capacity : over 600 Mbytes u Inexpensive to duplicate and durable u Weak aspects of CD-ROM u Poor seek performance (weak random access) u Magnetic disk: 30 msec, CD-ROM : 500 msec u Comparison of access time of a large file from several media u RAM: 20 sec, Disk: 58 days, CD-ROM: 2.5 years u We should have a good file structure avoiding seeks to an even greater extent that on magnetic disk

47 File StructureSNU-OOPSLA Lab47 Storage as a Hierarchy u Primary storage u 1/1,000,000,000 sec u semi-conductors: registers, RAM, RAM disk, and disk cache u Secondary storage u 1/1,000 sec u magnetic disks(DASD), tape and mass storage(Serial) u Off-line storage (archival & back-up) u 10 sec u removable magnetic disks, optical discs, tapes 3.7 Storage as a Hierarchy

48 File StructureSNU-OOPSLA Lab48 A Journey of a Byte(1) u What happens when a program writes a byte to a file on a disk? u WRITE(TEXT, c, 1); User Program ---> OS File manager --> I/O buffer ---> I/O processor ---> Disk controller ---> Disk 3.8 A Journey of a Byte

49 File StructureSNU-OOPSLA Lab49 A Journey of a Byte(2) 1 The program asks the OS to write the contents of the variable c to the next available position in TEXT. 2 The OS passes the job on to the file manager 3 The file manager looks up TEXT in a table containing information about it, such as whether the file is open and available for use, what types of access are allowed, if any, and what physical file the logical name TEXT corresponds to. 4 The file manager searches a FAT for the physical location of the sector that is to contain the byte. 5 The file manager makes sure that the last sector in the file has been stored in a system I/O buffer in RAM, then deposits the ‘P’ into its proper position in the buffer. Logical layer 3.8 A Journey of a Byte

50 File StructureSNU-OOPSLA Lab50 A Journey of a Byte(3) 6 The file manager gives instructions to the I/O processor about where the byte is stored in RAM and where it needs to be sent on the disk 7 The I/O processor finds a time when the drive is available to receive the data and puts the data in proper format for the disk. It may also buffer the data to send it out in chunks of the proper size for the disk 8 The I/O processor sends the data to the disk controller. 9 The controller instructs the drive to move the r/w head to the proper track, waits for the desired sector to come under the r/w head, then sends the byte to the drive to be deposited, bit-by-bit, on the surface of the disk. Physical layer 3.8 A Journey of a Byte

51 File StructureSNU-OOPSLA Lab51 What is buffer? u Definition - the part of main memory available for storage of copies of disk blocks u Program buffers vs. System I/O buffers u Buffer manager u subsystem responsible for the allocation for blocks u goal: u minimize the number of disk access u utilize the memory space effectively 3.8 A Journey of a Byte

52 File StructureSNU-OOPSLA Lab52 Buffer Management u Buffer bottlenecks u What if program performs both input and output on one character at a time, and only one I/O buffer is available? u At least two buffer - for input and output u I/O bound jobs: wait for I/O completion u Buffering Strategies for Performance u use more than one buffer u I/O system processes next block when CPU is processing current one 3.9 Buffer Management

53 File StructureSNU-OOPSLA Lab53 Buffering Strategies(1): Multiple buffering u Double buffering u swapping roles of two buffers after I/O finished u allows O/S operating on one buffer while the other buffer is being loaded or emptied u any number of buffers can be used u Buffer pooling u takes from a pool of available buffers u decides which buffer to take from a buffer pool é take buffer by LRU 3.9 Buffer Management

54 File StructureSNU-OOPSLA Lab54 Double Buffering Program data area To disk I/O buffer 1 I/O buffer 2 (a) (b) 3.9 Buffer Management

55 File StructureSNU-OOPSLA Lab55 Buffering Strategies(2): Move & Locate mode u Move mode (using both system buffer & program buffer) u moving data from one place in RAM to another before they can be accessed u sometimes, unnecessary data moves u Locate mode (using system buffer only or program buffer only) u perform I/O directly between secondary storage and program buffer (program’s data area) u system buffers handle all I/Os, but program uses locations through pointer variable 3.9 Buffer Management

56 File StructureSNU-OOPSLA Lab56 Move mode & Location mode system buffer program’s data area system buffer disk user’s program location (pointer) Move mode Locate mode 3.9 Buffer Management

57 File StructureSNU-OOPSLA Lab57 Buffering Strategies(3): Scatter/Gather IO u A file with many blocks and each block with header and data u Need to put the headers in one buffer and the data in a different buffer ==> may occur complication u Scatter-input mode u a single READ can scatter data into a collection of buffers u Gather-output mode u a single WRITE can gather several buffers and output 3.9 Buffer Management

58 File StructureSNU-OOPSLA Lab58 Scatter Input & Gather Output disk scatter input gather output read() write() buffer 1 buffer 2 buffer 1 buffer 3 buffer 2 3.9 Buffer Management

59 File StructureSNU-OOPSLA Lab59 I/O in UNIX I/O in UNIX 3.10 I/O in UNIX I/O system PROCESS KERNEL user programs librariesshell commands system call interface block I/O system (normal files) character I/O system (terminals, printers, etc.) network I/O system (sockets) block device drivers diskdisk... character device drivers consolesprinters... network interface drivers...networks...

60 File StructureSNU-OOPSLA Lab60 UNIX Kernel(1) u The central part of UNIX operating system u View all I/O as operating on a sequence of bytes u Make all operations below top layer independent of application’s logical view of file u Data structures related to unix files u file-descriptor table: owned by user process u open-file table: owned by kernel u index-node: a kind of FAT (one inode for each file in use) u index-node table: owned by kernel 3.10 I/O in UNIX

61 File StructureSNU-OOPSLA Lab61 UNIX Kernel(2) u File descriptor table u associates each file descriptor to open file table u every process has its own file descriptor table u Open file table u entries for every open file u file structures: r/w mode, offset, reference count 3.10 I/O in UNIX

62 File StructureSNU-OOPSLA Lab62 UNIX Kernel(3) file descriptor file table entry o(keyboard) 1(screen) 2(error) 3(normal) 4(normal) 5(normal). to open file table descriptor table 3.10 I/O in UNIX

63 File StructureSNU-OOPSLA Lab63 UNIX Kernel(4) R/W mode # of processes using it Offset of next access ptr to write routine inode table entry...... write1100...... to inode table write() routine for this type of file open file table 3.10 I/O in UNIX

64 File StructureSNU-OOPSLA Lab64 UNIX Kernel(5) u Inode(Index node) u data structure used to describe a file u when a file is opened, a copy of inode is loaded into RAM for rapid access u has a list of disk blocks of the file u this list is UNIX counterpart to FAT u Device driver u I/O processor program invoked by kernel performing I/O for devices 3.10 I/O in UNIX

65 File StructureSNU-OOPSLA Lab65 Inode Structure device permissions owner’s userid file size block count file allocation table............ An inode 3.10 I/O in UNIX

66 File StructureSNU-OOPSLA Lab66 Hard/Soft Link & File types u Linking file names to u Linking file names to files u hard link: direct reference u pointer from a directory to the inode of a file u ln src-file target-file u soft(symbolic) link: pathnames u link a file name to another file name u ln -s src-file target-file u File types u normal file (governed by block IO system) u special file: stream, signal control of device (governed by character IO system) u sockets: endpoints of IPC (governed by network IO system) 3.10 I/O in UNIX

67 File StructureSNU-OOPSLA Lab67 Let’s Review !!! 3.1 Disks 3.2 Magnetic Tape 3.3 Disk versus Tape 3.4 Introduction to CD-ROM 3.5 Physical Organization of CD-ROM 3.6 CD-ROM Strengths and Weaknesses 3.7 Storage as a Hierarchy 3.8 A Journey of a Byte 3.9 Buffer Management 3.10 I/O in UNIX


Download ppt "File StructureSNU-OOPSLA Lab1 Chap3. Secondary Storage and System Software File Structures by Folk, Zoellick, and Ricarrdi."

Similar presentations


Ads by Google