s Swap-Space Management s RAID Structure s Disk Attachment s Stable- Storage Implementation s Tertiary Storage Devices s Operating System Issues
Previous PDF | Next PDF |
[PDF] Tertiary Storage Devices
Module 14: Tertiary-Storage Structure Tertiary Storage Devices Removable Disks Removable Disks (Cont ) WORM Disks Tapes Operating System Issues Application Interface
[PDF] Chapter 5 Storage Devices
CSCA0101 Computing Basics 4 Storage Devices Types of Storage There are four type of storage: • Primary Storage • Secondary Storage • Tertiary Storage
[PDF] A Study on the Use of Tertiary Storage in Multimedia Systems
Tape media is still two orders of magnitude less expensive than magnetic disk storage; although, tape drives exhibit access latencies two to four orders of
[PDF] Tertiary Storage: An Evaluation of New Applications - UC Berkeley
Chapter 2 describes the technology of magnetic tape and optical disk devices It surveys existing drive and robot products and discusses tradeo s in the design of
[PDF] Chapter 12: Mass-Storage Systems
s Swap-Space Management s RAID Structure s Disk Attachment s Stable- Storage Implementation s Tertiary Storage Devices s Operating System Issues
[PDF] Mass-Storage Systems - Computer Science, FSU
Swap-Space Management ▫ RAID Structure ▫ Disk Attachment ▫ Stable- Storage Implementation ▫ Tertiary Storage Devices ▫ Operating System Issues
[PDF] Tertiary Storage in Multimedia Systems: Staging or Direct - CORE
an object that resides in tertiary storage can either be staged to and then played back from disks, or the object can be accessed directly from the tertiary drives
[PDF] Hierarchy and Characteristic of Storage Devices
Tertiary storage is a third level storage such as cloud storage Off-line storage is computer data storage on a medium or a device Primary storage is the only one
[PDF] tesco 2014 annual report
[PDF] tesco annual report 2013
[PDF] tesco annual report 2015
[PDF] tesco beef burgers halal
[PDF] tesla unit
[PDF] tesselaar roses
[PDF] test 10 7
[PDF] test 100 7
[PDF] test 7 14 olympus
[PDF] test 7 14 panasonic
[PDF] test and score data summary for toefl 2
[PDF] test anglais cecrl b2
[PDF] test anglais cecrl c1
[PDF] test anticorps coronavirus belgique
Chapter 12: Mass-Storage SystemsChapter 12: Mass-Storage Systems
12.2Silberschatz, Galvin and Gagne ©2009Operating System ConceptsChapter 12: Mass-Storage SystemsChapter 12: Mass-Storage Systems
nOverview of Mass Storage Structure nDisk Structure nDisk Attachment nDisk Scheduling nDisk Management nSwap-Space Management nRAID Structure nDisk Attachment nStable-Storage Implementation nTertiary Storage Devices nOperating System Issues nPerformance Issues12.3Silberschatz, Galvin and Gagne ©2009Operating System ConceptsObjectivesObjectives
nDescribe the physical structure of secondary and tertiary storage devices and the resulting effects on the uses of the devices nExplain the performance characteristics of mass-storage devices nDiscuss operating-system services provided for mass storage, including RAID and HSM12.4Silberschatz, Galvin and Gagne ©2009Operating System ConceptsOverview of Mass Storage StructureOverview of Mass Storage Structure
nMagnetic tape lWas early secondary-storage medium lRelatively permanent and holds large quantities of data lAccess time slowHAnd random access ~1000 times slower than disk
lMainly used for backup, storage of infrequently-used data, transfer medium between systems lKept in spool and wound or rewound past read-write head lOnce data under head, transfer rates comparable to disk l20-200GB typical storage lCommon technologies are 4mm, 8mm, 19mm, LTO-2 and SDLT nToday - primary mass storage structure is magnetic disks lTape used for backup lDepending on capacity required, CD/DVD also sometimes used for backup12.5Silberschatz, Galvin and Gagne ©2009Operating System ConceptsOverview of Mass Storage Structure (Cont.)Overview of Mass Storage Structure (Cont.)
nMagnetic disks provide bulk of secondary storage of modern computers lDrives rotate at 60 to 200 times per second lTransfer rate is rate at which data flow between drive and computerlPositioning time (random-access time) is time to move disk arm to desired cylinder (seek time) and time for desired sector to rotate under the disk head (rotational latency)
lHead crash results from disk head making contact with the disk surfaceHThat's bad
nDisks can be removable nDrive attached to computer via I/O bus lBusses vary, including EIDE, ATA, SATA, USB, Fibre Channel, SCSI lHost controller in computer uses bus to talk to disk controller built into drive or storage array12.6Silberschatz, Galvin and Gagne ©2009Operating System ConceptsDisk StructureDisk Structure
nDisk drives are addressed as large 1-dimensional arrays of logical blocks, where the logical block is the smallest unit of transfer. nThe 1-dimensional array of
logical blocks is mapped into the sectors of the disk sequentially.lSector 0 is the first sector
of the first track on the outermost cylinder.lMapping proceeds in
order through that track, then the rest of the tracks in that cylinder, and then through the rest of the cylinders from outermost to innermost.12.7Silberschatz, Galvin and Gagne ©2009Operating System ConceptsDisk AttachmentDisk Attachment
nHost-attached storage accessed through I/O ports talking to I/O busses nSCSI itself is a type of bus, up to 16 devices on one cable, SCSI initiator requests operation and SCSI targets perform tasks lEach target can have up to 8 logical units (disks attached to device controller nFibre Channel (FC) is high-speed serial architecture lCan be switched fabric with 24-bit address space - the basis of storage area networks (SANs) in which many hosts attach to many storage units lCan be arbitrated loop (FC-AL) of 126 devices12.8Silberschatz, Galvin and Gagne ©2009Operating System ConceptsNetwork-Attached StorageNetwork-Attached Storage
nNetwork-attached storage (NAS) is storage made available over a network rather than over a local connection (such as a bus) nAlthough often "mounted" as a filesystem, is really a remote server lNFS and CIFS are common protocols (with NFS's shortcomings) lImplemented via remote procedure calls (RPCs) between host and storage lNew iSCSI protocol uses IP network to carry the SCSI protocol12.9Silberschatz, Galvin and Gagne ©2009Operating System ConceptsStorage Area NetworkStorage Area Network
n"Super" version of NAS servers - SAN masks complexity - redundancy lLarge system / set of systems attached to nothing but storage (arrays) nCommon in large storage environments (and becoming more common)12.10Silberschatz, Galvin and Gagne ©2009Operating System ConceptsDisk SchedulingDisk Scheduling
nGoal of OS / controller is to minimize service time (provide quickest response time) - by using hardware most efficiently nTime factors in reading (or writing) a disk sector are: lSeek time is the time for the disk are to move the heads to the cylinder containing the desired sector. lRotational latency time is the additional time waiting for the disk to rotate the desired sector to the disk head. lTransfer time is the time to read data and move it to the system nCannot do much about transfer and latency time (HW) nCAN try to schedule reads/writes to minimize seek time lDo this by minimizing seek distance lIf system / disk is reasonably busy, can have significant impact lScheduling can be done by system, disk, or in between12.11Silberschatz, Galvin and Gagne ©2009Operating System ConceptsDisk Scheduling (Cont.)Disk Scheduling (Cont.)
nSeveral algorithms exist to schedule the servicing of disk I/O requests. nWe illustrate them with a request queue (0-199).98, 183, 37, 122, 14, 124, 65, 67
Head pointer = 53
(this is where the head is currently positioned) nEvaluate algorithms on how well they minimize seek time, on fairness, and on complexity12.12Silberschatz, Galvin and Gagne ©2009Operating System ConceptsFCFSFCFS Scheduling Scheduling
nFirst-come-first-served scheduling - FCFS lServe disk read / write requests in order in which they arrive nSimple, fair, OK if utilization is low (i.e., disk is not very busy) nBUT - it often results in excessive arm movement12.13Silberschatz, Galvin and Gagne ©2009Operating System ConceptsFCFSFCFS
Illustration shows total head movement of 640 cylinders/tracks.12.14Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSSTFSSTF Scheduling Scheduling
nShortest-seek-time-first scheduling lThe request chosen to be serviced next is the one closest to the current head position lSimilar in concept to SJF process scheduling nPROBLEM - may result in starvation nCan be considerable improvement over FCFS, but not optimal (because of possible starvation) nIllustration shows total head movement of 236 cylinders.12.15Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSSTF (Cont.)SSTF (Cont.)
Total tracks moved: 236
12.16Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSCANSCAN
nThe disk arm starts at one end of the disk, and moves toward the other end, servicing all possible requests along the way, then moves back toward the other end of the disk, doing the same lNote: it has to go all the way to the end of the disk (at each end) before it can turn around and start back nSometimes called the elevator algorithm. lThink of an elevator that must go to very top and very bottom floors before reversing directions nCan result in uneven service time, but no starvation nIllustration shows total head movement of 208 cylinders.12.17Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSCAN (Cont.)SCAN (Cont.)
Total tracks moved: 208
12.18Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSCAN (Cont.)SCAN (Cont.)
Total tracks moved: 208
NOTE - this example starts in the opposite direction from all the others, so is not a good comparison12.19Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSCAN (Cont.)SCAN (Cont.)
Total tracks moved: 331Note - this example goes in SAME DIRECTION as all the others12.20Silberschatz, Galvin and Gagne ©2009Operating System ConceptsC-SCANC-SCAN
nCircular SCAN nThe head moves from one end of the disk to the other, servicing requests as it goes. When it reaches the other end, however, it immediately returns to the beginning of the disk, without servicing any requests on the return trip. lOnly reads/writes when moving one direction lLike SCAN, goes from first to last sector of disk, even though there may be no requests that far nProvides a more uniform wait time than SCAN. nTreats the cylinders as a circular list that wraps around from the last cylinder to the first one.12.21Silberschatz, Galvin and Gagne ©2009Operating System ConceptsC-SCAN (Cont.)C-SCAN (Cont.)
Total tracks moved: 382
NOTE: when counting tracks moved, must also count while moving from one end to other, even if doing no I/O12.22Silberschatz, Galvin and Gagne ©2009Operating System ConceptsLOOKLOOK Scheduling Scheduling
nLike SCAN, but arm only goes as far as the last request in each direction (i.e., does not go all the way to each end of the disk) nSaves arm movement (and time) nC-LOOK has a similar relationship to C-SCAN12.23Silberschatz, Galvin and Gagne ©2009Operating System ConceptsC-LOOK (Cont.)C-LOOK (Cont.)
Total tracks moved: 322
12.24Silberschatz, Galvin and Gagne ©2009Operating System ConceptsComparison of Scheduling AlgorithmsComparison of Scheduling Algorithms
TOTAL TRACKS MOVED
FCFS - 640
SSTF - 236 - smallest, although some requests may starve SCAN - 208 - started opposite direction from others SCAN - 331 - going same direction as all the othersC-SCAN - 382
LOOK - 299 Best overall in this exampleC-LOOK - 322
12.25Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSelecting a Disk-Scheduling AlgorithmSelecting a Disk-Scheduling Algorithm
nSSTF is common and has a natural appeal lBut starvation possible nText says that SCAN and C-SCAN perform better than SSTF for systems that place a heavy load on the disk lBut only because they solve starvation problem - they don't really perform better nPerformance depends on the number and types of requests. nRequests for disk service can be influenced by the file-allocation method. nThe disk-scheduling algorithm should be written as a separate module of the operating system, allowing it to be replaced with a different algorithm if necessary.nEither SSTF or LOOK is a reasonable choice for the default algorithm (if modify SSTF to avoid starvation).
12.26Silberschatz, Galvin and Gagne ©2009Operating System ConceptsDisk ManagementDisk Management
nLow-level formatting, or physical formatting - Dividing a disk into sectors that the disk controller can read and write. nTo use a disk to hold files, the operating system still needs to record its own data structures on the disk. lPartition the disk into one or more groups of cylinders. lLogical formatting or "making a file system". nBoot block initializes system. lThe bootstrap is stored in ROM. lBootstrap loader program. nMethods such as sector sparing used to handle bad blocks (provide alternate sectors to replace bad ones) lLogically remove bad sectors or assign alternate sectors12.27Silberschatz, Galvin and Gagne ©2009Operating System ConceptsBooting from a Disk in Windows 2000Booting from a Disk in Windows 2000
MBR = "Master Boot Record"
12.28Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSwap-SpaceSwap-Space Management Management
nSwap-space - Virtual memory uses disk space as an extension of main memory. nSwap-space can be carved out of the normal file system,or, more commonly, it can be in a separate disk partition. nSwap-space management - various techniques l4.3BSD allocates swap space when process starts; holds text segment (the program) and data segment.HKernel uses swap maps to track swap-space use.
lLinux and late versions of Solaris only allocate swap space when a page is forced outlLinux supports multiple swap partitions, and suggests that they be at least 2X size of main memory (although there is some debate)
nOS/400 does not use separate swap space / swapping drive - all drives are paging devices in single-level store12.29Silberschatz, Galvin and Gagne ©2009Operating System ConceptsData Structures for Swapping on Linux SystemsData Structures for Swapping on Linux Systems
nSwap space only allocated (for a process) when page first swapped out, not when process created nSwap space is series of 4K page slots, used to hold swapped pages nSwap map - integer counters, show number of processes using page lIf = 0, then is available12.30Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRAID StructureRAID Structure
nRAID - Redundant Array of Independent Disks lMultiple disk drives provides reliability via redundancy. nSeveral improvements in disk-use techniques involve the use of multiple disks working cooperatively. nDisk striping uses a group of disks as one storage unit. nRAID schemes improve performance and / or improve the reliability of the storage system by storing redundant data. lMirroring or shadowing keeps duplicate copy of each disk. lBlock interleaved parity requires much less extra space for redundancy.12.31Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRedundancy for Improved ReliabilityRedundancy for Improved Reliability
Why do we care?
nSystems becoming larger and larger - now can have hundreds of disks attached, storing terabytes of data. lRecovery time following a failure is much longer for large systems than for smaller systems lLarge systems are often more mission critical than small system nEven with highly reliable hardware, as the number of instances of the hardware increases, the probability of failure increases lE.g., if mean time between failure for a disk drive is one failureevery five years, if have 100 disks on the system, probability is that will have a disk failure every three weeks
lThis is not good for your bank (or hospital, or inventory control system, or e-commerce system, or . . . ) lAnd many systems now have hundreds of drives attached nUse data redundancy to reduce impact of hardware failures12.32Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRAID Comes in Six LevelsRAID Comes in Six Levels
nLevel 0 - non-redundant striping Data spread across all disks - no redundancy
n Level 1 - mirroring Have duplicate copies of all disks nLevel 2 - memory-style ECC organization Store two or more extra bits to detect and correct errors
nLevel 3 - bit interleaved parity organization Parity on one disk, no striping of data disks
nLevel 4 - block interleaved parity organization Parity on one disk, data spread across n disks
nLevel 5 - block interleaved distributed parity Parity and data spread across all disks - quite common
nLevel 6 - P+Q redundancy Stores extra parity to protect against multiple failures
n Level 0 + 1 Spread data across all disks and mirror Levels 0, 1, & 5 most common, 6 starting to be usedAlso work on 5 + 1
12.33Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRAID Levels
12.34Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRAID 5 DetailsRAID 5 Details
nData and parity are stripedand interleaved across a set of diskslRequires one additional disk per parity setl
Parity is spread across disks in the set to avoid "hot spots"l Parity is the XOR of data in corresponding sectors on other disks of setnIf a disk in set fails, can recreate data by XORing together the corresponding sectors of remaining disks in the setl
If two disks fail, lose all the data on the entire setnReading data requires one I/O (just read the data)nWriting data requires 4 I/Os plus system processing lRead disk block that will be written, read corresponding parity block, XOR data block and parity block together, XOR result with data to be written to produce new parity value, write new data, write new parity 1 2 3 4 P
5 6 7 P 8
9 10 P 11 12
13 P 14 15 16
P 17 18 19 2012.35Silberschatz, Galvin and Gagne ©2009Operating System ConceptsSelecting A Protection LevelSelecting A Protection Level
nRAID 0 (striping) - popular for performance and ease of management no real protection from redundancy nRAID 1 (mirroring) - provides best redundancy protection lBut most expensive in terms of cost for extra disks nRAID 0 and RAID 1 combined - best of both worlds nBecause the hardware required for RAID 1 is expensive, RAID 5 is popular as an alternative lHowever, will not tolerate double failure in same RAID set and often does not perform as well as RAID 1 (since a write requires two reads and two writes - data and parity - although caches help)lHot spares help avoid double failures, but must have spare for each RAID set (and replace spare before a double failure occurs)
nRAID 6 is an improvement, since will tolerate a double failure without requiring 2x the disk space, but not widely available yet12.36Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRAID (0 + 1) and (1 + 0)RAID (0 + 1) and (1 + 0)
12.37Silberschatz, Galvin and Gagne ©2009Operating System ConceptsAdditional Data Protection OptionsAdditional Data Protection Options
nRedundancy through clustering nMaintain a redundant copy of data on another system in cluster lUse remote journaling or similar sort of replication lOr remote mirroring nWhen primary copy of data fails, switch to redundant copy on different system in the cluster nCan also use clustering to protect against system failure (not just disk failure)12.38Silberschatz, Galvin and Gagne ©2009Operating System ConceptsStable-StorageStable-Storage Implementation Implementation
nInformation in Stable Storage is never lost nStable storage is useful or necessary for a number of scenarios lWrite-ahead log scheme requires stable storage. nTo implement stable storage: lReplicate information on more than one nonvolatile storage media with independent failure modes. lUpdate information in a controlled manner to ensure that the stable data can be recovered after any failure during data transfer or recovery.HSuccessful completion
HPartial failure
HTotal failure
HFailures can occur anywhere in I/O hardware path
12.39Silberschatz, Galvin and Gagne ©2009Operating System ConceptsTertiaryTertiary Storage Devices Storage Devices
nExtention of secondary storage lExpand secondary storage - with tradeoffs nLow cost is the defining characteristic of tertiary storage. nGenerally, tertiary storage is built using removable media nTertiary storage is generally slower than secondary storage (disks) nCommon examples of removable media are floppy disks and CDs, DVDs, tape libraries, etc.; other types are available.12.40Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRemovable DisksRemovable Disks
nFloppy disk - thin flexible disk coated with magnetic material, enclosed in a protective plastic case. lMost floppies hold about 1 MB lSimilar technology is used for removable disks that hold more than 1 GB. nSome removable magnetic disks can be nearly as fast as hard disks, but they are at a greater risk of damage from exposure.12.41Silberschatz, Galvin and Gagne ©2009Operating System ConceptsRemovable Disks (Cont.)Removable Disks (Cont.)
nA magneto-optical disk records data on a rigid platter coated with magnetic material. lLaser heat is used to amplify a large, weak magnetic field to record a bit. lLaser light is also used to read data (Kerr effect). lThe magneto-optical head flies much farther from the disk surface than a magnetic disk head, and the magnetic material is covered with a protective layer of plastic or glass; resistant to head crashes. nOptical disks do not use magnetism lThey employ special materials that are altered by laser light nUSB attached, various forms lNVRAM "thumb drives" now commonly used instead of diskettes to save data for short periods of time or transfer between systems lAlso regular "laptop" drives in case with USB adapter lThey appear to the system as a removable disk12.42Silberschatz, Galvin and Gagne ©2009Operating System ConceptsOptical DisksOptical Disks
nWORM ("Write Once, Read Many Times") disks can be written only once. lThin aluminum film sandwiched between two glass or plastic platters.lTo write a bit, the drive uses a laser light to burn a small hole through the aluminum; information can be destroyed, but not altered. (i.e., "burn" a CD)
lVery durable and reliable (all these are very durable) nRead Only disks, such ad CD-ROM and DVD, come from the factory with the data pre-recorded. lPits are pressed, not burned, into media nCD-R and DVD-R disks can be written only once lSimilar to WORM disk but use organic polymer dye nCD-RW and DVD-RW are phase change disks lDisk coated with material that can be crystalline or amorphous - crystalline is more transparent lLaser uses different power levels to read, erase, write