Chapter 12: Mass-Storage Systems
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
Silberschatz, Galvin and Gagne ©2007
Chapter 12: Mass-Storage Systems












Overview of Mass Storage Structure
Disk Structure
Disk Formatting
Disk Scheduling
Disk Management
RAID
Stable Storage
Swap-Space Management
Disk Attachment
Tertiary Storage Devices
Operating System Issues
Performance Issues
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.2
Silberschatz, Galvin and Gagne ©2007
Objectives



Describe the physical structure of secondary and
tertiary storage devices and the resulting effects
on the uses of the devices
Explain the performance characteristics of massstorage devices
Discuss operating-system services provided for
mass storage, including RAID and HSM
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.3
Silberschatz, Galvin and Gagne ©2007
Overview of Mass Storage Structure


Magnetic disks provide bulk of secondary storage of modern
computers
 Drives rotate at 60 to 200 times per second
 Transfer rate is rate at which data flow between drive and
computer
 Positioning 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)
 Head crash results from disk head making contact with the
disk surface
Drive attached to computer via I/O bus
 Busses vary, including EIDE, ATA, SATA, USB, SCSI

Host controller in computer uses bus to talk to disk
controller built into drive or storage array
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.4
Silberschatz, Galvin and Gagne ©2007
Overview of Mass Storage (Cont.)

Magnetic tape
 Was early secondary-storage medium
 Relatively permanent and holds large quantities of data
 Access time slow
 Random access ~1000 times slower than disk
 Mainly used for backup, storage of infrequently-used data,
transfer medium between systems
 Kept in spool and wound or rewound past read-write head
 Once data under head, transfer rates comparable to disk
 20-200GB typical storage
 Common technologies are 4mm, 8mm, 19mm, LTO-2 and
SDLT
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.5
Silberschatz, Galvin and Gagne ©2007
What Does The Disk Look Like?
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.6
Silberschatz, Galvin and Gagne ©2007
Some Parameters

2-30 heads (platters * 2)
 diameter 14’’ to 2.5’’
700-20480 tracks per surface
16-1600 sectors per track
Sector size:
 64-8k bytes
 512 for most PCs
 Note: inter-sector gaps
Capacity: 20M-300G

Main adjectives: BIG, slow




Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.7
Silberschatz, Galvin and Gagne ©2007
Disk Overheads


To read from disk, we must specify:
 cylinder #, surface #, sector #, transfer size, memory address
Latency includes:
 Seek time: to get to the track
 Latency time: to get to the sector (rotation delay)
 Transfer time: get bits off the disk
Track
Sector
Rotation
Delay
Seek Time
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.8
Silberschatz, Galvin and Gagne ©2007
Modern Disks
Barracuda 180
Cheetah X15 36LP
Capacity
181GB
36.7GB
Disk/Heads
12/24
4/8
Cylinders
24,247
18,479
Sectors/track
~609
~485
Speed
7200 RPM
15000 RPM
Rotational latency (ms)
4.17
2.0
Avg seek (ms)
7.4
3.6
Track-2-track(ms)
0.8
0.3
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.9
Silberschatz, Galvin and Gagne ©2007
50 Years Ago…

On 13th September 1956, IBM 305 RAMAC computer
system first to use disk storage

80000 times more data on the 8GB 1-inch drive in his
right hand than on the 24-inch RAMAC one in his left…
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.10
Silberschatz, Galvin and Gagne ©2007
Disks vs. Memory






Smallest write: sector
Atomic write = sector
Random access: 5ms
 not on a good curve
Sequential access: 200MB/s
Cost $.002MB
Crash: doesn’t matter
(“non-volatile”)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006






(usually) bytes
byte, word
50 ns
 faster all the time
200-1000MB/s
$.10MB
contents gone (“volatile”)
12.11
Silberschatz, Galvin and Gagne ©2007
Disk Structure



Disk drives addressed as 1-dim arrays of logical blocks

the logical block is the smallest unit of transfer

Logical block addressing
This array mapped sequentially onto disk sectors

Address 0 is 1st sector of 1st track of the outermost
cylinder

Addresses incremented within track, then within
tracks of the cylinder, then across cylinders, from
innermost to outermost
Translation is theoretically possible, but usually difficult

Some sectors might be defective

Number of sectors per track is not a constant
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.12
Silberschatz, Galvin and Gagne ©2007
Non-Uniform # Sectors / Track


Maintain same data rate with Constant Linear Velocity
Approaches:
 Reduce bit density per track for outer layers
 Have more sectors per track on the outer layers (virtual
geometry)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.13
Silberschatz, Galvin and Gagne ©2007
How is the Disk Formatted?



After manufacturing disk has no information
 Is stack of platters coated with magnetizable metal oxide
Before use, each platter receives low-level format
 Format has series of concentric tracks
 Each track contains some sectors
 There is a short gap between sectors
Preamble allows h/w to recognize start of sector
 Also contains cylinder and sector numbers
 Data is usually 512 bytes
 ECC field used to detect and recover from read errors
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.14
Silberschatz, Galvin and Gagne ©2007
Cylinder Skew

Why cylinder skew?

How much skew?
Example, if
 10000 rpm
 Drive rotates in 6 ms
 Track has 300 sectors
 New sector every 20 µs
 If track seek time 800 µs
 40 sectors pass on seek

Cylinder skew: 40 sectors
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.15
Silberschatz, Galvin and Gagne ©2007
Formatting and Performance


If 10K rpm, 300 sectors of 512 bytes per track
 153,600 bytes every 6 ms  24.4 MB/sec transfer
If disk controller buffer can store only one sector
 For 2 consecutive reads, 2nd sector flies past during
memory transfer of 1st track
 Idea: Use single/double interleaving
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.16
Silberschatz, Galvin and Gagne ©2007
Disk Partitioning




Each partition is like a separate disk
Sector 0 is Master Boot Record (MBR)
 Contains boot code + partition table (starting sector and size)
High-level formatting
 Done for each partition
 Specifies boot block, free list, root directory, empty file system
What happens on boot?
 BIOS loads MBR, boot program checks to see active partition
 Reads boot sector from that partition that then boots OS kernel,
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.17
Silberschatz, Galvin and Gagne ©2007
Handling Errors


A disk track with a bad sector
Solutions:
 Substitute a spare for the bad sector (sector sparing)
 Shift all sectors to bypass bad one (sector forwarding)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.18
Silberschatz, Galvin and Gagne ©2007
Disk Scheduling





The operating system tries to use hardware efficiently
 for disk drives  having fast access time, disk
bandwidth
Access time has two major components
 Seek time is time to move the heads to the cylinder
containing the desired sector
 Rotational latency is additional time waiting to rotate
the desired sector to the disk head.
Want to minimize seek time
Seek time  seek distance
Disk bandwidth is total number of bytes transferred,
divided by the total time between the first request for
service and the completion of the last transfer.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.19
Silberschatz, Galvin and Gagne ©2007
Disk Scheduling (Cont.)


Several algorithms exist to schedule the servicing
of disk I/O requests.
We illustrate them with a request queue (0-199).
98, 183, 37, 122, 14, 124, 65, 67
Head pointer starts at 53
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.20
Silberschatz, Galvin and Gagne ©2007
FCFS (First-Come First Serve)
Illustration shows total head movement of 640 cylinders.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.21
Silberschatz, Galvin and Gagne ©2007
SSTF (Shortest Seek Time First)



Selects request with minimum seek time from current
head position
SSTF scheduling is a form of SJF scheduling
 may cause starvation of some requests.
Illustration shows total head movement of 236 cylinders.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.22
Silberschatz, Galvin and Gagne ©2007
SSTF (Cont.)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.23
Silberschatz, Galvin and Gagne ©2007
SCAN



The disk arm starts at one end of the disk,
 moves toward the other end, servicing requests
 head movement is reversed when it gets to the other
end of disk
 servicing continues.
Sometimes called the elevator algorithm.
Illustration shows total head movement of 208 cylinders.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.24
Silberschatz, Galvin and Gagne ©2007
SCAN (Cont.)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.25
Silberschatz, Galvin and Gagne ©2007
C-SCAN



Provides a more uniform wait time than SCAN.
The head moves from one end of the disk to the other.
 servicing requests as it goes.
 When it reaches the other end it immediately returns
to beginning of the disk
 No requests serviced on the return trip.
Treats the cylinders as a circular list
 that wraps around from the last cylinder to the first
one.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.26
Silberschatz, Galvin and Gagne ©2007
C-SCAN (Cont.)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.27
Silberschatz, Galvin and Gagne ©2007
C-LOOK


Version of C-SCAN
Arm only goes as far as last request in each direction,
 then reverses direction immediately,
 without first going all the way to the end of the disk.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.28
Silberschatz, Galvin and Gagne ©2007
C-LOOK (Cont.)
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.29
Silberschatz, Galvin and Gagne ©2007
Selecting a Good Algorithm

SSTF is common and has a natural appeal

SCAN and C-SCAN perform better under heavy load

Performance depends on number and types of requests

Requests for disk service can be influenced by the fileallocation method.

Disk-scheduling algo should be a separate OS module


allowing it to be replaced with a different algorithm if
necessary.
Either SSTF or C-LOOK is a reasonable default algo
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.30
Silberschatz, Galvin and Gagne ©2007
RAID Motivation





Disks are improving, but not as fast as CPUs
 1970s seek time: 50-100 ms.
 2000s seek time: <5 ms.
 Factor of 20 improvement in 3 decades
Use multiple disks to improve performance
 By striping files across multiple disks (placing parts of each
file on a different disk), parallel I/O can improve access time
Problem: striping reduces reliability
 100 disks have 1/100th the reliability of one disk
 MTBF: mean time between failures
So, we need striping for performance, but we need something to
help with reliability / availability
To improve reliability, we can add redundant data to the disks, in
addition to striping
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.31
Silberschatz, Galvin and Gagne ©2007
RAID





RAID: Redundant Array of Inexpensive Disks
 In industry, “I” is for “Independent”
 Alternative: SLED, single large expensive disk
Disks are small and cheap, so it’s easy to put lots of
disks (10s to 100s) in one box for increased storage,
performance, and availability
The RAID box with a RAID controller looks just like a
SLED to the computer
Data plus some redundant information is striped
across the disks in some way
How that striping is done is key to performance and
reliability.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.32
Silberschatz, Galvin and Gagne ©2007
RAID Level 0





Level 0 is nonredundant disk
array
Files are striped across disks,
no redundant info
 Stripe is 1-N sectors,
sometimes also called a
block
High read throughput
Best write throughput (no
redundant info to write)
Any disk failure results in data
loss
 Reliability worse than SLED
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.33
Silberschatz, Galvin and Gagne ©2007
RAID Level 1




Mirrored Disks
Data is written to two
places
 On failure, just use
surviving disk
On read, choose fastest
disk to read
 Write performance is
same as single drive,
read performance is
2x better
Expensive
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.34
Silberschatz, Galvin and Gagne ©2007
Parity and Hamming Code



What do you need to do in order to detect and correct a one-bit
error ?
 Suppose you have a binary number, represented as a
collection of bits: <b3, b2, b1, b0>, e.g. 0110
Detection is easy
Parity:
 Count the number of bits that are on, see if it’s odd or even
 EVEN parity is 0 if the number of 1 bits is even
 Parity(<b3, b2, b1, b0 >) = P0 = b0  b1  b2  b3
 Parity(<b3, b2, b1, b0, p0>) = 0 if all bits are intact
 Parity(0110) = 0, Parity(01100) = 0
 Parity(11100) = 1 => ERROR!
 Parity can detect a single error, but can’t tell you which of
the bits got flipped
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.35
Silberschatz, Galvin and Gagne ©2007
Parity and Hamming Code



Detection and correction require more work
Hamming codes can detect double bit errors and detect &
correct single bit errors
7/4 Hamming Code
 h0 = b0  b1  b3
 h1 = b0  b2  b3
 h2 = b1  b2  b3
 H0(<1101>) = 0
 H1(<1101>) = 1
 H2(<1101>) = 0
 Hamming(<1101>) = <b3, b2, b1, h2, b0, h1, h0> =
<1100110>
 If a bit is flipped, e.g. <1110110>
 Hamming(<1111>) = <h2, h1, h0> = <111> compared to
<010>, <101> are in error. Error occurred in bit 5.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.36
Silberschatz, Galvin and Gagne ©2007
RAID Level 2





Bit-level Striping with Hamming (ECC) codes for error correction
All 7 disk arms are synchronized and move in unison
Complicated controller
Single access at a time
Tolerates only one error, but with no performance degradation
Bit 0
Bit 1
Bit 2
Bit 3
Bit 4
data disks
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
Bit 5
Bit 6
ECC disks
12.37
Silberschatz, Galvin and Gagne ©2007
RAID Level 3
Single parity disk
can be used
to detect and
correct errors
after crash





Simplified version of RAID level 2; drives again synchronized
Uses a parity disk
 Each bit on the parity disk is a parity function of the corresponding
bits on all the other disks
A read accesses all the data disks
A write accesses all data disks plus the parity disk
On disk failure, read remaining disks plus parity disk to compute the
missing data
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.38
Silberschatz, Galvin and Gagne ©2007
RAID Level 4




Combines level 0
and 3 – block-level
parity with stripes
 Parity stripe is
XOR of all other
stripes
A read accesses all
the data disks
A write accesses all
data disks plus the
parity disk
Heavy load on the
parity disk
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.39
Silberschatz, Galvin and Gagne ©2007
RAID Level 5



Block Interleaved
Distributed Parity
Like parity scheme,
but distribute the
parity info over all
disks (as well as data
over all disks)
Better read
performance, large
write performance
 Reads can
outperform SLEDs
and RAID-0
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.40
Silberschatz, Galvin and Gagne ©2007
RAID Level 6



Level 5 with an extra parity bit
Can tolerate two failures
 What are the odds of having two concurrent
failures ?
May outperform Level-5 on reads, slower on writes
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.41
Silberschatz, Galvin and Gagne ©2007
Nested RAID: RAID 0 + 1

RAID 0 + 1:
 Set of disks are striped,
stripe is mirrored to 2nd
array
 Performance of RAID 0,
reliability of RAID 1
 Expensive
 If First RAID 0 A2 fails,
fall back to 2nd RAID 0
disk
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.42
Silberschatz, Galvin and Gagne ©2007
Nested RAID: RAID 1 + 0

RAID 1 + 0:
 Disks are mirrored first,
then striped
 If disk fails, mirror is still
available
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.43
Silberschatz, Galvin and Gagne ©2007
Stable Storage

Handling disk write errors:
 Write lays down bad data
 Crash during a write corrupts original data

What we want to achieve? Stable Storage
 When a write is issued, the disk either correctly writes data, or it
does nothing, leaving existing data intact

Model:
 An incorrect disk write can be detected by looking at the ECC
 It is very rare that same sector goes bad on multiple disks
 CPU is fail-stop
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.44
Silberschatz, Galvin and Gagne ©2007
Stable Storage: Approach


Use 2 identical disks
 corresponding blocks on both drives are the same
3 operations:
 Stable write: retry on 1st until successful, then try 2nd disk
 Stable read: read from 1st. If ECC error, then try 2nd
 Crash recovery: scan corresponding blocks on both disks
 If one block is bad, replace with good one
 If both are good, replace block in 2nd with the one in 1st
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.45
Silberschatz, Galvin and Gagne ©2007
Network-Attached Storage



Network-attached storage (NAS) is storage made available over a
network rather than over a local connection (such as a bus)
 NFS and CIFS are common protocols
Implemented via remote procedure calls (RPCs) between host and
storage
New iSCSI protocol uses IP network to carry the SCSI protocol
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.46
Silberschatz, Galvin and Gagne ©2007
Storage Area Network


Common in large storage environments (and
becoming more common)
Multiple hosts attached to multiple storage arrays flexible
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.47
Silberschatz, Galvin and Gagne ©2007
Swap-Space Management



Swap-space — Virtual memory uses disk space as an
extension of main memory.
Swap-space can be carved out of the normal file system,or,
more commonly, it can be in a separate disk partition.
Swap-space management
 4.3BSD allocates swap space when process starts; holds
text segment (the program) and data segment.
 Kernel uses swap maps to track swap-space use.
 Solaris 2 allocates swap space only when a page is
forced out of physical memory, not when the virtual
memory page is first created.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.48
Silberschatz, Galvin and Gagne ©2007
Swapping on Linux



Linux uses 4K page slots
Swap can be in dedicated swap partition or swap file on FS
Numbers above indicate ref count; e.g., 3 means 3 processes
sharing
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.49
Silberschatz, Galvin and Gagne ©2007
Tertiary Storage Devices

Low cost is the defining characteristic of tertiary
storage.

Generally, tertiary storage is built using removable
media

Common examples of removable media are floppy
disks and CD-ROMs; other types are available.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.50
Silberschatz, Galvin and Gagne ©2007
Removable Disks

Floppy disk — thin flexible disk coated with
magnetic material, enclosed in a protective plastic
case.


Most floppies hold about 1 MB; similar
technology is used for removable disks that hold
more than 1 GB.
Removable magnetic disks can be nearly as fast
as hard disks, but they are at a greater risk of
damage from exposure.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.51
Silberschatz, Galvin and Gagne ©2007
Removable Disks (Cont.)

A magneto-optic disk records data on a rigid platter coated
with magnetic material.
 Laser heat is used to amplify a large, weak magnetic field
to record a bit.
 Laser light is also used to read data (Kerr effect).
 The magneto-optic 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.

Optical disks do not use magnetism; they employ special
materials that are altered by laser light.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.52
Silberschatz, Galvin and Gagne ©2007
WORM Disks






The data on read-write disks can be modified over and over.
WORM (“Write Once, Read Many Times”) disks can be
written only once.
Thin aluminum film sandwiched between two glass or plastic
platters.
To write a bit, the drive uses a laser light to burn a small hole
through the aluminum; information can be destroyed by not
altered.
Very durable and reliable.
Read Only disks, such ad CD-ROM and DVD, come from
the factory with the data pre-recorded.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.53
Silberschatz, Galvin and Gagne ©2007
Tapes




Compared to a disk, a tape is less expensive and holds more
data, but random access is much slower.
Tape is an economical medium for purposes that do not
require fast random access, e.g., backup copies of disk data,
holding huge volumes of data.
Large tape installations typically use robotic tape changers
that move tapes between tape drives and storage slots in a
tape library.
 stacker – library that holds a few tapes
 silo – library that holds thousands of tapes
A disk-resident file can be archived to tape for low cost
storage; the computer can stage it back into disk storage for
active use.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.54
Silberschatz, Galvin and Gagne ©2007
Application Interface





Most OSs handle removable disks almost exactly like fixed
disks — a new cartridge is formatted and an empty file system
is generated on the disk.
Tapes are presented as a raw storage medium, i.e., and
application does not not open a file on the tape, it opens the
whole tape drive as a raw device.
Usually the tape drive is reserved for the exclusive use of that
application.
Since the OS does not provide file system services, the
application must decide how to use the array of blocks.
Since every application makes up its own rules for how to
organize a tape, a tape full of data can generally only be used
by the program that created it.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.55
Silberschatz, Galvin and Gagne ©2007
Tape Drives






The basic operations for a tape drive differ from
those of a disk drive.
locate positions the tape to a specific logical block,
not an entire track (corresponds to seek).
The read position operation returns the logical
block number where the tape head is.
The space operation enables relative motion.
Tape drives are “append-only” devices; updating a
block in the middle of the tape also effectively
erases everything beyond that block.
An EOT mark is placed after a block that is written.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.56
Silberschatz, Galvin and Gagne ©2007
Naming on Removable Media



The issue of naming files on removable media is
especially difficult when we want to write data on a
removable cartridge on one computer, and then
use the cartridge in another computer.
Contemporary OSs generally leave the name
space problem unsolved for removable media, and
depend on applications and users to figure out
how to access and interpret the data.
Some kinds of removable media (e.g., CDs) are so
well standardized that all computers use them the
same way.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.57
Silberschatz, Galvin and Gagne ©2007
Hierarchical Storage Management



A hierarchical storage system extends the storage hierarchy
beyond primary memory and secondary storage to incorporate
tertiary storage — usually implemented as a jukebox of tapes
or removable disks.
Usually incorporate tertiary storage by extending the file
system.
 Small and frequently used files remain on disk.
 Large, old, inactive files are archived to the jukebox.
HSM is usually found in supercomputing centers and other
large installations that have enormous volumes of data.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.58
Silberschatz, Galvin and Gagne ©2007
Speed in Tertiary Storage

Two aspects of speed in tertiary storage are
bandwidth and latency.

Bandwidth is measured in bytes per second.
 Sustained bandwidth – average data rate during
a large transfer; # of bytes/transfer time.
Data rate when the data stream is actually
flowing.
 Effective bandwidth – average over the entire
I/O time, including seek or locate, and cartridge
switching.
Drive’s overall data rate.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.59
Silberschatz, Galvin and Gagne ©2007
Speed in Tertiary Storage (Cont.)



Access latency – amount of time needed to locate data.
 Access time for a disk – move the arm to the selected
cylinder and wait for the rotational latency; < 35
milliseconds.
 Access on tape requires winding the tape reels until the
selected block reaches the tape head; tens or hundreds of
seconds.
 Generally say that random access within a tape cartridge
is about a thousand times slower than random access on
disk.
The low cost of tertiary storage is a result of having many
cheap cartridges share a few expensive drives.
A removable library is best devoted to the storage of
infrequently used data, because the library can only satisfy a
relatively small number of I/O requests per hour.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.60
Silberschatz, Galvin and Gagne ©2007
Reliability



A fixed disk drive is likely to be more reliable
than a removable disk or tape drive.
An optical cartridge is likely to be more reliable
than a magnetic disk or tape.
A head crash in a fixed hard disk generally
destroys the data, whereas the failure of a tape
drive or optical disk drive often leaves the data
cartridge unharmed.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.61
Silberschatz, Galvin and Gagne ©2007
Disk Attachment



Host-attached storage accessed through I/O ports talking to
I/O busses
SCSI itself is a bus, up to 16 devices on one cable, SCSI
initiator requests operation and SCSI targets perform tasks
 Each target can have up to 8 logical units (disks attached
to device controller
Fiber Channel is high-speed serial architecture
 Can be switched fabric with 24-bit address space – the
basis of storage area networks (SANs) in which many
hosts attach to many storage units
 Can be arbitrated loop (FC-AL) of 126 devices
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.62
Silberschatz, Galvin and Gagne ©2007
Cost




Main memory is much more expensive than
disk storage
The cost per megabyte of hard disk storage is
competitive with magnetic tape if only one tape
is used per drive.
The cheapest tape drives and the cheapest disk
drives have had about the same storage
capacity over the years.
Tertiary storage gives a cost savings only when
the number of cartridges is considerably larger
than the number of drives.
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.63
Silberschatz, Galvin and Gagne ©2007
DRAM Price per MB, 1981 to 2004
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.64
Silberschatz, Galvin and Gagne ©2007
Hard Disk Price per MB, 1981 to 2004
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.65
Silberschatz, Galvin and Gagne ©2007
Tape Price per MB, 1984-2004
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.66
Silberschatz, Galvin and Gagne ©2007
End of Chapter 12
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
Silberschatz, Galvin and Gagne ©2007
CD-ROMs (1)
Spiral makes 22,188 revolutions around disk (approx 600/mm).
Will be 5.6 km long. Rotation rate: 530 rpm to 200 rpm
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.68
Silberschatz, Galvin and Gagne ©2007
CD-ROMs (2)
Logical data layout on a CD-ROM.
Picture from Tanenbaum, Modern Operating Systems 3 e, (c) 2008 Prentice-Hall, Inc. All rights reserved. 0-13-6006639
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.69
Silberschatz, Galvin and Gagne ©2007
CD-Recordables
Cross section of a CD-R disk and laser. A silver
CD-ROM has similar structure, except without dye layer and with
pitted aluminum layer instead of gold layer.
Picture from Tanenbaum, Modern Operating Systems 3 e, (c) 2008 Prentice-Hall, Inc. All rights reserved. 0-13-6006639
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.70
Silberschatz, Galvin and Gagne ©2007
DVD (1)
DVD Improvements on CDs
1.
2.
3.
Smaller pits
(0.4 microns versus 0.8 microns for CDs).
A tighter spiral
(0.74 microns between tracks versus 1.6
microns for CDs).
A red laser
(at 0.65 microns versus 0.78 microns for CDs).
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.71
Silberschatz, Galvin and Gagne ©2007
DVD (2)
DVD Formats
1.
2.
3.
4.
Single-sided, single-layer (4.7 GB).
Single-sided, dual-layer (8.5 GB).
Double-sided, single-layer (9.4 GB).
Double-sided, dual-layer (17 GB).
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.72
Silberschatz, Galvin and Gagne ©2007
DVD (3)
A double-sided, dual-layer DVD disk.
Picture from Tanenbaum, Modern Operating Systems 3 e, (c) 2008 Prentice-Hall, Inc. All rights reserved. 0-13-6006639
Operating System Concepts with Java – 7th Edition, Nov 15, 2006
12.73
Silberschatz, Galvin and Gagne ©2007