Database System Concepts, 7th Ed.
©Silberschatz, Korth and Sudarshan
See www.db-book.com for conditions on re-use
Cover-6Ed
Chapter 12: Physical Storage Systems

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Classification of Physical Storage Media
§Can differentiate storage into:
•volatile storage: loses contents when power is switched off
•non-volatile storage:
§Contents persist even when power is switched off.
§Includes secondary and tertiary storage, as well as batter-backed up main-memory.
§Factors affecting choice of storage media include
•Speed with which data can be accessed
•Cost per unit of data
•Reliability
•

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Storage Hierarchy

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Storage Hierarchy (Cont.)
§primary storage: Fastest media but volatile (cache, main memory).
§secondary storage: next level in hierarchy, non-volatile, moderately fast access time
•Also called on-line storage
•E.g., flash memory, magnetic disks
§tertiary storage: lowest level in hierarchy, non-volatile, slow access time
•also called off-line storage and used for archival storage
•e.g., magnetic tape, optical storage
•Magnetic tape
§ Sequential access, 1 to 12 TB capacity
§A few drives with many tapes
§Juke boxes with petabytes (1000’s of TB) of storage

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Storage Interfaces
§Disk interface standards families
•SATA (Serial ATA - Advanced Technology Attachment)
§SATA 3 supports data transfer speeds of up to 6 gigabits/sec
•SAS (Serial Attached SCSI – Small Computer System Interconnect - scuzzy)
§SAS Version 3 supports 12 gigabits/sec
•NVMe (Non-Volatile Memory Express) interface
§Works with PCIe connectors to support lower latency and higher transfer rates
§Supports data transfer rates of up to 24 gigabits/sec
§Disks usually connected directly to computer system
§In Storage Area Networks (SAN), a large number of disks are connected by a high-speed network to a
number of servers
§In Network Attached Storage (NAS) networked storage provides a file system interface using
networked file system protocol, instead of providing a disk system interface
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Magnetic Hard Disk Mechanism
Schematic diagram of magnetic disk drive
Photo of magnetic disk drive

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Magnetic Disks
§Read-write head
§Surface of platter divided into circular tracks
•Over 50K-100K tracks per platter on typical hard disks
§Each track is divided into sectors.
•A sector is the smallest unit of data that can be read or written.
•Sector size typically 512 bytes
•Typical sectors per track: 500 to 1000 (on inner tracks) to 1000 to 2000 (on outer tracks)
§To read/write a sector
•disk arm swings to position head on right track
•platter spins continually; data is read/written as sector passes under head
§Head-disk assemblies
•multiple disk platters on a single spindle (1 to 5 usually)
•one head per platter, mounted on a common arm.
§Cylinder i consists of ith track of all the platters

Mention:  checksums, bad sector remapping

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Magnetic Disks (Cont.)
§Disk controller – interfaces between the computer system and the disk drive hardware.
•accepts high-level commands to read or write a sector
•initiates actions such as moving the disk arm to the right track and actually reading or writing
the data
•Computes and attaches checksums to each sector to verify that data is read back correctly
§If data is corrupted, with very high probability stored checksum won’t match recomputed checksum
•Ensures successful writing by reading back sector after writing it
•Performs remapping of bad sectors
n

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Performance Measures of Disks
§Access time – the time it takes from when a read or write request is issued to when data transfer
begins.  Consists of:
•Seek time – time it takes to reposition the arm over the correct track.
§ Average seek time is 1/2 the worst case seek time.
•Would be 1/3 if all tracks had the same number of sectors, and we ignore the time to start and
stop arm movement
§4 to 10 milliseconds on typical disks
•Rotational latency – time it takes for the sector to be accessed to appear under the head.
§4 to 11 milliseconds on typical disks (5400 to 15000 r.p.m.)
§Average latency is 1/2 of the above latency.
•Overall latency is 5 to 20 msec depending on disk model
§Data-transfer rate – the rate at which data can be retrieved from or stored to the disk.
•25 to 200 MB per second max rate, lower for inner tracks

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Performance Measures (Cont.)
§Disk block is a logical unit for storage allocation and retrieval
•4 to 16 kilobytes typically
§Smaller blocks: more transfers from disk
§Larger blocks:  more space wasted due to partially filled blocks
§Sequential access pattern
•Successive requests are for successive disk blocks
•Disk seek required only for first block
§Random access pattern
•Successive requests are for blocks that can be anywhere on disk
•Each access requires a seek
•Transfer rates are low since a lot of time is wasted in seeks
§I/O operations per second (IOPS)
•Number of random block reads that a disk can support per second
•50 to 200 IOPS on current generation magnetic disks

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Performance Measures (Cont.)
§Mean time to failure (MTTF) – the average time the disk is expected to run continuously without
any failure.
•Typically 3 to 5 years
•Probability of failure of new disks is quite low, corresponding to a
“theoretical MTTF” of 500,000 to 1,200,000 hours for a new disk
§E.g., an MTTF of 1,200,000 hours for a new disk means that given 1000 relatively new disks, on an
average one will fail every 1200 hours
•MTTF decreases as disk ages
l
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Flash Storage
§NOR flash vs NAND flash
§NAND flash
•used widely for storage, cheaper than NOR flash
•requires page-at-a-time read (page: 512 bytes to 4 KB)
§20 to 100 microseconds for a page read
§Not much difference between sequential and random read
•Page can only be written once
§Must be erased to allow rewrite
§Solid state disks
•Use standard block-oriented disk interfaces, but store data on multiple flash storage devices
internally
•Transfer rate of up to 500 MB/sec using SATA, and
up to 3 GB/sec using NVMe PCIe

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Flash Storage (Cont.)
§Erase happens in units of erase block
•Takes 2 to 5 millisecs
•Erase block typically 256 KB to 1 MB (128 to 256 pages)
§Remapping of logical page addresses to physical page addresses avoids waiting for erase
§Flash translation table tracks mapping
•also stored in a label field of flash page
•remapping carried out by flash translation layer
•
•
•
•
l
§After 100,000 to 1,000,000 erases, erase block becomes unreliable and cannot be used
•wear leveling

Where is FTL stored?  How is it reconstructed on system startup?
How to track number of erases per block (extra storage per block)

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
SSD Performance Metrics
§Random reads/writes per second
•Typical 4 KB reads:  10,000 reads per second (10,000 IOPS)
•Typical  4KB writes: 40,000 IOPS
•SSDs support parallel reads
§Typical 4KB reads:
•100,000 IOPS with 32 requests in parallel (QD-32) on SATA
•350,000 IOPS with QD-32 on NVMe PCIe
§Typical 4KB writes:
•100,000 IOPS with QD-32, even higher on some models
§ Data transfer rate for sequential reads/writes
•400 MB/sec for SATA3, 2 to 3 GB/sec using NVMe PCIe
§Hybrid disks: combine small amount of flash cache with larger magnetic disk

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Storage Class Memory
§3D-XPoint memory technology pioneered by Intel
§Available as Intel Optane
•SSD interface shipped from 2017
§Allows lower latency than flash SSDs
•Non-volatile memory interface announced in 2018
§Supports direct access to words, at speeds comparable to main-memory speeds

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID
§RAID: Redundant Arrays of Independent Disks
•disk organization techniques that manage a large numbers of disks, providing a view of a single
disk of
§high capacity and high speed  by using multiple disks in parallel,
§high reliability by storing data redundantly, so that data can be recovered even if  a disk fails
§The chance that some disk out of a set of N disks will fail is much higher than the chance that a
specific single disk will fail.
•E.g., a system with 100 disks, each with MTTF of 100,000 hours (approx.  11 years), will have a
system MTTF of 1000 hours (approx. 41 days)
•Techniques for using redundancy to avoid data loss are critical with large numbers of disks

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Improvement of Reliability via Redundancy
§Redundancy – store extra information that can be used to rebuild information lost in a disk
failure
§E.g., Mirroring (or shadowing)
•Duplicate every disk.  Logical disk consists of two physical disks.
•Every write is carried out on both disks
§Reads can take place from either disk
•If one disk in a pair fails, data still available in the other
§Data loss would occur only if a disk fails, and its mirror disk also fails before the system is
repaired
•Probability of combined event is very small
§Except for dependent failure modes such as fire or building collapse or electrical power surges
§Mean time to data loss depends on mean time to failure,
and mean time to repair
•E.g., MTTF of 100,000 hours, mean time to repair of 10 hours gives mean time to data loss of
500*106 hours (or 57,000 years) for a mirrored pair of disks (ignoring dependent failure modes)
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Improvement in Performance via Parallelism
§Two main goals of parallelism in a disk system:
l1. Load balance multiple small accesses to increase throughput
l2. Parallelize large accesses to reduce response time.
§Improve transfer rate by striping data across multiple disks.
§Bit-level striping – split the bits of each byte across multiple disks
•In an array of eight disks, write bit i of each byte to disk i.
•Each access can read data at eight times the rate of a single disk.
•But seek/access time worse than for a single disk
§Bit level striping is not used much any more
§Block-level striping – with n disks, block i of a file goes to disk (i mod n) + 1
•Requests for different blocks can run in parallel if the blocks reside on different disks
•A request for a long sequence of blocks can utilize all disks in parallel

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID Levels
§Schemes to provide redundancy at lower cost by using disk striping combined with parity bits
•Different RAID organizations, or RAID levels, have differing cost, performance and reliability
characteristics
§RAID Level 0:  Block striping; non-redundant.
• Used in high-performance applications where data loss is not critical.
§RAID Level 1:  Mirrored disks with block striping
•Offers best write performance.
•Popular for applications such as storing log files in a database system.
§
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID Levels (Cont.)
§Parity blocks: Parity block j stores XOR of bits from block j  of each disk
•When writing data to a block j, parity block j must also be computed and written to disk
§Can be done by using old parity block, old value of current block and new value of current block
(2 block reads + 2 block writes)
§Or by recomputing the parity value using the new values of blocks corresponding to the parity
block
•More efficient for writing large amounts of data sequentially
•To recover data for a block, compute XOR of bits from all other blocks in the set including the
parity block
•
§
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID Levels (Cont.)
§RAID Level 5:  Block-Interleaved Distributed Parity; partitions data and parity among all N + 1
disks, rather than storing data in N disks and parity in 1 disk.
•E.g., with 5 disks, parity block for nth set of blocks is stored on disk (n mod 5) + 1, with the
data blocks stored on the other 4 disks.

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID Levels (Cont.)
§RAID Level 5 (Cont.)
•Block writes occur in parallel if the blocks and their parity blocks are on different disks.
§RAID Level 6: P+Q Redundancy scheme; similar to Level 5, but stores two error correction blocks
(P, Q) instead of single parity block to guard against multiple disk failures.
• Better reliability than Level 5 at a higher cost
§Becoming more important as storage sizes increase

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
RAID Levels (Cont.)
§Other levels (not used in practice):
•RAID Level 2:  Memory-Style Error-Correcting-Codes (ECC) with bit striping.
•RAID Level 3: Bit-Interleaved Parity
•RAID Level 4:  Block-Interleaved Parity; uses block-level striping, and keeps a parity block on a
separate parity disk for corresponding blocks from N other disks.
§RAID 5 is better than RAID 4, since with RAID 4 with random writes, parity disk gets much higher
write load than other disks and becomes a bottleneck
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Choice of RAID Level
§Factors in choosing RAID level
•Monetary cost
•Performance: Number of I/O operations per second, and bandwidth during normal operation
•Performance during failure
•Performance during rebuild of failed disk
§Including time taken to rebuild failed disk
§RAID 0 is used only when data safety is not important
•E.g., data can be recovered quickly from other sources

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Choice of RAID Level (Cont.)
§Level 1 provides much better write performance than level 5
•Level 5 requires at least 2 block reads and 2 block writes to write a single block, whereas Level
1 only requires 2 block writes
§Level 1 had higher storage cost than level 5
§Level 5 is preferred for applications where writes are sequential and large (many blocks), and
need large amounts of data storage
§RAID 1 is preferred for applications with many random/small updates
§Level 6 gives better data protection than RAID 5 since it can tolerate two disk (or disk block)
failures
•Increasing in importance since latent block failures on one disk, coupled with a failure of
another disk can result in data loss with RAID 1 and RAID 5.

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Hardware Issues
§Software RAID:  RAID implementations done entirely in software, with no special hardware support
§Hardware RAID:  RAID implementations with special hardware
•Use non-volatile RAM to record writes that are being executed
•Beware:  power failure during write can result in corrupted disk
§E.g., failure after writing one block but before writing the second in a mirrored system
§Such corrupted data must be detected when power is restored
•Recovery from corruption is similar to recovery from failed disk
•NV-RAM helps to efficiently detected potentially corrupted blocks
§Otherwise all blocks of disk must be read and compared with mirror/parity block
•

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Hardware Issues (Cont.)
§Latent failures: data successfully written earlier gets damaged
•can result in data loss even if only one disk fails
§Data scrubbing:
•continually scan for latent failures, and recover from copy/parity
§Hot swapping: replacement of disk while system is running, without power down
•Supported by some hardware RAID systems,
•reduces time to recovery, and improves availability greatly
§Many systems maintain spare disks which are kept online, and used as replacements for failed disks
immediately on detection of failure
•Reduces time to recovery greatly
§Many hardware RAID systems ensure that a single point of failure will not stop the functioning of
the system by using
•Redundant power supplies with battery backup
•Multiple controllers and multiple interconnections to guard against controller/interconnection
failures

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Optimization of Disk-Block Access
§Buffering: in-memory buffer to cache disk blocks
§Read-ahead: Read extra blocks from a track in anticipation that they will be requested soon
§Disk-arm-scheduling algorithms re-order block requests so that disk arm movement is minimized
•elevator algorithm
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Optimization of Disk-Block Access
§Buffering: in-memory buffer to cache disk blocks
§Read-ahead: Read extra blocks from a track in anticipation that they will be requested soon
§Disk-arm-scheduling algorithms re-order block requests so that disk arm movement is minimized
•elevator algorithm
§
R1
R5
R2
R4
R3
R6
Inner track
Outer track

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Optimization of Disk Block Access (Cont.)
§File organization
• Allocate blocks of a file in as contiguous a manner as possible
•Allocation in units of extents
•Files may get fragmented
§E.g., if free blocks on disk are scattered, and newly created file has its blocks scattered over
the disk
§Sequential access to a fragmented file results in increased disk arm movement
§Some systems have utilities to defragment the file system, in order to speed up file access
§Non-volatile write buffers
§
§

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
End of Chapter 12

©Silberschatz, Korth and Sudarshan
12.‹#›
Database System Concepts - 7th Edition
Cover-6Ed
Magnetic Tapes
§Hold large volumes of data and provide high transfer rates
•Few GB for DAT (Digital Audio Tape) format, 10-40 GB with DLT (Digital Linear Tape) format, 100
GB+ with Ultrium format, and 330 GB with Ampex helical scan format
•Transfer rates from few to 10s of MB/s
§Tapes are cheap, but cost of drives is very high
§Very slow access time in comparison to magnetic and optical disks
•limited to sequential access.
•Some formats (Accelis) provide faster seek (10s of seconds) at cost of lower capacity
§Used mainly for backup, for storage of infrequently used information, and as an off-line medium
for transferring information from one system to another.
§Tape jukeboxes used for very large capacity storage
•Multiple petabyes (1015 bytes)