cs 245: database system principles notes 03: disk organization
DESCRIPTION
CS 245: Database System Principles Notes 03: Disk Organization. Hector Garcia-Molina. Topics for today. How to lay out data on disk How to move it to memory. What are the data items we want to store?. a salary a name a date a picture. What we have available: Bytes. 8 bits. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/1.jpg)
CS 245 Notes 3 1
CS 245: Database System Principles
Notes 03: Disk Organization
Hector Garcia-Molina
![Page 2: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/2.jpg)
CS 245 Notes 3 2
• How to lay out data on disk• How to move it to memory
Topics for today
![Page 3: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/3.jpg)
CS 245 Notes 3 3
What are the data items we want to store?
• a salary• a name• a date• a picture
![Page 4: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/4.jpg)
CS 245 Notes 3 4
What are the data items we want to store?
• a salary• a name• a date• a pictureWhat we have available: Bytes
8bits
![Page 5: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/5.jpg)
CS 245 Notes 3 5
To represent:
• Integer (short): 2 bytese.g., 35 is
00000000 00100011
• Real, floating pointn bits for mantissa, m for exponent….
![Page 6: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/6.jpg)
CS 245 Notes 3 6
• Characters various coding schemes suggested,
most popular is ascii
To represent:
Example:A: 1000001a: 11000015: 0110101LF: 0001010
![Page 7: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/7.jpg)
CS 245 Notes 3 7
• Booleane.g., TRUE
FALSE
1111 1111
0000 0000
To represent:
• Application specifice.g., RED 1 GREEN 3
BLUE 2 YELLOW 4 …
![Page 8: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/8.jpg)
CS 245 Notes 3 8
• Booleane.g., TRUE
FALSE
1111 1111
0000 0000
To represent:
• Application specifice.g., RED 1 GREEN 3
BLUE 2 YELLOW 4 …Can we use less than 1
byte/code?Yes, but only if desperate...
![Page 9: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/9.jpg)
CS 245 Notes 3 9
• Datese.g.: - Integer, # days since Jan 1, 1900
- 8 characters, YYYYMMDD - 7 characters, YYYYDDD
(not YYMMDD! Why?)• Time
e.g. - Integer, seconds since midnight - characters, HHMMSSFF
To represent:
![Page 10: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/10.jpg)
CS 245 Notes 3 10
• String of characters– Null terminated
e.g.,
– Length givene.g.,
- Fixed length
c ta
c ta3
To represent:
![Page 11: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/11.jpg)
CS 245 Notes 3 11
• Bag of bits
Length Bits
To represent:
![Page 12: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/12.jpg)
CS 245 Notes 3 12
Key Point
• Fixed length items
• Variable length items- usually length given at beginning
![Page 13: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/13.jpg)
CS 245 Notes 3 13
• Type of an item: Tells us how to interpret(plus size if
fixed)
Also
![Page 14: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/14.jpg)
CS 245 Notes 3 14
Data Items
Records
Blocks
Files
Memory
Overview
![Page 15: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/15.jpg)
CS 245 Notes 3 15
Record - Collection of related data
items (called FIELDS)E.g.: Employee record:
name field,salary field,date-of-hire field, ...
![Page 16: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/16.jpg)
CS 245 Notes 3 16
Types of records:
• Main choices:– FIXED vs VARIABLE FORMAT– FIXED vs VARIABLE LENGTH
![Page 17: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/17.jpg)
CS 245 Notes 3 17
A SCHEMA (not record) containsfollowing information
- # fields- type of each field- order in record- meaning of each field
Fixed format
![Page 18: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/18.jpg)
CS 245 Notes 3 18
Example: fixed format and length
Employee record(1) E#, 2 byte integer(2) E.name, 10 char. Schema(3) Dept, 2 byte code
55 s m i t h 02
83 j o n e s 01
Records
![Page 19: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/19.jpg)
CS 245 Notes 3 19
• Record itself contains format“Self Describing”
Variable format
![Page 20: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/20.jpg)
CS 245 Notes 3 20
Example: variable format and length
4I52 4S DROF46
Field name codes could also be strings, i.e. TAGS
# F
ield
s
Cod
e id
enti
fyin
g
field
as
E#
Inte
ger
typ
e
Cod
e f
or
En
am
eS
trin
g t
yp
eLe
ng
th o
f st
r.
![Page 21: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/21.jpg)
CS 245 Notes 3 21
Variable format useful for:
• “sparse” records• repeating fields• evolving formats
But may waste space...
![Page 22: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/22.jpg)
CS 245 Notes 3 22
• EXAMPLE: var format record with repeating fields
Employee one or more children
3 E_name: Fred Child: SallyChild: Tom
![Page 23: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/23.jpg)
CS 245 Notes 3 23
Note: Repeating fields does not imply- variable format, nor- variable size
John Sailing Chess --
![Page 24: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/24.jpg)
CS 245 Notes 3 24
Note: Repeating fields does not imply- variable format, nor- variable size
John Sailing Chess --
• Key is to allocate maximum number ofrepeating fields (if not used null)
![Page 25: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/25.jpg)
CS 245 Notes 3 25
Many variants betweenfixed - variable format:
Example: Include record type in record
record type record lengthtells me whatto expect(i.e. points to schema)
5 27 . . . .
![Page 26: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/26.jpg)
CS 245 Notes 3 26
Record header - data at beginning
that describes recordMay contain:
- record type- record length- time stamp- other stuff ...
![Page 27: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/27.jpg)
Exercise: How to store XML data?
<table><description> people on the fourth floor <\description><people>
<person><name> Alan <\name><age> 42 <\age><email> [email protected] <\email>
<\person><person>
<name> Sally <\name><age> 30 <\age><email> [email protected] <\email>
<\person><\people><\table>
CS 245 Notes 3 27
from: Data on the Web,Abiteboul et al
![Page 28: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/28.jpg)
CS 245 Notes 3 28
Other interesting issues:
• Compression– within record - e.g. code selection– collection of records - e.g. find
common patterns
• Encryption
![Page 29: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/29.jpg)
CS 245 Notes 3 29
Encrypting Records
trustedprocessor
newrecord
rdbms
E(r)
E(r1)E(r2)E(r3)E(r4)...
![Page 30: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/30.jpg)
CS 245 Notes 3 30
Encrypting Records
trustedprocessor
searchF(r)=x
dbms??
E(r1)E(r2)E(r3)E(r4)...
![Page 31: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/31.jpg)
CS 245 Notes 3 31
Search Key in the Clear
trustedprocessor
searchk=2
dbmsQ: k=2
[1, E(b1)][2, E(b2)][3, E(b3)][4, E(b4)]
...• each record is [k,b]• store [k, E(b)]• can search for records with k=x
A: [2, E(b2)]
![Page 32: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/32.jpg)
CS 245 Notes 3 32
Encrypt Key
trustedprocessor
searchk=2
dbmsQ: k’=E(2)
[E(1), E(b1)][E(2), E(b2)][E(3), E(b3)][E(4), E(b4)]
...• each record is [k,b]• store [E(k), E(b)]• can search for records with k=E(x)
A: [E(2), E(b2)]
![Page 33: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/33.jpg)
CS 245 Notes 3 33
Issues
• Hard to do range queries• Encryption not good• Better to use encryption that does not
always generate same cyphertext
E
k
DE(k, random) k
simplification
![Page 34: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/34.jpg)
CS 245 Notes 3 34
How Do We Search Now?
trustedprocessor
searchk=2
dbmsQ: k’=E(2)
[E(1, abc), E(b1)][E(2, dhe), E(b2)][E(3, nft), E(b3)][E(2, lkz), E(b4)]
...• each record is [k,b]• store [E(k, rand), E(b)]• can search for records with k=E(x,???)?
A: [E(2,dhe), E(b2)] [E(2, lkz), E(b4)]
???
![Page 35: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/35.jpg)
CS 245 Notes 3 35
Solution?• Develop new decryption function:
D(f(k1), E(k2, rand)) is true if k1=k2
![Page 36: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/36.jpg)
CS 245 Notes 3 36
Solution?• Develop new decryption function:
D(f(k1), E(k2, rand)) is true if k1=k2
trustedprocessor
searchk=2
dbms
Q: check if D(f(2),*) true
[E(1, abc), E(b1)][E(2, dhe), E(b2)][E(3, nft), E(b3)][E(2, lkz), E(b4)]
...
A: [E(2,dhe), E(b2)] [E(2, lkz), E(b4)]
![Page 37: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/37.jpg)
CS 245 Notes 3 37
Issues?
• Cannot do non-equality predicates• Hard to build indexes
![Page 38: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/38.jpg)
What are choices/issues with data compression?
• Leaving search keys uncompressed not as bad
• Larger compression units:– better for compression efficiency– worse for decompression overhead
• Similar data compresses better – compress columns?
CS 245 Notes 3 38
![Page 39: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/39.jpg)
CS 245 Notes 3 39
Next: placing records into blocks
blocks ...
a file
![Page 40: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/40.jpg)
CS 245 Notes 3 40
Next: placing records into blocks
blocks ...
a file
assume fixedlength blocks
assume a single file (for now)
![Page 41: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/41.jpg)
CS 245 Notes 3 41
(1) separating records(2) spanned vs. unspanned(3) sequencing(4) indirection
Options for storing records in blocks:
![Page 42: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/42.jpg)
CS 245 Notes 3 42
Block
(a) no need to separate - fixed size recs.(b) special marker(c) give record lengths (or offsets)
- within each record- in block header
(1) Separating records
R2R1 R3
![Page 43: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/43.jpg)
CS 245 Notes 3 43
• Unspanned: records must be within one block
block 1 block 2
...
• Spannedblock 1 block 2
...
(2) Spanned vs. Unspanned
R1 R2
R1
R3 R4 R5
R2 R3(a)
R3(b) R6R5R4 R7
(a)
![Page 44: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/44.jpg)
CS 245 Notes 3 44
need indication need indication
of partial record of continuation
“pointer” to rest (+ from where?)
R1 R2 R3(a)
R3(b) R6R5R4 R7
(a)
With spanned records:
![Page 45: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/45.jpg)
CS 245 Notes 3 45
• Unspanned is much simpler, but may waste space…
• Spanned essential if record size > block size
Spanned vs. unspanned:
![Page 46: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/46.jpg)
CS 245 Notes 3 46
• Ordering records in file (and block) by some key value
Sequential file ( sequenced)
(3) Sequencing
![Page 47: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/47.jpg)
CS 245 Notes 3 47
Why sequencing?
Typically to make it possible to efficiently read records in order(e.g., to do a merge-join — discussed later)
![Page 48: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/48.jpg)
CS 245 Notes 3 48
Sequencing Options
(a) Next record physically contiguous
...
(b) Linked
Next (R1)R1
R1 Next (R1)
![Page 49: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/49.jpg)
CS 245 Notes 3 49
(c) Overflow area
Recordsin sequence
R1
R2
R3
R4
R5
Sequencing Options
![Page 50: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/50.jpg)
CS 245 Notes 3 50
(c) Overflow area
Recordsin sequence
R1
R2
R3
R4
R5
Sequencing Options
header
R2.1
R1.3
R4.7
![Page 51: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/51.jpg)
CS 245 Notes 3 51
• How does one refer to records?
(4) Indirection
Rx
![Page 52: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/52.jpg)
CS 245 Notes 3 52
• How does one refer to records?
(4) Indirection
Rx
Many options: Physical Indirect
![Page 53: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/53.jpg)
CS 245 Notes 3 53
Purely Physical
Device IDE.g., Record Cylinder #
Address = Track #or ID Block #
Offset in block
Block ID
![Page 54: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/54.jpg)
CS 245 Notes 3 54
Fully Indirect
E.g., Record ID is arbitrary bit string
maprec ID r address
a
Physicaladdr.Rec ID
![Page 55: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/55.jpg)
CS 245 Notes 3 55
Tradeoff
Flexibility Costto move records of indirection(for deletions, insertions)
![Page 56: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/56.jpg)
CS 245 Notes 3 56
Physical Indirect
Many optionsin between …
![Page 57: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/57.jpg)
CS 245 Notes 3 57
Example: Indirection in block
Header
A block: Free
space
R3
R4
R1 R2
![Page 58: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/58.jpg)
CS 245 Notes 3 58
Block header - data at beginning that
describes blockMay contain:
- File ID (or RELATION or DB ID) - This block ID - Record directory
- Pointer to free space- Type of block (e.g. contains recs type 4;
is overflow, …)- Pointer to other blocks “like it”- Timestamp ...
![Page 59: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/59.jpg)
CS 245 Notes 3 59
(1) separating records(2) spanned vs. unspanned(3) sequencing(4) indirection
Options for storing records in blocks:
![Page 60: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/60.jpg)
CS 245 Notes 3 60
Case Study: salesforce.com
• salesforce.com provides CRM services• salesforce customers are tenants• Tenants run apps and DBMS as service
tenant A
tenant B
tenant C
salesforce.com
dataCRM App
![Page 61: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/61.jpg)
CS 245 Notes 3 61
Options for Hosting
• Separate DBMS per tenant• One DBMS, separate tables per
tenant• One DBMS, shared tables
![Page 62: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/62.jpg)
CS 245 Notes 3 62
Tenants have similar data
customer A B C D E F a1 b1 c1 d1 e1 - a2 b2 c2 - e2 f2
customer A B C D G a3 b3 c2 - - a1 b1 c1 - g1 a4 - - d1
tenant 1:
tenant 2:
![Page 63: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/63.jpg)
CS 245 Notes 3 63
salesforce.com solutioncustomer tenant A B C 1 a1 b1 c1 1 a2 b2 c2 2 a3 b3 c2 2 a1 b1 c1
cust-other tenant A f1 v1 f2 v2 ... 1 a1 D d1 E e1 1 a2 E e2 F f2 2 a1 G g1 3 a4 D d1
fixed schema forall tenants
var schema forall tenants
![Page 64: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/64.jpg)
CS 245 Notes 3 64
(1) Insertion/Deletion(2) Buffer Management(3) Comparison of Schemes
Other Topics
![Page 65: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/65.jpg)
CS 245 Notes 3 65
Block
Deletion
Rx
![Page 66: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/66.jpg)
CS 245 Notes 3 66
Options:
(a) Immediately reclaim space(b) Mark deleted
![Page 67: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/67.jpg)
CS 245 Notes 3 67
Options:
(a) Immediately reclaim space(b) Mark deleted
– May need chain of deleted records
(for re-use)– Need a way to mark:
• special characters• delete field• in map
![Page 68: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/68.jpg)
CS 245 Notes 3 68
As usual, many tradeoffs...
• How expensive is to move valid record to free space for immediate reclaim?
• How much space is wasted?– e.g., deleted records, delete fields,
free space chains,...
![Page 69: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/69.jpg)
CS 245 Notes 3 69
Dangling pointers
Concern with deletions
R1 ?
![Page 70: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/70.jpg)
CS 245 Notes 3 70
Solution #1: Do not worry
![Page 71: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/71.jpg)
CS 245 Notes 3 71
E.g., Leave “MARK” in map or old location
Solution #2: Tombstones
![Page 72: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/72.jpg)
CS 245 Notes 3 72
E.g., Leave “MARK” in map or old location
Solution #2: Tombstones
• Physical IDs
A block
This space This space cannever re-used be re-used
![Page 73: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/73.jpg)
CS 245 Notes 3 73
• Logical IDs
ID LOC
7788
map
Never reuseID 7788 nor
space in map...
E.g., Leave “MARK” in map or old location
Solution #2: Tombstones
![Page 74: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/74.jpg)
CS 245 Notes 3 74
Easy case: records not in sequence Insert new record at end of
file or in deleted slot If records are variable size,
not as easy...
Insert
![Page 75: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/75.jpg)
CS 245 Notes 3 75
Hard case: records in sequence If free space “close by”, not too bad... Or use overflow idea...
Insert
![Page 76: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/76.jpg)
CS 245 Notes 3 76
Interesting problems:
• How much free space to leave in each block, track, cylinder?
• How often do I reorganize file + overflow?
![Page 77: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/77.jpg)
CS 245 Notes 3 77
Freespace
![Page 78: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/78.jpg)
CS 245 Notes 3 78
• DB features needed• Why LRU may be bad Read• Pinned blocks Textbook!• Forced output• Double buffering• Swizzling
Buffer Management
in Notes02
![Page 79: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/79.jpg)
CS 245 Notes 3 79
Swizzling
Memory Disk
Rec A
block 1
block 2
block 1
![Page 80: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/80.jpg)
CS 245 Notes 3 80
Swizzling
Memory Disk
Rec A
block 1
Rec Ablock 2 block 2
block 1
![Page 81: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/81.jpg)
CS 245 Notes 3 81
Row vs Column Store• So far we assumed that fields of a
record are stored contiguously (row store)...
• Another option is to store like fields together (column store)
![Page 82: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/82.jpg)
CS 245 Notes 3 82
• Example: Order consists of– id, cust, prod, store, price, date, qty
Row Store
id1 cust1 prod1 store1 price1 date1 qty1
id2 cust2 prod2 store2 price2 date2 qty2
id3 cust3 prod3 store3 price3 date3 qty3
![Page 83: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/83.jpg)
CS 245 Notes 3 83
• Example: Order consists of– id, cust, prod, store, price, date, qty
Column Store
id1 cust1id2 cust2id3 cust3id4 cust4... ...
id1 prod1id2 prod2id3 prod3id4 prod4... ...
id1 price1 qty1id2 price2 qty2id3 price3 qty3id4 price4 qty4... ... ...
ids may or may not be stored explicitly
![Page 84: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/84.jpg)
CS 245 Notes 3 84
Row vs Column Store• Advantages of Column Store
– more compact storage (fields need not start at byte boundaries)
– efficient reads on data mining operations
• Advantages of Row Store– writes (multiple fields of one
record)more efficient– efficient reads for record access (OLTP)
![Page 85: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/85.jpg)
CS 245 Notes 3 85
Interesting paper to read:
• Mike Stonebreaker, Elizabeth (Betty) O'Neil, Pat O’Neil, Xuedong Chen, et al. " C-Store: A Column-oriented DBMS," Presented at the 31st VLDB Conference, September 2005.
• http://www.cs.umb.edu/%7Eponeil/vldb05_cstore.pdf
![Page 86: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/86.jpg)
CS 245 Notes 3 86
• There are 10,000,000 ways to organize my data on disk…
Which is right for me?
Comparison
![Page 87: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/87.jpg)
CS 245 Notes 3 87
Issues:
Flexibility Space Utilization
Complexity Performance
![Page 88: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/88.jpg)
CS 245 Notes 3 88
To evaluate a given strategy, compute following parameters:-> space used for expected data-> expected time to
- fetch record given key- fetch record with next key- insert record- append record- delete record- update record- read all file- reorganize file
![Page 89: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/89.jpg)
CS 245 Notes 3 89
Example
How would you design Megatron 3000 storage system? (for a relational DB, low end)– Variable length records?– Spanned?– What data types?– Fixed format?– Record IDs ?– Sequencing?– How to handle deletions?
![Page 90: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/90.jpg)
CS 245 Notes 3 90
• How to lay out data on disk
Data Items
Records
Blocks
Files
Memory
DBMS
Summary
![Page 91: CS 245: Database System Principles Notes 03: Disk Organization](https://reader035.vdocuments.us/reader035/viewer/2022062408/56813add550346895da3262e/html5/thumbnails/91.jpg)
CS 245 Notes 3 91
How to find a record quickly,given a key
Next