CS 4432lecture #61 CS4432: Database Systems II Lecture #6 Professor Elke A. Rundensteiner.

CS 4432 lecture #6 1

CS4432: Database Systems IILecture #6

Professor Elke A. Rundensteiner


PROJECT 1

• Project-1 teams assigned (my.wpi)

• If any issues, let us know sooner rather than later …

• Must state each member’s contributions …


Storage Layout :How to lay out data on disk. ( chapter 3)


Placing records into blocks

blocks ...

a file

assume fixedlength blocks

assume a single file (for now)


(1) separating records(2) spanned vs. unspanned(3) mixed record types – clustering(4) split records(5) sequencing(6) indirection

Options for storing records in blocks:


Fixed part in one block

Typically forhybrid format

Variable part in another block

(4) Split records


Block with fixed recs.

R1 (a)R1 (b)

Block with variable recs.

R2 (a)

R2 (b)

R2 (c)


• Ordering records in file (and block) by some key value– Sequential file ( - sequenced file)

• Why sequencing ?– Typically to make it possible to efficiently read

records in order

(5) Sequencing


Sequencing Options

(a) Next record physically contiguous

...

(b) Linked

What about INSERT/ DELETE ?

Next (R1)R1

R1 Next (R1)


(c) Overflow area

Recordsin sequence

R1

R2

R3

R4

R5

Sequencing Options

header

R2.1

R1.3

R4.7


• How does one refer to records?

• Problem: Records can be on disk or in (virtual) memory. Need common address, but have different physical locations.

(6) Indirection Addressing

Rx

Many options: Physical Indirect


Purely Physical Addressing

Device ID E.g., Record Cylinder #

Address = Track #( ID ) Block #

Offset in block

Block ID


Fully Indirect Addressing

Solution: Record ID (Oracle: ROWID) as global address, maintain a map table.

Map Tablerec ID r address

a

Physicaladdr.Rec ID


Tradeoff

Flexibility Costto move records of indirection(for deletions, insertions) (lookup)

What to do : Options inbetween ?

Physical Indirect


Ex #1 : Indirection in block

Block Header

A block: Free

space

R3

R4

R1 R2


Ex. #2 Use logical block #’s understood by file system

instead of direct disk accessREC ID File ID

Block # Record # or Offset

File ID, PhysicalBlock # Block ID

File System Map


(1) Separating records(2) Spanned vs. Unspanned(3) Mixed record types - Clustering(4) Split records(5) Sequencing(6) Indirection

Recap: Storing records in blocks


(1) Insertion/Deletion(2) Buffer Management(3) Comparison of Schemes

Other Topics in Chapter 3


Block

Deletion

Rx


Options:

(a) Deleted and 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


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,...


Dangling pointers

Note: If pointers point to physical locations (rather than ROWIDs), storing new data in deleted block corrupts data.

Concern with deletions

R1 ?


Solution #1: Do not worry


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


• 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


• Place record ID within every record• When you follow a pointer, check if

it leads to correct record

Solution #3 (?):

Does this work??? If space reused, won’t new record have same ID?

to3-77

rec-id:3-77


Easy case: Records fixed length/not in sequence Insert new record at end of file

or, in deleted slotA little harder:

If records are variable size, not as easy may not be able to reuse space – fragmentationHard case: records in sequence

If free space “close by”, not too bad... Or, use overflow idea...

Or worst case, reorganize file ...

Insert


Interesting problems:

• How much free space to leave in each block, track, cylinder?

• How often do I reorganize file + overflow?

Freespace


• DB features needed• Why LRU may be bad Read• Pinned blocks Textbook!• Forced output• Double buffering• Swizzling

Buffer Management

in Notes03


Pointer Swizzling

Memory Disk

Rec A

block 1

Rec Ablock 2 block 2

block 1

Issue : If records (objects) contain pointers to otherobjects, translate locations when load objects into memory.


Translation DB Addr Mem Addr Table Rec-A Rec-A-inMem

One Option:

Solution: Insert fields that represent pointers into map table. Translate pointers as needed.


In memory pointers - need “type” bit

to disk

to memoryM

Another Option:


Swizzling Issues

• Automatic• On-demand• No swizzling / program control

Swizzling Options

• Must ‘unswizzle’

• Updating/writing of records


• There are 1,000,001 ways to organize my data on disk…

Which is right for me?

Comparison


Issues:

Flexibility Space Utilization

Complexity Performance


To evaluate a given strategy, compute following parameters:

-> space used for expected data - on average

-> expected time to :- fetch record given key- fetch record with next key- insert/delete/update record- read complete file- reorganize file (maybe sort)

-> usage patterns / workload: - how many/which user queries/updates


Example : What about Project 1?

Design decisions for CAPE queue storage system? – What data types?– Variable/Fixed length records?– Fixed/Variable format?– Record IDs ?– Sequencing? What’s special here ?– Deletions?– Insertions? – Headers/Meta-data for Queues?– Buffering/Disk Pushing?– Reorganize files and Overflow? – Defragmentation?– …


Chapter 4 (Chapter 13 in ‘compete book’ )

How to find a record quickly, given a key

TOMORROW

Date post:	20-Dec-2015
Category:	Documents
View:	214 times
Download:	1 times

CS 4432lecture #61 CS4432: Database Systems II Lecture #6 Professor Elke A. Rundensteiner.

Documents