Distributed Database Management Systems: Lecture 31

Distributed Database Management Systems: Lecture 31. The main topics covered in this chapter include: query decomposition; steps in query processing; SQL query on distributed relations; optimized fragment query with communication operations;...

Distributed Database Management Systems

Lecture 31

In the previous lecture

• Basic Concepts of Query

Optimization

• QP in centralized and

Distributed DBs

In this Lecture

• Query Decomposition

• Its Phases

SQL Query on Distributed Relations

QUERY DECOMPOSITION GLOBALSCHEMAAlgebraic Query on Distributed

Relations

DATA LOCALIZATION FRAGMENTSCHEMAFragment Query

GLOBAL OPTIMIZATION STAT OFFRAGMENTSOptimized Fragment Query with

Communication Operations LOCAL

OPTIMIZATION SCHEMALOCAL

Optimized Local Query

Steps in Query Processing

1- Normalization

complex

Analysis (like compilers)

Clause

V into U and ^ into

V into U and ^ into or ⋈

.

• SELECT eName

FROM EMP, ASG

WHERE EMP.eNo = ASG.eNo AND ASG.pNo = ‘P1’

AND dur = 12 OR dur = 24

^ASG.pNo = ‘P1’ ^ dur = 24)

2- Analysis

• Reject incorrect ones

• Type incorrect

–Relations/attributes not exist

Query Graph

or operand relations

• Links represent joins or

projection (result node)

• Self join/select on

operand nodes

• Select eName, resp

FROM EMP, ASG, PROJ

WHERE EMP.eNo = ASG.eNo AND ASG.pNo = PROJ.pNo AND pName = ‘CAD/CAM’ AND dur ≥ 36

AND title = ‘Programmer’

Query Graph

RESULT EMP

Dur ≥ 36

ASG.pNo = PROJ.pNo

pName =

‘CAD/CAM’ resp

Semantically Incorrect

• Select eName, resp

FROM EMP, ASG, PROJ

WHERE EMP.eNo =ASG.eNo AND pName = ‘CAD/CAM’ AND dur ≥ 36

AND title = ‘Programmer’

Query Graph

RESULT EMP

Dur ≥ 36

pName =

‘CAD/CAM’ resp

3- Elimination of Redundancy

already used in views

• User mistake or this

replacement may contain

redundant predicates

rules

• Select title

FROM EMP

WHERE (title = ‘Prog’

AND ((not (title = ‘Prog’))

OR title = ‘Elect Engr’)

AND (not (title = ‘Elect Engr’)))

OR eName = ‘Saleem’

(false) v (false) v p3 = p3

steps-Query Tree

• Select eName

FROM EMP, ASG, PROJ

WHERE EMP.eNo = ASG.eNo AND ASG.pNo = PROJ.pNo

AND pName = ‘CAD/CAM’ AND (dur = 36 or dur = 24)

PROJ ASG EMP

(R x S) x T S x (R x T)

(R (R S) T ⋈ ⋈ S (R ⋈ ⋈ T)

3- Idempotency of unary Ops

i) A’( A”(R)) A’(R)

ii) σp1(A1)(σp2(A2) (R))

σp1(A1) ∧ p2(A2)

(R)-4- Commuting selection with projection

A1, ….,An ( p(Ap)(R)) A1,

….,An (( p(Ap) A1, ….,An, Ap

• Many eq query trees

can be generated

• Comparing all such

trees to select best is not feasible

• Heuristic is applied

1 Separation of Unary Ops

relation grouped together

binary ops

ASN PROJ EMP

x

⋈ pNo^eNo

(pName = ‘CAD/CAM’)^ ( dur = 12 v dur = 24)^ eName ’Saleem’

eName

PROJ ASG EMP

pNo, eName eName

• This concludes QP that

comprises 4 steps

• Now we move to the

second phase of Query Optimization; Data

Localization of DD

• QD at global level, this

phase transform into

local ones

• A Nạve rule…

• However, it won’t be

an efficient one

Reduction During Data

Localization

Example Schema

• Reduction with Selection

• Rule 1:

pi (Rj) = Ø if x in R ∀ j: (pi(x) ^ pj(x))

• That is, there exist

conflicting predicates

• Select * from EMP where

Reduction on Join

• Distributing joins over

unions and avoiding

unnecessary joins

• (R1UR2) ⋈ R3= (R1 ⋈

R3) U (R2UR3)

• Rule2:

• Ri⋈Rj = Ø if x in R∀ iand y in R∀ j: (pi(x) ^

pj(x))

• Useless joins can be

determined viewing the join predicates

• Remember! Reduced query

is not always better We

have to be watchful

• ASG1 = eNo ≤ ‘E3’ (ASG)ASG1 =

• ASG2 = ’eNo > ‘E3’ (ASG).ASG2 =

–θ {=, <, ≠ ≤, >, ≥} ∈ ∈ ∀– ←Ø ⇒ ⋉ ⋈ ∪ ^

–