Dividing a cluster

p5

Indexes

Access

Predicate

vector

0

0

0

Bit Vector

p1

p2 p3 p4 p5

p1

list of clusters

A

B

C

p1

p2

p3

p4

p5

S1: p1,p2,p4

S2: p1,p3,p5

S3: p1,p4

s1

p4

detail of cluster 1.2

s3

0

0

0

p1p3

p1p2

S2

S1

1.1

1.2

p1p2

AB

p1p3

hash(pA,pB)

Average Cost of cluster 1.2:

Additional Index_cost:

check(AB,p1) + ?(AB,p1) Hash(p1,pB)

Gain in checking cost for S1

?(p1)checkpred(3) - ?(p1,p5) checkpred(2)

We sum the checking cost gain for all subscriptions that are moved

Diapositive précédente

Diapositive suivante

Revenir à la première diapositive

Afficher la version graphique