Self-Tuning Memory Management of A Database System

IBM T.J. Watson Research Center

Self-Tuning Memory Management of A Database System

Yixin Diao

diao@us.

IM 2009 Tutorial: Recent Advances in the Application of Control Theory to Network and Service Management

IBM T.J. Watson Research Center

DB2 Self-Tuning Memory Management

DB2 UDB Server

Memory pools

Disks

Agents

Memory pools

DB2 Clients

Technical problems

? Large systems with varying workloads and many configuration parameters

? Autonomic computing: systems self-management

Challenges from systems aspects ? Heterogeneous memory pools ? Dissimilar usage characteristics

Challenges from control aspects ? Adaptation and self-design ? Reliability and robustness

2

IM2009:RecentAdvancesintheApplica6onofControlTheorytoNetworkandServiceManagement ? 2009 IBM Corporation

IBM T.J. Watson Research Center

Load Balancing for Database Memory

Measured Output 1

Resource

Allocation 1 Resource Consumer 1

Load Balancer

Resource

Resource Allocation N

Consumer N

Measured Output N

Load Balancing ? Fairness optimal ? ? Common measured output ?

Saved System Time (xi )

Resource

BenefitPerPage (yi )

Benefit (sec/page)

0.16

0.14

OLTP

0.12

0.1

0.08

0.06

0.04

0.02

0 0

1000 2000 3000 4000 5000

Entry size (Page)

savedTime simPages

Memory Pool Size (ui )

3

IM2009:RecentAdvancesintheApplica6onofControlTheorytoNetworkandServiceManagement ? 2009 IBM Corporation

IBM T.J. Watson Research Center

Constrained Optimization and Regulatory Control

Saved DSyisskteTmimTeim( exi()xi ) Overall

Regulatory Control

Mem pool 1 (x1) BenefitPerPage (y1)

Mem pool 2 (x2)

MMeemmsoizreyP1o(oul11) Mem size 2 (u2)

Optimal memory allocation

Constrained Optimization

Karush-Kuhn-Tucker conditions

4

IM2009:RecentAdvancesintheApplica6onofControlTheorytoNetworkandServiceManagement ? 2009 IBM Corporation

IBM T.J. Watson Research Center

Dynamic State Feedback Controller

State space model Control error Integral control error Feedback control law

5

IM2009:RecentAdvancesintheApplica6onofControlTheorytoNetworkandServiceManagement ? 2009 IBM Corporation

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download