Scribd
Upload
298 views

Database Performance and Tuning

This document discusses database performance and tuning. It provides an overview of the database market, with Oracle leading the market share. It also shows that Unix/Linux platforms lead as the hosting operating system. The document then discusses Oracle architecture and the components of the system global area (SGA). It provides a roadmap for tuning methodology, covering business rules, database design, operations, I/O, memory, contention, and the underlying platform. Additional sections cover database indexing techniques and best practices, prepared statements, and example applications. A case study discusses issues faced by a database containing 27 lakh records and the actions taken to address performance problems.

Uploaded by

subhashsirsi
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
298 views

Database Performance and Tuning

This document discusses database performance and tuning. It provides an overview of the database market, with Oracle leading the market share. It also shows that Unix/Linux platforms lead as the hosting operating system. The document then discusses Oracle architecture and the components of the system global area (SGA). It provides a roadmap for tuning methodology, covering business rules, database design, operations, I/O, memory, contention, and the underlying platform. Additional sections cover database indexing techniques and best practices, prepared statements, and example applications. A case study discusses issues faced by a database containing 27 lakh records and the actions taken to address performance problems.

Uploaded by

subhashsirsi
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 13

Database performance and

tuning

By:Subhash Hegde

June 27th 2009 1


Market proposition
 $25 Billion/120,000 Crore industry
 Oracle leads the market share
 Unix/Linux lead the market as the
hosting operating system

June 27th 2009 Karthavya technologies 2


Market share

15%
Oracle
1%
IBM-DB2
44%
20% Microsoft-Sql
server
Mysql

Others
20%

June 27th 2009 3


Hosting operating system

7%
8%
Unix/Linux
Windows
45%
Mainframe
Others
40%

June 27th 2009 4


Oracle architecture

Control File
Redo Log Buffer
Datafile
Bufer Cache

Shared Pool
Redo Log

init.ora

Archived log file

Password File

DBWR LGWR LCKO Alert log file

Trace files
SGA

Instance Database
June 27th 2009 5
Tuning methodology- Road map

 Tuning business rules


 Tuning Database design
 Tuning database operations
 Tuning I/O and memory allocation
 Tuning resource contention
 Tuning the underlying platform

June 27th 2009 6


Database indexing- How it works
 B tree structure
 Oracle decides to use index or table scan depending
on its calculations.

June 27th 2009 7


Example: Performance boost
achieved by indexing
 Search without index
 Search with index
 Performance comparison against a
index on a column with less
cardinality
 Tools used to verify query
performance – sqlplus autotrace
utility, sql developer etc.

June 27th 2009 8


Database indexing (cntd..) Best
practices
 Index columns having maximum cardinality
 Limit the number of indexes on each table
 Create an index if you frequently want to retrieve
less than 15% of the rows
 To improve performance on joins of multiple tables,
index columns used for joins.
 The order of columns in the CREATE INDEX
statement can affect query performance. In
general, specify the most frequently used columns
first.
 Drop indexes on a table before migration

June 27th 2009 9


Prepared statements
 Can help increase security by
separating SQL logic from the data
being supplied
 Pre-prepares execution plan
 Not recommended for rarely used
inserts.
 Hibernate uses prepared statements

June 27th 2009 10


Buzzwords / applications
 RBO (Rule based optimizer)
 CBO (Cost based optimizer)
 Enterprise manager
 RMAN (Recovery manager)
 Schema
 Tablespace

June 27th 2009 11


Case study- BLIS
Issues faced and action taken
 BLIS database contains approximately
27 Lakh records
 Initial approach to attain search speed
failed
 Database went down because of issues
like row lock and index failures
 Data backup method

June 27th 2009 12


Questions?

June 27th 2009 13

You might also like