PostGreSQL Performance Tuning

POSTGRESQL
PERFORMANCE TUNING
Presented by: Majid Tahir
Maven Logix Pvt Ltd

AGENDA
 Introduction
 What is Database Performance Tuning
 Factors affecting database performance
 Tuning PGSQL Database Parameters
 Performance Tips
 PGSQL Tuning Tools

DATABASE PERFORMANCE TUNING
 Database performance tuning encompasses the
steps you can take to optimize performance with
the goal of maximizing the use of system resources
for greater efficiency.
 At a high level, database performance can be
defined as the rate at which a
database management system (DBMS) supplies
information to users.

FACTORS AFFECTING DATABASE
PERFORMANCE
 There are five factors that influence database
performance:
1. Workload
2. Throughput
3. Resources
4. Optimization
5. Contention.

WORKLOAD
 The workload that is requested of the DBMS defines
the demand.
 It is a combination of
 Online transactions
 Batch jobs
 Ad hoc Queries
 Workload can be predicted (such as heavy month-end
processing of payroll, or very light access after 7 pm,
but at other times it is unpredictable.
 The overall workload can have a major impact on
database performance.

THROUGHPUT
 Throughput defines the overall capability of the
computer to process data.
 It is a composite of
 I/O speed
 CPU speed
 Parallel capabilities of the machine
 The efficiency of the operating system and system
software.

RESOURCES
 The hardware and software tools at the disposal of
the system are known as the resources of the
system.
 Examples include
 Memory (such as that allocated to buffer pools)
 Disk

OPTIMIZATION
 Database optimization involves maximizing the
speed and efficiency with which data is retrieved.
 Relational database management systems
(RDBMS) rely on an optimizer (or relational
optimizer) that transforms SQL statements into
executable code.
 Other factors that need to be optimized
 Database Parameters
 System parameters, etc.

CONTENTION
 When the demand (workload) for a particular
resource is high, Contention can result.
 Contention is the condition in which two or more
components of the workload are attempting to use
a single resource in a conflicting way
 (for example, dual updates to the same piece of
data).
 The DBMS uses a locking mechanism
 The DBMS locking strategies permit multiple users
from multiple environments to access and modify
data in the database at the same time.

TUNING IN PGSQL
 First step is to understand lifecycle of a query
 Following are the steps:
1. Transmission of query string to database backend
2. Parsing of query string
3. Planning of query to optimize retrieval of data
4. Retrieval of data from hardware
5. Transmission of results to client

TRANSMISSION OF QUERY STRING TO DATABASE
BACKEND
 The first step is the sending of the query string ( the
actual SQL command you type in or your
application uses ) to the database backend.
 Select * from emp

PARSING OF QUERY STRING
 Once the SQL query is inside the database server it
is parsed into tokens.
 During the parse call, the database performs the
following checks:
 Syntax Check
e.g. SELECT * FORM employees;
SELECT * FORM employees
* ERROR at line 1: FROM keyword not found where
expected
 Semantic Check whether a statement is meaningful)

PLANNING OF QUERY TO OPTIMIZE RETRIEVAL OF
DATA
 The planning of the query is where PostgreSQL
really starts to do some work.
 It also analyzes your SQL to determine what the
most efficient way of retrieving your data is.
 Should we use an index?
 Maybe a hash join on those two tables is appropriate?

RETRIEVAL OF DATA FROM HARDWARE
 Now that PostgreSQL has a plan of what it believes
to be the best way to retrieve the data, it is time to
actually get it.
 This step is mostly effected by your hardware
configuration.

TRANSMISSION OF RESULTS TO CLIENT
 And finally the last step is to transmit the results to
the client.
 All of the data that you are returning is pulled from
the disk and sent over the wire to your client.
 Minimizing the number of rows and columns to only
those that are necessary can often increase your
performance.

DATABASE PARAMETERS
 The default PostgreSQL configuration is not tuned
for any particular workload.
 Default values are set to ensure that PostgreSQL
runs everywhere, with the least resources it can
consume and so that it doesn’t cause any
vulnerabilities.
 It has default settings for all of the database
parameters.
 The parameters can be changed in the
postgresql.conf file located in the installation folder

MAX_CONNECTIONS
 Determines the maximum number of concurrent
connections to the database server. The default is
typically 100 connections
 The change in max_connections require restart

SHARED_BUFFERS
 This parameter sets how much dedicated memory
will be used by PostgreSQL for cache.
 This should be set to roughly 25% of available RAM
on the system.
Editing this option is the simplest way to improve the
performance of your database server.

WAL_BUFFERS
 PostgreSQL writes its WAL (write ahead log) record
into the buffers and then these buffers are flushed
to disk.
 The default size of the buffer, defined
by wal_buffers, is 16MB, but if you have a lot of
concurrent connections then a higher value can
give better performance.
 Change in this parameter requires restart

EFFECTIVE_CACHE_SIZE
 This value tells PostgreSQL's optimizer how much
memory PostgreSQL has available for caching data
and helps in determing whether or not it use an
index or not.
 The larger the value increases the likely hood of
using an index.
 This should be set to the amount of memory
allocated to shared_buffers plus the amount of OS
cache available.
 Often this is more than 50% of the total system
memory.
 The larger value is recommended by experts

WORK_MEM
 This configuration is used for complex sorting.
 If you have to do complex sorting then increase the
value of work_mem for good results.
 In-memory sorts are much faster than sorts spilling
to disk.

MAX_FSM_PAGES
 This option helps to control the free space map.
 When something is deleted from a table it isn't
removed from the disk immediately, it is simply
marked as "free" in the free space map.
 The space can then be reused for any new
INSERTs that you do on the table.
 If your setup has a high rate of DELETEs and
INSERTs it may be necessary increase this value to
enhance performance.

MAINTENANCE_WORK_MEM
 maintenance_work_mem is a memory setting used
for maintenance tasks.
 The default value is 64MB.
 Setting a large value helps in tasks like RESTORE,
CREATE INDEX, ADD FOREIGN KEY and ALTER
TABLE.

USEFUL TUNING TOOLS
 EXPLAIN ANALYZE
 POSTGRESQLTUNER
 PGBADGER
 Analyze logs to generate performance reports
 PGTUNE

DB
Performance
Tuning
Approaches
Monitoring
Tools
Optimization
Partition Table
Index
Managment
Table
Reorganization
Parameter
Configuration

PERFORMANCE TIPS
 Use tool that best fit for your environment
 One index per query
 Use Multiple columns in index
 There are more parameters that can be tuned to
gain better performance.
 In the end, we must always keep in mind that not all
parameters are relevant for all applications types.
 Some applications perform better by tuning a
parameter and some don’t.
 Database parameters must be tuned for the specific
needs of an application and the OS it runs on.

PostGreSQL Performance Tuning

More Related Content

What's hot (20)

Similar to PostGreSQL Performance Tuning (20)

More from Maven Logix (6)

Recently uploaded (20)

PostGreSQL Performance Tuning