Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Tree: 2be6f02d06
Fetching contributors…

Cannot retrieve contributors at this time

698 lines (558 sloc) 24.513 kB
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "">
<title>pgpool-II Tutorial</title>
<meta http-equiv="Content-Type" content="text/html" />
<link href="pgpool.css" rel="stylesheet" type="text/css" />
<h1>pgpool-II Tutorial</h1>
<p>Welcome to the Tutorial for pgpool-II. From here, you can learn how
to install, setup, and run parallel queries or do replication using
pgpool-II. We assume that you already know basic PostreSQL operations,
so please refer to the PostgreSQL document if needed.</p>
<dt><em>Table of Contents</em></dt>
<dt>1. <a href="#start">Let's Begin!</a></dt>
<dd>1.1. <a href="#install">Installing pgpool-II</a></dd>
<dd>1.2. <a href="#config">Configuration Files</a></dd>
<dd>1.3. <a href="#pcp-config">Configuring PCP commands</a></dd>
<dd>1.4. <a href="#db-node">Preparing Database Nodes</a></dd>
<dd>1.5. <a href="#start-shutdown">Starting/Stopping pgpool-II</a></dd>
<dt>2. <a href="#replication">Your First Replication</a></dt>
<dd>2.1. <a href="#replication-config">Configuring Replication</a></dd>
<dd>2.2. <a href="#replication-check">Checking Replication</a></dd>
<dt>3. <a href="#parallel">Your First Parallel Query</a></dt>
<dd>3.1. <a href="#parallel-config">Configuring Parallel Query</a></dd>
<dd>3.2. <a href="#system-db">Configuring the System Database</a></dd>
<dd>3.3. <a href="#dist-def">Partitioning Rule Definition</a></dd>
<dd>3.4. <a href="#replicate-def">Replication Rule Definition</a></dd>
<dd>3.5. <a href="#parallel-check">Checking Parallel Query</a></dd>
<h2>1. <a name="start">Let's Begin!</a></h2>
<p>First, we must learn how to install, configure pgpool-II and
database nodes before using replication or parallel query.</p>
<h3>1.1. <a name="install">Installing pgpool-II</a></h3>
<p>Installing pgpool-II is very easy. In the directory which you have
extracted the source tar ball, execute the following commands.</p>
$ ./configure
$ make
$ make install
<p><code>configure</code> script collects your system information
and use it for the compilation procedure. You can pass command line
arguments to <code>configure</code> script to change the default
behavior, such as the installation directory. pgpool-II will be
installed to <code>/usr/local</code> directory by default.</p>
<p><code>make</code> command compiles the source code, and <code>make
install</code> will install the executables. You must have write
permission on the installation directory.</p>
<p>In this tutorial, we will install pgpool-II in the default
<code>/usr/local</code> directory.</p>
<p><em>Note</em>: pgpool-II requires libpq library in PostgreSQL 7.4
or later (version 3 protocol). If <code>configure</code> script
displays the following error message, libpq library may not be
installed, or it is not of version 3.</p>
<pre>configure: error: libpq is not installed or libpq is old</pre>
<p>If the library is version 3, but above message is still displayed,
your libpq library is probably not recognized by
<code>configure</code> script.</p>
<p><code>configure</code> script searches for libpq library under
<code>/usr/local/pgsql</code> libaray. If you have installed
PostgreSQL to a directory other than <code>/usr/local/pgsql</code>,
use <code>--with-pgsql</code>, or <code>--with-pgsql-includedir</code>
and <code>--with-pgsql-libdir</code> command line options when you
execute <code>configure</code>.</p>
<h3>1.2. <a name="config">Configuration Files</a></h3>
<p>pgpool-II configuration parameters are saved in
<code>pgpool.conf</code> file. The file is in "<code>parameter =
value</code>" per line format. When you install pgpool-II,
<code>pgpool.conf.sample</code> is automatically created. We recommend
copying and renaming it to <code>pgpool.conf</code>, and edit it as
you like.</p>
<pre>$ cp /usr/local/etc/pgpool.conf.sample /usr/local/etc/pgpool.conf</pre>
<p>pgpool-II only accepts connections from the local host using port
9999. If you wish to receive conenctions from other hosts, set
<code>listen_addresses</code> to '*'.</code></p>
listen_addresses = 'localhost'
port = 9999
<p>We will use the default parameters in thie tutorial.</p>
<h3>1.3. <a name="pcp-config">Configuring PCP Commands</a></h3>
<p>pgpool-II has an interface for administration purpose to retrieve
information on database nodes, shutdown pgpool-II, etc. via
network. To use PCP commands, user authentication is required. This
authentication is different from PostgreSQL's user authentication. A
username and password need to be defined in <code>pcp.conf</code>
file. In the file, a username and password are listed as a pair on
each line, and they are separated by a colon (:). Passwords are
encrypted in md5 hash format.</p>
When you install pgpool-II, <code>pcp.conf.sample</code> is
automatically created. We recommend copying and renaming it to
<code>pcp.conf</code>, and edit it.
<pre>$ cp /usr/local/etc/pcp.conf.sample /usr/local/etc/pcp.conf</pre>
<p>To encrypt your password into md5 hash format, use pg_md5 command,
which is installed as a part of pgpool-II
executables. <code>pg_md5</code> takes text as an command line
argument, and displays its md5-hashed text.</p>
<p>For example, give "postgres" as the command line argument, at
<code>pg_md5</code> displays md5-hashed text to the standard
$ /usr/bin/pg_md5 postgres
<p>PCP commands are executed via network, so the port number must be
configured with <code>pcp_port</code> parameter in
<code>pgpool.conf</code> file.</p>
<p>We will use the default 9898 for <code>pcp_port</code> in this tutorial.</p>
<pre>pcp_port = 9898</pre>
<h3>1.4. <a name="db-node">Preparing Database Nodes</a></h3>
<p>Now, we need to set up backend PostgreSQL servers for
pgpool-II. These servers can be placed within the same host as
pgpool-II, or on separate machines. If you decide to place the servers
on the same host, different port numbers must be assigned for each
server. If the servers are placed on separate machines, they must be
configured properly so that they can accept network connections from
<p>In this tutorial, we will place three servers within the same host
as pgpool-II, and assign 5432, 5433, 5434 port numbers
respectively. To configure pgpool-II, edit <code>pgpool.conf</code> as
backend_hostname0 = 'localhost'
backend_port0 = 5432
backend_weight0 = 1
backend_hostname1 = 'localhost'
backend_port1 = 5433
backend_weight1 = 1
backend_hostname2 = 'localhost'
backend_port2 = 5434
backend_weight2 = 1
<p>For <code>backend_hostname</code>, <code>backend_port</code>,
<code>backend_weight</code>, set the node's hostname, port number, and
ratio for load balancing. At the end of each parameter string, node ID
must be specified by adding positive integers starting with 0 (i.e. 0,
1, 2, &hellip;).</p>
<p><code>backend_weight</code> parameters are all 1, meaning that
SELECT queries are equally distributed among three servers.</p>
<h3>1.5. <a name="start-shutdown">Starting/Stopping pgpool-II</a></h3>
<p>To fire up pgpool-II, execute the following command on a terminal.</p>
<pre>$ pgpool</pre>
<p>The above command, however, prints no log messages because pgpool
detaches the terminal. If you want to show pgpool log messages, you
pass <code>-n</code> option to pgpool command. pgpool-II is executed
as non-daemon process, and the terminal will not be detached.
$ pgpool -n &
The log messages are printed on the terminal, so the recommended
options to use are like the following.
<pre>$ pgpool -n -d &gt; /tmp/pgpool.log 2&gt;&1 &</pre>
<p><code>-d</code> option enables debug messages to be generated.</p>
The above command keeps appending log messages to /tmp/pgpool.log. If
you need to rotate log files, pass the logs to a external command
which have log rotation function.
For example, you can use rotatelogs coming with Apache2:
$ pgpool -n 2>&1 | /usr/local/apache2/bin/rotatelogs \
-l -f /var/log/pgpool/pgpool.log.%A 86400 &
This will generate a log file named "pgpool.log.Thursday" everyday then
rotate it 00:00 at midnight. Rotatelogs adds log to a file if it already
exists. To delete old log files before rotation, you could use cron:
55 23 * * * /usr/bin/find /var/log/pgpool -type f -mtime +5 -exec /bin/rm -f '{}' \;
Please note that rotaelogs may exist as /usr/sbin/rotatelogs2 in some
-f option generates a log file as soon as rotatelogs starts and is
available apache2 2.2.9 or greater.
Also <a
href=""><code>cronolog</code></a> helps you.
$ pgpool -n 2>&1 | /usr/sbin/cronolog \
--hardlink=/var/log/pgsql/pgpool.log \
'/var/log/pgsql/%Y-%m-%d-pgpool.log' &
<p>To stop pgpool-II process, execute the following command.</p>
<pre>$ pgpool stop</pre>
<p>If any client is still connected, pgpool-II waits for them to
disconnect, and then terminate itself. Execute the following command
instead if you want to shutdown pgpool-II forcibly.
<pre>$ pgpool -m fast stop</pre>
<h2>2. <a name="replication">Your First Replication</a></h2>
<p>Replication enables the same data to be copied to multiple database
<p>In this section, we'll use three database nodes, which we have
already set up in section "1. <a href="#start">Let's Begin!</a>", and
takes you step by step to create a database replication system. Sample
data to be replicated will be generated by pgbench benchmark
<h3>2.1. <a name="replication-config">Configuring Replication</a></h3>
<p>To enable the database replication function, set
<code>replication_mode</code> to true in <code>pgpool.conf</code>
<pre>replication_mode = true</pre>
<p>When <code>replication_mode</code> is set to true, pgpool-II will send a
copy of a received query to all the database nodes.</p>
<p>When <code>load_balance_mode</code> is set to true, pgpool-II will
distribute SELECT queries among the database nodes.</p>
<pre>load_balance_mode = true</pre>
<p>In this section, we enable both <code>replication_mode</code> and
<h3>2.2. <a name="replication-check">Checking Replication</a></h3>
<p>To reflect the changes in <code>pgpool.conf</code>, pgpool-II must
be restarted. Please refer to section "1.5 <a
href="#start-shutdown">Starting/Stopping pgpool-II</a>".</p>
<p>After configuring <code>pgpool.conf</code> and restarting
pgpool-II, let's try the actual replication and see if everything is
working OK.</p>
<p>First, we need to create a database to be replicated. We will name
it "bench_replication". This database needs to be created on all the
nodes. Use <code>createdb</code> commands through pgpool-II, and the
database will be created on all the nodes.</p>
<pre>$ createdb -p 9999 bench_replication</pre>
<p>Then, we'll execute pgbench with <code>-i</code>
option. <code>-i</code> option initializes the database with
pre-defined tables and data.</p>
<pre>$ pgbench -i -p 9999 bench_replication</pre>
<p>The following table is the summary of tables and data, which will
be created by <code>pgbench -i</code>. If, on all the nodes, the
listed tables and data are created, replication is working correctly.
<table border="1" align="center">
<th>Table Name</th>
<th>Number of Rows</th>
<p>Let's use a simple shell script to check the above on all the
nodes. The following script will display the number of rows in
branches, tellers, accounts, and history tables on all the nodes (5432,
5433, 5434).</p>
<pre>$ for port in 5432 5433 5434; do
&gt; echo $port
&gt; for table_name in branches tellers accounts history; do
&gt; echo $table_name
&gt; psql -c &quot;SELECT count(*) FROM $table_name&quot; -p $port bench_replication
&gt; done
&gt; done
<h2>3. <a name="parallel">Your First Parallel Query</a></h2>
Data within the different range is stored in two or more data base nodes in a parallel Query. This is called a partitioning. Moreover, the same data as two or more data base nodes can be reproduced with partitioning.
<p>To enable parallel query in pgpool-II, you must set up another
database called "System Database" (we will denote it as SystemDB from
this point).</p>
<p>SystemDB holds the user-defined rules to decide what data will be
saved in which database node. Another use of SystemDB is to merge
results sent back from the database nodes using dblink.</p>
<p>In this section, we will use three database nodes which we have set
up in section "1. <a href="#start">Let's Begin!</a>", and takes you
step by step to create a parallel query database system. We will use
pgbench again to create sample data.</p>
<h3>3.1. <a name="parallel-config">Configuring Parallel Query</a></h3>
<p>To enable the parallel query function, set <code>parallel_mode</code> to true in <code>pgpool.conf</code> file.</p>
<pre>parallel_mode = true</pre>
<p>Setting <code>paralle_mode</code> to true does not start parallel
query automatically. pgpool-II needs SystemDB and the rules
to know how to distribute data to the database nodes.</p>
<p>Also, dblink used by SystemDB makes connections to
pgpool-II. Therefore, <code>listen_addresses</code> needs to be
configured so that pgpool-II accepts those connections.</p>
<pre>listen_addresses = '*'</pre>
Attention: The replication is not done for the table that does the partitioning
though a parallel Query and the replication can be made effective at the same time.
The data base made by Moreover, because the composition of the data stored in
the data base is different in a parallel Query and the replication, that is,
"bench_replication" created in section "2. <a href="#replication">Your
First Replication</a>" cannot be reused.</p>
replication_mode = true
load_balance_mode = false
replication_mode = false
load_balance_mode = true
<p>In this section, we will set <code>parallel_mode and load_balance_mode</code> to true,
<code>listen_addresses</code> to '*', <code>replication_mode</code>to false.</p>
<h3>3.2. <a name="system-db">Configuring SystemDB</a></h3>
There is no difference in the data base the system data base and usually.
However, the function of dblink is defined in the system data base, and
the table that stores a distribution rule.. It is necessary to define dist_def. Moreover,
the data base node One can make the system data base, and pgpool-II can be
distributed in the load by connecting the cascade.
<p>In this section, we will create SystemDB within the 5432 port
node. The following list is the configuration parameters for
system_db_hostname = 'localhost'
system_db_port = 5432
system_db_dbname = 'pgpool'
system_db_schema = 'pgpool_catalog'
system_db_user = 'pgpool'
system_db_password = ''
<p>Actually, the above are the default settings of
<code>pgpool.conf</code>. Now, we must create a user called "pgpool",
and a database called "pgpool" owned by user "pgpool".</p>
$ createuser -p 5432 pgpool
$ createdb -p 5432 -O pgpool pgpool
<h4><p>3.2.1. Installing dblink</p></h4>
<p>Next, we must install dblink into "pgpool" database. dblink is one
of the tools included in <code>contrib</code> directory in the
PostgreSQL source code.</p>
<p>To install dblink to your system, execute the following commands.</p>
$ USE_PGXS=1 make -C contrib/dblink
$ USE_PGXS=1 make -C contrib/dblink install
<p>After dblink has been installed into your system, we will define
dblink functions in "pgpool" database. If PostgreSQL is installed in
<code>/usr/local/pgsql</code>, <code>dblink.sql</code> (a file with
function definitions) should have been installed in
<code>/usr/local/pgsql/share/contrib</code>. Now, execute the
following command to define dblink functions.</p>
<pre>$ psql -f /usr/local/pgsql/share/contrib/dblink.sql -p 5432 pgpool</pre>
<h4><p>3.2.2. Defining dist_def table</p></h4>
<p>Next, we will define a table called "dist_def" to hold the
distribution rules. When pgpool-II was installed, a file called
<code>system_db.sql</code> should have been installed in
<code>/usr/local/share/system_db.sql</code> (note that in this
tutorial, we are using the default installation directory,
<code>/usr/local</code>). <code>systeym_db.sql</code> contains
definitions to create special tables including "dist_def"
table. Execute the following command to define "dist_def" table.</p>
<pre>$ psql -f /usr/local/share/system_db.sql -p 5432 -U pgpool pgpool</pre>
<p>In <code>system_db.sql</code>, tables including "dist_def" are
installed in "pgpool_catalog" schema. If you have configured
<code>system_db_schema</code> to use other schema, you need to edit
<code>system_db.sql</code> accordingly.</p>
<p>The definition for "dist_def" is as shown here, and the table name
cannot be changed.</p>
CREATE TABLE pgpool_catalog.dist_def (
dbname text, -- database name
schema_name text, -- schema name
table_name text, -- table name
col_name text NOT NULL CHECK (col_name = ANY (col_list)), -- distribution key-column
col_list text[] NOT NULL, -- list of column names
type_list text[] NOT NULL, -- list of column types
dist_def_func text NOT NULL, -- distribution function name
PRIMARY KEY (dbname, schema_name, table_name)
<p>A tuple stored in "dist_def" can be classified into two types.</p>
<li>Distribution Rule (col_name, dist_def_func)</li>
<li>Table's meta-information (dbname, schema_name, table_name, col_list, type_list)</li>
<p>A distribution rule decides how to distribute data to a
particular node. Data will be distributed depending on the value of
"col_name" column. "dist_def_func" is a function that takes the value
of "col_name" as its argument, and returns an integer which
points to the appropriate database node ID where the data should be
<p>A meta-information is used to rewrite queries. Parallel query
must rewrite queries so that the results sent back from the backend
nodes can be merged into one result.</p>
<h4><p>3.2.2. Defining replicate_def table</p></h4>
When the table that does the replication to one SQL sentence with the table
registered in dist_def by uniting tables is specified, information on the
table that does the replication (replication rule) is registered in the table
named replicate_def beforehand.
The replicate_def table has already been made when making it from the
system_db.sql file when dist_def is defined.
The replicate_def table is defined as follows.
CREATE TABLE pgpool_catalog.replicate_def (
dbname text, -- database name
schema_name text, -- schema name
table_name text, -- table name
col_list text[] NOT NULL, -- list of column names
type_list text[] NOT NULL, -- list of column types
PRIMARY KEY (dbname, schema_name, table_name)
<h3>3.3. <a name="dist-def">Defining Distribution Rules</a></h3>
<p>In this tutorial, we will define rules to distribute pgbench's
sample data into three database nodes. The sample data will be created
by "pgbench -i -s 3" (i.e. scale factor of 3). We will create a new
database called "bench_parallel" for this section.</p>
<p>In pgpool-II's source code, you can find
<code>dist_def_pgbench.sql</code> file in <code>sample</code>
directoy. We will use this sample file here to create distribution
rules for pgbench. Execute the following command in extracted
pgpool-II source code directory.</p>
<pre>$ psql -f sample/dist_def_pgbench.sql -p 5432 pgpool</pre>
<p>Here is the explanation of <code>dist_def_pgbench.sql</code>.</p>
<p>Inside <code>dist_def_pgbench.sql</code>, we are inserting one
row into "dist_def" table. There is a distribution
function for accounts table.
For key-column, aid is defined for accounts respectively (which is primary keys)</p>
INSERT INTO pgpool_catalog.dist_def VALUES (
ARRAY['aid', 'bid', 'abalance', 'filler'],
ARRAY['integer', 'integer', 'integer', 'character(84)'],
<p>Now, we must define the distribution function for accounts table. Note
that you can use the same function from different tables. Also, you
can define functions using languages other than SQL (e.g. PL/pgSQL,
PL/Tcl, etc.).</p>
The accounts table when data is initialized specifying 3 scale factor,
The value of the aid is 1 to 300000.
The function is defined so that data is evenly distributed to three data base nodes.
SQL function will be defined as the return of the number of the data base node.
<pre>CREATE OR REPLACE FUNCTION pgpool_catalog.dist_def_branches(anyelement)
RETURNS integer AS $$
SELECT CASE WHEN $1 &gt; 0 AND $1 &lt;= 1 THEN 0
WHEN $1 &gt; 1 AND $1 &lt;= 2 THEN 1
$$ LANGUAGE sql;
<h3>3.4. <a name="replicate-def">Defining Replication Rules</a></h3>
The replication rule is the one that which table decides the replication whether to be done.
Here, it is made with pgbench With the branches table and tellers table are registered.
As a result, the accounts table and the inquiry that uses the branches table
and the tellers table become possible.
INSERT INTO pgpool_catalog.replicate_def VALUES (
ARRAY['bid', 'bbalance', 'filler'],
ARRAY['integer', 'integer', 'character(88)']
INSERT INTO pgpool_catalog.replicate_def VALUES (
ARRAY['tid', 'bid', 'tbalance', 'filler'],
ARRAY['integer', 'integer', 'integer', 'character(84)']
Replicate_def_pgbench.sql is prepared in sample directory.
In the directory that progresses the source code to define a replicate rule by using this as follows The psql command is executed.
$ psql -f sample/replicate_def_pgbench.sql -p 5432 pgpool
<h3>3.5. <a name="parallel-check">Checking Parallel Query</a></h3>
<p>To reflect the changes in <code>pgpool.conf</code>, pgpool-II must
be restarted. Please refer to section "1.5 <a
href="#start-shutdown">Starting/Stopping pgpool-II</a>".</p>
<p>After configuring <code>pgpool.conf</code> and restarting
pgpool-II, let's try and see if parallel query is working OK.</p>
<p>First, we need to create a database to be distributed. We will name
it "bench_parallel". This database needs to be created on all the
nodes. Use <code>createdb</code> commands through pgpool-II, and the
database will be created on all the nodes.</p>
<pre>$ createdb -p 9999 bench_parallel</pre>
<p>Then, we'll execute pgbench with <code>-i -s 3</code>
options. <code>-i</code> option initializes the database with
pre-defined tables and data. <code>-s</code> option specifies the
scale factor for initialization.</p>
<pre>$ pgbench -i -s 3 -p 9999 bench_parallel</pre>
<p>The tables and data created are shown in "3.3. <a
href="#dist-def">Defining Distribution Rules</a>".</p>
<p>One way to check if the data have been distributed correctly is to
execute a SELECT query via pgpool-II and directly on the backend, and
compare two results. If everything is configured right,
"bench_parallel" should be distributed as follows.</p>
<table border="1" align ="center">
<th >Table Name</th>
<th >the number of lines</th>
<p>Let's use a simple shell script to check the above on all the nodes
and via pgpool-II. The following script will display the minimum and
maximum values in accounts table using port 5432, 5433, 5434, and
<pre>$ for port in 5432 5433 5434 9999; do
&gt; echo $port
&gt; psql -c &quot;SELECT min(aid), max(aid) FROM accounts&quot; -p $port bench_parallel
&gt; done
<div class="copyright">
Copyright &copy; 2003 &ndash; 2008 PgPool Global Development Group
Jump to Line
Something went wrong with that request. Please try again.