<!--

$Header: /cvsroot/pgsql/doc/src/sgml/ref/Attic/pg_upgrade.sgml,v 1.19 2002/01/11 20:48:47 momjian Exp $

$Header: /cvsroot/pgsql/doc/src/sgml/ref/Attic/pg_upgrade.sgml,v 1.20 2002/01/13 01:13:59 momjian Exp $

PostgreSQL documentation

-->

<refsynopsisdivinfo>

<date>1999-07-31</date>

</refsynopsisdivinfo>

<synopsis>

pg_upgrade -s <replaceable class="parameter">filename</replaceable> <replaceable class="parameter">old_data_dir</replaceable>

</synopsis>

<cmdsynopsis>

<command>pg_upgrade</command>

<group choice="plain"><arg>-1</arg><arg>-2</arg></group>

</cmdsynopsis>

</refsynopsisdiv>

<refsect1 id="R1-APP-PG-UPGRADE-1">

<step performance="required">

<para>

Then do:

<programlisting>

$ pg_dumpall -s > schema.out

</programlisting>

to dump out your old database's table definitions without any data.

</para>

</step>

<step performance="required">

<para>

<command>VACUUM</command> your entire database using

<command>vacuumdb -a</command.>

Copy the program <filename>pgsql/src/bin/pg_dump/pg_upgrade</filename>

from the current PostgreSQL distribution into somewhere in your path.

</para>

</step>

<step performance="required">

<para>

Stop the old postmaster and all backends.

</para>

</step>

<step performance="required">

<para>

Rename (using mv) your old pgsql <filename>data/</filename> directory to

<filename>data.old/</filename>.

Change your working directory to the

pgsql main directory, and type:

<programlisting>

$ pg_upgrade -1

</programlisting>

to collect information about the old database needed for the

upgrade.

</para>

</step>

<step performance="required">

<para>

Run <application>initdb</application> to create a new template1 database

containing the system tables for the new release.

containing the system tables for the new release. Make sure you use

settings similar to those used in your previous version.

</para>

</step>

<step performance="required">

<para>

Start the new postmaster. (Note: it is critical that no users connect

to thedatabase until the upgrade is complete. You may wish to start

to theserver until the upgrade is complete. You may wish to start

the postmaster without -i and/or alter <filename>pg_hba.conf</filename>

temporarily.)

</para>

Change your working directory to the

pgsql main directory, and type:

<programlisting>

$ pg_upgrade -s schema.out data.old

$ pg_upgrade -2

</programlisting>

The program will do some checking to make sure everything is properly

configured, and will run your db.out script to recreate all the databases

<step performance="required">

<para>

<emphasis>Carefully</emphasis> examine the contents of the upgraded

database. If you detect problems, you'll need to recover by restoring

databases. If you detect problems, you'll need to recover by restoring

from your full pg_dump backup.

You can delete the <filename>data.old/</filename> directory when you

are satisfied.

<step performance="required">

<para>

The upgradeddatabase will be in an un-vacuumed state. You will probably

The upgradeddatabases will be in an un-vacuumed state. You will probably

want to run a <command>VACUUM ANALYZE</command> before beginning

production work.

</para>

</step>

<note>

<para>

pg_upgrade does not handle custom tables/indexes/sequences in template1.

It does handle other template1 object customizations.

</para>

</note>

</procedure>

</refsect1>

</refentry>