| Server IP : 103.119.228.120 / Your IP : 3.137.159.134 Web Server : Apache System : Linux v8.techscape8.com 3.10.0-1160.119.1.el7.tuxcare.els2.x86_64 #1 SMP Mon Jul 15 12:09:18 UTC 2024 x86_64 User : nobody ( 99) PHP Version : 5.6.40 Disable Function : shell_exec,symlink,system,exec,proc_get_status,proc_nice,proc_terminate,define_syslog_variables,syslog,openlog,closelog,escapeshellcmd,passthru,ocinum cols,ini_alter,leak,listen,chgrp,apache_note,apache_setenv,debugger_on,debugger_off,ftp_exec,dl,dll,myshellexec,proc_open,socket_bind,proc_close,escapeshellarg,parse_ini_filepopen,fpassthru,exec,passthru,escapeshellarg,escapeshellcmd,proc_close,proc_open,ini_alter,popen,show_source,proc_nice,proc_terminate,proc_get_status,proc_close,pfsockopen,leak,apache_child_terminate,posix_kill,posix_mkfifo,posix_setpgid,posix_setsid,posix_setuid,dl,symlink,shell_exec,system,dl,passthru,escapeshellarg,escapeshellcmd,myshellexec,c99_buff_prepare,c99_sess_put,fpassthru,getdisfunc,fx29exec,fx29exec2,is_windows,disp_freespace,fx29sh_getupdate,fx29_buff_prepare,fx29_sess_put,fx29shexit,fx29fsearch,fx29ftpbrutecheck,fx29sh_tools,fx29sh_about,milw0rm,imagez,sh_name,myshellexec,checkproxyhost,dosyayicek,c99_buff_prepare,c99_sess_put,c99getsource,c99sh_getupdate,c99fsearch,c99shexit,view_perms,posix_getpwuid,posix_getgrgid,posix_kill,parse_perms,parsesort,view_perms_color,set_encoder_input,ls_setcheckboxall,ls_reverse_all,rsg_read,rsg_glob,selfURL,dispsecinfo,unix2DosTime,addFile,system,get_users,view_size,DirFiles,DirFilesWide,DirPrintHTMLHeaders,GetFilesTotal,GetTitles,GetTimeTotal,GetMatchesCount,GetFileMatchesCount,GetResultFiles,fs_copy_dir,fs_copy_obj,fs_move_dir,fs_move_obj,fs_rmdir,SearchText,getmicrotime MySQL : ON | cURL : ON | WGET : ON | Perl : ON | Python : ON | Sudo : ON | Pkexec : ON Directory : /usr/local/ssl/local/ssl/share/doc/postgresql-9.2.24/html/ |
Upload File : |
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Populating a Database</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PostgreSQL 9.2.24 Documentation"
HREF="index.html"><LINK
REL="UP"
TITLE="Performance Tips"
HREF="performance-tips.html"><LINK
REL="PREVIOUS"
TITLE="Controlling the Planner with Explicit JOIN Clauses"
HREF="explicit-joins.html"><LINK
REL="NEXT"
TITLE="Non-Durable Settings"
HREF="non-durability.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2017-11-06T22:43:11"></HEAD
><BODY
CLASS="SECT1"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
><A
HREF="index.html"
>PostgreSQL 9.2.24 Documentation</A
></TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Controlling the Planner with Explicit JOIN Clauses"
HREF="explicit-joins.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="performance-tips.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Chapter 14. Performance Tips</TD
><TD
WIDTH="20%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Non-Durable Settings"
HREF="non-durability.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="POPULATE"
>14.4. Populating a Database</A
></H1
><P
> One might need to insert a large amount of data when first populating
a database. This section contains some suggestions on how to make
this process as efficient as possible.
</P
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="DISABLE-AUTOCOMMIT"
>14.4.1. Disable Autocommit</A
></H2
><P
> When using multiple <TT
CLASS="COMMAND"
>INSERT</TT
>s, turn off autocommit and just do
one commit at the end. (In plain
SQL, this means issuing <TT
CLASS="COMMAND"
>BEGIN</TT
> at the start and
<TT
CLASS="COMMAND"
>COMMIT</TT
> at the end. Some client libraries might
do this behind your back, in which case you need to make sure the
library does it when you want it done.) If you allow each
insertion to be committed separately,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> is doing a lot of work for
each row that is added. An additional benefit of doing all
insertions in one transaction is that if the insertion of one row
were to fail then the insertion of all rows inserted up to that
point would be rolled back, so you won't be stuck with partially
loaded data.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-COPY-FROM"
>14.4.2. Use <TT
CLASS="COMMAND"
>COPY</TT
></A
></H2
><P
> Use <A
HREF="sql-copy.html"
>COPY</A
> to load
all the rows in one command, instead of using a series of
<TT
CLASS="COMMAND"
>INSERT</TT
> commands. The <TT
CLASS="COMMAND"
>COPY</TT
>
command is optimized for loading large numbers of rows; it is less
flexible than <TT
CLASS="COMMAND"
>INSERT</TT
>, but incurs significantly
less overhead for large data loads. Since <TT
CLASS="COMMAND"
>COPY</TT
>
is a single command, there is no need to disable autocommit if you
use this method to populate a table.
</P
><P
> If you cannot use <TT
CLASS="COMMAND"
>COPY</TT
>, it might help to use <A
HREF="sql-prepare.html"
>PREPARE</A
> to create a
prepared <TT
CLASS="COMMAND"
>INSERT</TT
> statement, and then use
<TT
CLASS="COMMAND"
>EXECUTE</TT
> as many times as required. This avoids
some of the overhead of repeatedly parsing and planning
<TT
CLASS="COMMAND"
>INSERT</TT
>. Different interfaces provide this facility
in different ways; look for <SPAN
CLASS="QUOTE"
>"prepared statements"</SPAN
> in the interface
documentation.
</P
><P
> Note that loading a large number of rows using
<TT
CLASS="COMMAND"
>COPY</TT
> is almost always faster than using
<TT
CLASS="COMMAND"
>INSERT</TT
>, even if <TT
CLASS="COMMAND"
>PREPARE</TT
> is used and
multiple insertions are batched into a single transaction.
</P
><P
> <TT
CLASS="COMMAND"
>COPY</TT
> is fastest when used within the same
transaction as an earlier <TT
CLASS="COMMAND"
>CREATE TABLE</TT
> or
<TT
CLASS="COMMAND"
>TRUNCATE</TT
> command. In such cases no WAL
needs to be written, because in case of an error, the files
containing the newly loaded data will be removed anyway.
However, this consideration only applies when
<A
HREF="runtime-config-wal.html#GUC-WAL-LEVEL"
>wal_level</A
> is <TT
CLASS="LITERAL"
>minimal</TT
> as all commands
must write WAL otherwise.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-RM-INDEXES"
>14.4.3. Remove Indexes</A
></H2
><P
> If you are loading a freshly created table, the fastest method is to
create the table, bulk load the table's data using
<TT
CLASS="COMMAND"
>COPY</TT
>, then create any indexes needed for the
table. Creating an index on pre-existing data is quicker than
updating it incrementally as each row is loaded.
</P
><P
> If you are adding large amounts of data to an existing table,
it might be a win to drop the indexes,
load the table, and then recreate the indexes. Of course, the
database performance for other users might suffer
during the time the indexes are missing. One should also think
twice before dropping a unique index, since the error checking
afforded by the unique constraint will be lost while the index is
missing.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-RM-FKEYS"
>14.4.4. Remove Foreign Key Constraints</A
></H2
><P
> Just as with indexes, a foreign key constraint can be checked
<SPAN
CLASS="QUOTE"
>"in bulk"</SPAN
> more efficiently than row-by-row. So it might be
useful to drop foreign key constraints, load data, and re-create
the constraints. Again, there is a trade-off between data load
speed and loss of error checking while the constraint is missing.
</P
><P
> What's more, when you load data into a table with existing foreign key
constraints, each new row requires an entry in the server's list of
pending trigger events (since it is the firing of a trigger that checks
the row's foreign key constraint). Loading many millions of rows can
cause the trigger event queue to overflow available memory, leading to
intolerable swapping or even outright failure of the command. Therefore
it may be <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>necessary</I
></SPAN
>, not just desirable, to drop and re-apply
foreign keys when loading large amounts of data. If temporarily removing
the constraint isn't acceptable, the only other recourse may be to split
up the load operation into smaller transactions.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-WORK-MEM"
>14.4.5. Increase <TT
CLASS="VARNAME"
>maintenance_work_mem</TT
></A
></H2
><P
> Temporarily increasing the <A
HREF="runtime-config-resource.html#GUC-MAINTENANCE-WORK-MEM"
>maintenance_work_mem</A
>
configuration variable when loading large amounts of data can
lead to improved performance. This will help to speed up <TT
CLASS="COMMAND"
>CREATE
INDEX</TT
> commands and <TT
CLASS="COMMAND"
>ALTER TABLE ADD FOREIGN KEY</TT
> commands.
It won't do much for <TT
CLASS="COMMAND"
>COPY</TT
> itself, so this advice is
only useful when you are using one or both of the above techniques.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-CHECKPOINT-SEGMENTS"
>14.4.6. Increase <TT
CLASS="VARNAME"
>checkpoint_segments</TT
></A
></H2
><P
> Temporarily increasing the <A
HREF="runtime-config-wal.html#GUC-CHECKPOINT-SEGMENTS"
>checkpoint_segments</A
> configuration variable can also
make large data loads faster. This is because loading a large
amount of data into <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> will
cause checkpoints to occur more often than the normal checkpoint
frequency (specified by the <TT
CLASS="VARNAME"
>checkpoint_timeout</TT
>
configuration variable). Whenever a checkpoint occurs, all dirty
pages must be flushed to disk. By increasing
<TT
CLASS="VARNAME"
>checkpoint_segments</TT
> temporarily during bulk
data loads, the number of checkpoints that are required can be
reduced.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-PITR"
>14.4.7. Disable WAL Archival and Streaming Replication</A
></H2
><P
> When loading large amounts of data into an installation that uses
WAL archiving or streaming replication, it might be faster to take a
new base backup after the load has completed than to process a large
amount of incremental WAL data. To prevent incremental WAL logging
while loading, disable archiving and streaming replication, by setting
<A
HREF="runtime-config-wal.html#GUC-WAL-LEVEL"
>wal_level</A
> to <TT
CLASS="LITERAL"
>minimal</TT
>,
<A
HREF="runtime-config-wal.html#GUC-ARCHIVE-MODE"
>archive_mode</A
> to <TT
CLASS="LITERAL"
>off</TT
>, and
<A
HREF="runtime-config-replication.html#GUC-MAX-WAL-SENDERS"
>max_wal_senders</A
> to zero.
But note that changing these settings requires a server restart.
</P
><P
> Aside from avoiding the time for the archiver or WAL sender to
process the WAL data,
doing this will actually make certain commands faster, because they
are designed not to write WAL at all if <TT
CLASS="VARNAME"
>wal_level</TT
>
is <TT
CLASS="LITERAL"
>minimal</TT
>. (They can guarantee crash safety more cheaply
by doing an <CODE
CLASS="FUNCTION"
>fsync</CODE
> at the end than by writing WAL.)
This applies to the following commands:
<P
></P
></P><UL
><LI
><P
> <TT
CLASS="COMMAND"
>CREATE TABLE AS SELECT</TT
>
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>CREATE INDEX</TT
> (and variants such as
<TT
CLASS="COMMAND"
>ALTER TABLE ADD PRIMARY KEY</TT
>)
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>ALTER TABLE SET TABLESPACE</TT
>
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>CLUSTER</TT
>
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>COPY FROM</TT
>, when the target table has been
created or truncated earlier in the same transaction
</P
></LI
></UL
><P>
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-ANALYZE"
>14.4.8. Run <TT
CLASS="COMMAND"
>ANALYZE</TT
> Afterwards</A
></H2
><P
> Whenever you have significantly altered the distribution of data
within a table, running <A
HREF="sql-analyze.html"
>ANALYZE</A
> is strongly recommended. This
includes bulk loading large amounts of data into the table. Running
<TT
CLASS="COMMAND"
>ANALYZE</TT
> (or <TT
CLASS="COMMAND"
>VACUUM ANALYZE</TT
>)
ensures that the planner has up-to-date statistics about the
table. With no statistics or obsolete statistics, the planner might
make poor decisions during query planning, leading to poor
performance on any tables with inaccurate or nonexistent
statistics. Note that if the autovacuum daemon is enabled, it might
run <TT
CLASS="COMMAND"
>ANALYZE</TT
> automatically; see
<A
HREF="routine-vacuuming.html#VACUUM-FOR-STATISTICS"
>Section 23.1.3</A
>
and <A
HREF="routine-vacuuming.html#AUTOVACUUM"
>Section 23.1.6</A
> for more information.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="POPULATE-PG-DUMP"
>14.4.9. Some Notes About <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
></A
></H2
><P
> Dump scripts generated by <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> automatically apply
several, but not all, of the above guidelines. To reload a
<SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> dump as quickly as possible, you need to
do a few extra things manually. (Note that these points apply while
<SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>restoring</I
></SPAN
> a dump, not while <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>creating</I
></SPAN
> it.
The same points apply whether loading a text dump with
<SPAN
CLASS="APPLICATION"
>psql</SPAN
> or using <SPAN
CLASS="APPLICATION"
>pg_restore</SPAN
> to load
from a <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> archive file.)
</P
><P
> By default, <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> uses <TT
CLASS="COMMAND"
>COPY</TT
>, and when
it is generating a complete schema-and-data dump, it is careful to
load data before creating indexes and foreign keys. So in this case
several guidelines are handled automatically. What is left
for you to do is to:
<P
></P
></P><UL
><LI
><P
> Set appropriate (i.e., larger than normal) values for
<TT
CLASS="VARNAME"
>maintenance_work_mem</TT
> and
<TT
CLASS="VARNAME"
>checkpoint_segments</TT
>.
</P
></LI
><LI
><P
> If using WAL archiving or streaming replication, consider disabling
them during the restore. To do that, set <TT
CLASS="VARNAME"
>archive_mode</TT
>
to <TT
CLASS="LITERAL"
>off</TT
>,
<TT
CLASS="VARNAME"
>wal_level</TT
> to <TT
CLASS="LITERAL"
>minimal</TT
>, and
<TT
CLASS="VARNAME"
>max_wal_senders</TT
> to zero before loading the dump.
Afterwards, set them back to the right values and take a fresh
base backup.
</P
></LI
><LI
><P
> Consider whether the whole dump should be restored as a single
transaction. To do that, pass the <TT
CLASS="OPTION"
>-1</TT
> or
<TT
CLASS="OPTION"
>--single-transaction</TT
> command-line option to
<SPAN
CLASS="APPLICATION"
>psql</SPAN
> or <SPAN
CLASS="APPLICATION"
>pg_restore</SPAN
>. When using this
mode, even the smallest of errors will rollback the entire restore,
possibly discarding many hours of processing. Depending on how
interrelated the data is, that might seem preferable to manual cleanup,
or not. <TT
CLASS="COMMAND"
>COPY</TT
> commands will run fastest if you use a single
transaction and have WAL archiving turned off.
</P
></LI
><LI
><P
> If multiple CPUs are available in the database server, consider using
<SPAN
CLASS="APPLICATION"
>pg_restore</SPAN
>'s <TT
CLASS="OPTION"
>--jobs</TT
> option. This
allows concurrent data loading and index creation.
</P
></LI
><LI
><P
> Run <TT
CLASS="COMMAND"
>ANALYZE</TT
> afterwards.
</P
></LI
></UL
><P>
</P
><P
> A data-only dump will still use <TT
CLASS="COMMAND"
>COPY</TT
>, but it does not
drop or recreate indexes, and it does not normally touch foreign
keys.
<A
NAME="AEN24507"
HREF="#FTN.AEN24507"
><SPAN
CLASS="footnote"
>[1]</SPAN
></A
>
So when loading a data-only dump, it is up to you to drop and recreate
indexes and foreign keys if you wish to use those techniques.
It's still useful to increase <TT
CLASS="VARNAME"
>checkpoint_segments</TT
>
while loading the data, but don't bother increasing
<TT
CLASS="VARNAME"
>maintenance_work_mem</TT
>; rather, you'd do that while
manually recreating indexes and foreign keys afterwards.
And don't forget to <TT
CLASS="COMMAND"
>ANALYZE</TT
> when you're done; see
<A
HREF="routine-vacuuming.html#VACUUM-FOR-STATISTICS"
>Section 23.1.3</A
>
and <A
HREF="routine-vacuuming.html#AUTOVACUUM"
>Section 23.1.6</A
> for more information.
</P
></DIV
></DIV
><H3
CLASS="FOOTNOTES"
>Notes</H3
><TABLE
BORDER="0"
CLASS="FOOTNOTES"
WIDTH="100%"
><TR
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="5%"
><A
NAME="FTN.AEN24507"
HREF="populate.html#AEN24507"
><SPAN
CLASS="footnote"
>[1]</SPAN
></A
></TD
><TD
ALIGN="LEFT"
VALIGN="TOP"
WIDTH="95%"
><P
> You can get the effect of disabling foreign keys by using
the <TT
CLASS="OPTION"
>--disable-triggers</TT
> option — but realize that
that eliminates, rather than just postpones, foreign key
validation, and so it is possible to insert bad data if you use it.
</P
></TD
></TR
></TABLE
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="explicit-joins.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="non-durability.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Controlling the Planner with Explicit <TT
CLASS="LITERAL"
>JOIN</TT
> Clauses</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="performance-tips.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Non-Durable Settings</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>