1 <!DOCTYPE html PUBLIC "-//W3C//DTD html 4.01 transitional//EN">
4 <!-- DOCTYPE html PUBLIC "-//W3C//DTD HTML 3.2//EN" -->
8 <META name="generator" content="HTML Tidy, see www.w3.org">
10 <TITLE>PostgreSQL FAQ</TITLE>
13 <BODY bgcolor="#ffffff" text="#000000" link="#ff0000" vlink="#a00000"
15 <H1>Frequently Asked Questions (FAQ) for PostgreSQL</H1>
17 <P>Last updated: Mon Jun 24 21:45:50 EDT 2002</P>
19 <P>Current maintainer: Bruce Momjian (<A href=
20 "mailto:pgman@candle.pha.pa.us">pgman@candle.pha.pa.us</A>)<BR>
23 <P>The most recent version of this document can be viewed at <A
25 "http://www.Postgresql.org/docs/faq-english.html">http://www.PostgreSQL.org/docs/faq-english.html</A>.</P>
27 <P>Platform-specific questions are answered at <A href=
28 "http://www.PostgreSQL.org/users-lounge/docs/faq.html">http://www.PostgreSQL.org/users-lounge/docs/faq.html</A>.</P>
31 <H2 align="center">General Questions</H2>
32 <A href="#1.1">1.1</A>) What is PostgreSQL? How is it
34 <A href="#1.2">1.2</A>) What is the copyright on PostgreSQL?<BR>
35 <A href="#1.3">1.3</A>) What Unix platforms does PostgreSQL run
37 <A href="#1.4">1.4</A>) What non-Unix ports are available?<BR>
38 <A href="#1.5">1.5</A>) Where can I get PostgreSQL?<BR>
39 <A href="#1.6">1.6</A>) Where can I get support?<BR>
40 <A href="#1.7">1.7</A>) What is the latest release?<BR>
41 <A href="#1.8">1.8</A>) What documentation is available?<BR>
42 <A href="#1.9">1.9</A>) How do I find out about known bugs or
44 <A href="#1.10">1.10</A>) How can I learn <SMALL>SQL</SMALL>?<BR>
45 <A href="#1.11">1.11</A>) Is PostgreSQL Y2K compliant?<BR>
46 <A href="#1.12">1.12</A>) How do I join the development team?<BR>
47 <A href="#1.13">1.13</A>) How do I submit a bug report?<BR>
48 <A href="#1.14">1.14</A>) How does PostgreSQL compare to other
49 <SMALL>DBMS</SMALL>s?<BR>
50 <A href="#1.15">1.15</A>) How can I financially assist
54 <H2 align="center">User Client Questions</H2>
55 <A href="#2.1">2.1</A>) Are there <SMALL>ODBC</SMALL> drivers for
57 <A href="#2.2">2.2</A>) What tools are available for using
58 PostgreSQL with Web pages?<BR>
59 <A href="#2.3">2.3</A>) Does PostgreSQL have a graphical user
60 interface? A report generator? An embedded query language
62 <A href="#2.4">2.4</A>) What languages are available to
63 communicate with PostgreSQL?<BR>
66 <H2 align="center">Administrative Questions</H2>
67 <A href="#3.1">3.1</A>) How do I install PostgreSQL somewhere other
68 than <I>/usr/local/pgsql</I>?<BR>
69 <A href="#3.2">3.2</A>) When I start <I>postmaster</I>, I get a
70 <I>Bad System Call</I> or core dumped message. Why?<BR>
71 <A href="#3.3">3.3</A>) When I try to start <I>postmaster</I>, I
72 get <I>IpcMemoryCreate</I> errors. Why?<BR>
73 <A href="#3.4">3.4</A>) When I try to start <I>postmaster</I>, I
74 get <I>IpcSemaphoreCreate</I> errors. Why?<BR>
75 <A href="#3.5">3.5</A>) How do I control connections from other
77 <A href="#3.6">3.6</A>) How do I tune the database engine for
78 better performance?<BR>
79 <A href="#3.7">3.7</A>) What debugging features are available?<BR>
80 <A href="#3.8">3.8</A>) Why do I get <I>"Sorry, too many
81 clients"</I> when trying to connect?<BR>
82 <A href="#3.9">3.9</A>) What are the <I>pg_sorttempNNN.NN</I>
83 files in my database directory?<BR>
84 <A href="#3.10">3.10</A>) Why do I need to do a dump and restore
85 to upgrade PostgreSQL?<BR>
88 <H2 align="center">Operational Questions</H2>
89 <A href="#4.1">4.1</A>) What is the difference between binary
90 cursors and normal cursors?<BR>
91 <A href="#4.2">4.2</A>) How do I <SMALL>SELECT</SMALL> only the
92 first few rows of a query?<BR>
93 <A href="#4.3">4.3</A>) How do I get a list of tables or other
94 things I can see in <I>psql</I>?<BR>
95 <A href="#4.4">4.4</A>) How do you remove a column from a
97 <A href="#4.5">4.5</A>) What is the maximum size for a row, a
98 table, and a database?<BR>
99 <A href="#4.6">4.6</A>) How much database disk space is required
100 to store data from a typical text file?<BR>
101 <A href="#4.7">4.7</A>) How do I find out what tables, indexes,
102 databases, and users are defined?<BR>
103 <A href="#4.8">4.8</A>) My queries are slow or don't make use of
104 the indexes. Why?<BR>
105 <A href="#4.9">4.9</A>) How do I see how the query optimizer is
106 evaluating my query?<BR>
107 <A href="#4.10">4.10</A>) What is an R-tree index?<BR>
108 <A href="#4.11">4.11</A>) What is the Genetic Query Optimizer?<BR>
109 <A href="#4.12">4.12</A>) How do I perform regular expression
110 searches and case-insensitive regular expression searches? How do I
111 use an index for case-insensitive searches?<BR>
112 <A href="#4.13">4.13</A>) In a query, how do I detect if a field
113 is <SMALL>NULL</SMALL>?<BR>
114 <A href="#4.14">4.14</A>) What is the difference between the
115 various character types?<BR>
116 <A href="#4.15.1">4.15.1</A>) How do I create a
117 serial/auto-incrementing field?<BR>
118 <A href="#4.15.2">4.15.2</A>) How do I get the value of a
119 <SMALL>SERIAL</SMALL> insert?<BR>
120 <A href="#4.15.3">4.15.3</A>) Don't <I>currval()</I> and
121 <I>nextval()</I> lead to a race condition with other users?<BR>
122 <A href="#4.15.4">4.15.4</A>) Why aren't my sequence numbers
123 reused on transaction abort? Why are there gaps in the numbering of
124 my sequence/SERIAL column?<BR>
125 <A href="#4.16">4.16</A>) What is an <SMALL>OID</SMALL>? What is a
126 <SMALL>TID</SMALL>?<BR>
127 <A href="#4.17">4.17</A>) What is the meaning of some of the terms
128 used in PostgreSQL?<BR>
129 <A href="#4.18">4.18</A>) Why do I get the error <I>"ERROR: Memory
130 exhausted in AllocSetAlloc()"</I>?<BR>
131 <A href="#4.19">4.19</A>) How do I tell what PostgreSQL version I
133 <A href="#4.20">4.20</A>) Why does my large-object operations get
134 <I>"invalid large obj descriptor"</I>?<BR>
135 <A href="#4.21">4.21</A>) How do I create a column that will
136 default to the current time?<BR>
137 <A href="#4.22">4.22</A>) Why are my subqueries using
138 <CODE><SMALL>IN</SMALL></CODE> so slow?<BR>
139 <A href="#4.23">4.23</A>) How do I perform an outer join?<BR>
140 <A href="#4.24">4.24</A>) How do I perform queries using multiple
142 <A href="#4.25">4.25</A>) How do I return multiple rows or columns
144 <A href="#4.26">4.26</A>) Why can't I reliably create/drop
145 temporary tables in PL/PgSQL functions?<BR>
148 <H2 align="center">Extending PostgreSQL</H2>
149 <A href="#5.1">5.1</A>) I wrote a user-defined function. When I run
150 it in <I>psql</I>, why does it dump core?<BR>
151 <A href="#5.2">5.2</A>) How can I contribute some nifty new types
152 and functions to PostgreSQL?<BR>
153 <A href="#5.3">5.3</A>) How do I write a C function to return a
155 <A href="#5.4">5.4</A>) I have changed a source file. Why does the
156 recompile not see the change?<BR>
160 <H2 align="center">General Questions</H2>
162 <H4><A name="1.1">1.1</A>) What is PostgreSQL?</H4>
164 <P>PostgreSQL is pronounced <I>Post-Gres-Q-L</I>.</P>
166 <P>PostgreSQL is an enhancement of the POSTGRES database management
167 system, a next-generation <SMALL>DBMS</SMALL> research prototype.
168 While PostgreSQL retains the powerful data model and rich data
169 types of POSTGRES, it replaces the PostQuel query language with an
170 extended subset of <SMALL>SQL</SMALL>. PostgreSQL is free and the
171 complete source is available.</P>
173 <P>PostgreSQL development is performed by a team of Internet
174 developers who all subscribe to the PostgreSQL development mailing
175 list. The current coordinator is Marc G. Fournier (<A href=
176 "mailto:scrappy@PostgreSQL.org">scrappy@PostgreSQL.org</A>). (See
177 below on how to join). This team is now responsible for all
178 development of PostgreSQL.</P>
180 <P>The authors of PostgreSQL 1.01 were Andrew Yu and Jolly Chen.
181 Many others have contributed to the porting, testing, debugging,
182 and enhancement of the code. The original Postgres code, from which
183 PostgreSQL is derived, was the effort of many graduate students,
184 undergraduate students, and staff programmers working under the
185 direction of Professor Michael Stonebraker at the University of
186 California, Berkeley.</P>
188 <P>The original name of the software at Berkeley was Postgres. When
189 <SMALL>SQL</SMALL> functionality was added in 1995, its name was
190 changed to Postgres95. The name was changed at the end of 1996 to
193 <H4><A name="1.2">1.2</A>) What is the copyright on
196 <P>PostgreSQL is subject to the following COPYRIGHT:</P>
198 <P>PostgreSQL Data Base Management System</P>
200 <P>Portions copyright (c) 1996-2002, PostgreSQL Global Development
201 Group Portions Copyright (c) 1994-6 Regents of the University of
204 <P>Permission to use, copy, modify, and distribute this software
205 and its documentation for any purpose, without fee, and without a
206 written agreement is hereby granted, provided that the above
207 copyright notice and this paragraph and the following two
208 paragraphs appear in all copies.</P>
210 <P>IN NO EVENT SHALL THE UNIVERSITY OF CALIFORNIA BE LIABLE TO ANY
211 PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL
212 DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE USE OF THIS
213 SOFTWARE AND ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF
214 CALIFORNIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.</P>
216 <P>THE UNIVERSITY OF CALIFORNIA SPECIFICALLY DISCLAIMS ANY
217 WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
218 OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE
219 SOFTWARE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE
220 UNIVERSITY OF CALIFORNIA HAS NO OBLIGATIONS TO PROVIDE MAINTENANCE,
221 SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.</P>
223 <P>The above is the BSD license, the classic open-source license.
224 It has no restrictions on how the source code may be used. We like
225 it and have no intention of changing it.</P>
227 <H4><A name="1.3">1.3</A>) What Unix platforms does PostgreSQL run
230 <P>In general, a modern Unix-compatible platform should be able to
231 run PostgreSQL. The platforms that had received explicit testing at
232 the time of release are listed in the installation
235 <H4><A name="1.4">1.4</A>) What non-Unix ports are available?</H4>
237 <P><STRONG>Client</STRONG></P>
239 <P>It is possible to compile the <I>libpq</I> C library, psql, and
240 other interfaces and binaries to run on MS Windows platforms. In
241 this case, the client is running on MS Windows, and communicates
242 via TCP/IP to a server running on one of our supported Unix
243 platforms. A file <I>win31.mak</I> is included in the distribution
244 for making a Win32 <I>libpq</I> library and <I>psql</I>. PostgreSQL
245 also communicates with <SMALL>ODBC</SMALL> clients.</P>
247 <P><STRONG>Server</STRONG></P>
249 <P>The database server can run on Windows NT and Win2k using
250 Cygwin, the Cygnus Unix/NT porting library. See
251 <I>pgsql/doc/FAQ_MSWIN</I> in the distribution or the <A href=
252 "http://www.postgresql.org/docs/faq-mswin.html">MS Windows FAQ</A>
253 on our web site. We have no plan to do a native port to any
254 Microsoft platform.</P>
256 <H4><A name="1.5">1.5</A>) Where can I get PostgreSQL?</H4>
258 <P>The primary anonymous ftp site for PostgreSQL is <A href=
259 "ftp://ftp.PostgreSQL.org/pub">ftp://ftp.PostgreSQL.org/pub</A>.
260 For mirror sites, see our main web site.</P>
262 <H4><A name="1.6">1.6</A>) Where can I get support?</H4>
264 <P>The main mailing list is: <A href=
265 "mailto:pgsql-general@PostgreSQL.org">pgsql-general@PostgreSQL.org</A>.
266 It is available for discussion of matters pertaining to PostgreSQL.
267 To subscribe, send mail with the following lines in the body (not
268 the subject line):</P>
275 "mailto:pgsql-general-request@PostgreSQL.org">pgsql-general-request@PostgreSQL.org</A>.</P>
277 <P>There is also a digest list available. To subscribe to this
278 list, send email to: <A href=
279 "mailto:pgsql-general-digest-request@PostgreSQL.org">pgsql-general-digest-request@PostgreSQL.org</A>
285 Digests are sent out to members of this list whenever the main list
286 has received around 30k of messages.
288 <P>The bugs mailing list is available. To subscribe to this list,
289 send email to <A href=
290 "mailto:pgsql-bugs-request@PostgreSQL.org">pgsql-bugs-request@PostgreSQL.org</A>
296 There is also a developers discussion mailing list available. To
297 subscribe to this list, send email to <A href=
298 "mailto:pgsql-hackers-request@PostgreSQL.org">pgsql-hackers-request@PostgreSQL.org</A>
305 <P>Additional mailing lists and information about PostgreSQL can be
306 found via the PostgreSQL WWW home page at:</P>
309 <A href="http://www.PostgreSQL.org">http://www.PostgreSQL.org</A>
312 <P>There is also an IRC channel on EFNet, channel
313 <I>#PostgreSQL</I>. I use the Unix command <CODE>irc -c
314 '#PostgreSQL' "$USER" irc.phoenix.net.</CODE></P>
316 <P>A list of commercial support companies is available at <A href=
317 "http://www.postgresql.org/users-lounge/commercial-support.html">http://www.postgresql.org/users-lounge/commercial-support.html</A>.</P>
319 <H4><A name="1.7">1.7</A>) What is the latest release?</H4>
321 <P>The latest release of PostgreSQL is version 7.2.1.</P>
323 <P>We plan to have major releases every four months.</P>
325 <H4><A name="1.8">1.8</A>) What documentation is available?</H4>
327 <P>Several manuals, manual pages, and some small test examples are
328 included in the distribution. See the <I>/doc</I> directory. You
329 can also browse the manual online at <A href=
330 "http://www.PostgreSQL.org/users-lounge/docs/">http://www.PostgreSQL.org/users-lounge/docs/</A>.</P>
332 <P>There are two PostgreSQL books available online at <A href=
333 "http://www.PostgreSQL.org/docs/awbook.html">http://www.PostgreSQL.org/docs/awbook.html</A>
335 "http://www.commandprompt.com/ppbook/">http://www.commandprompt.com/ppbook/</A>.
336 There is a list of PostgreSQL books available for purchase at <A
338 "http://www.postgresql.org/books/">http://www.postgresql.org/books/</A>.
339 There is also a collection of PostgreSQL technical articles at <A
341 "http://techdocs.postgresql.org/">http://techdocs.postgresql.org/</A>.</P>
343 <P><I>psql</I> has some nice \d commands to show information about
344 types, operators, functions, aggregates, etc.</P>
346 <P>Our web site contains even more documentation.</P>
348 <H4><A name="1.9">1.9</A>) How do I find out about known bugs or
349 missing features?</H4>
351 <P>PostgreSQL supports an extended subset of <SMALL>SQL</SMALL>-92.
352 See our <A href="http://developer.PostgreSQL.org/todo.php">TODO</A>
353 list for known bugs, missing features, and future plans.</P>
355 <H4><A name="1.10">1.10</A>) How can I learn
356 <SMALL>SQL</SMALL>?</H4>
358 <P>The PostgreSQL book at <A href=
359 "http://www.PostgreSQL.org/docs/awbook.html">http://www.PostgreSQL.org/docs/awbook.html</A>
360 teaches <SMALL>SQL</SMALL>. There is another PostgreSQL book at <A
362 "http://www.commandprompt.com/ppbook/">http://www.commandprompt.com/ppbook.</A>
363 There is a nice tutorial at <A href=
364 "http://www.intermedia.net/support/sql/sqltut.shtm">http://www.intermedia.net/support/sql/sqltut.shtm,</A>
366 "http://ourworld.compuserve.com/homepages/graeme_birchall/HTM_COOK.HTM">
367 http://ourworld.compuserve.com/homepages/graeme_birchall/HTM_COOK.HTM,</A>
369 "http://sqlcourse.com/">http://sqlcourse.com.</A></P>
371 <P>Another one is "Teach Yourself SQL in 21 Days, Second Edition"
373 "http://members.tripod.com/er4ebus/sql/index.htm">http://members.tripod.com/er4ebus/sql/index.htm</A></P>
375 <P>Many of our users like <I>The Practical SQL Handbook</I>,
376 Bowman, Judith S., et al., Addison-Wesley. Others like <I>The
377 Complete Reference SQL</I>, Groff et al., McGraw-Hill.</P>
379 <H4><A name="1.11">1.11</A>) Is PostgreSQL Y2K compliant?</H4>
381 <P>Yes, we easily handle dates past the year 2000 AD, and before
384 <H4><A name="1.12">1.12</A>) How do I join the development
387 <P>First, download the latest source and read the PostgreSQL
388 Developers documentation on our web site, or in the distribution.
389 Second, subscribe to the <I>pgsql-hackers</I> and
390 <I>pgsql-patches</I> mailing lists. Third, submit high quality
391 patches to pgsql-patches.</P>
393 <P>There are about a dozen people who have commit privileges to the
394 PostgreSQL <SMALL>CVS</SMALL> archive. They each have submitted so
395 many high-quality patches that it was impossible for the existing
396 committers to keep up, and we had confidence that patches they
397 committed were of high quality.</P>
399 <H4><A name="1.13">1.13</A>) How do I submit a bug report?</H4>
401 <P>Please visit the <A href=
402 "http://www.postgresql.org/bugs/bugs.php">PostgreSQL BugTool</A>
403 page, which gives guidelines and directions on how to submit a
406 <P>Also check out our ftp site <A href=
407 "ftp://ftp.PostgreSQL.org/pub">ftp://ftp.PostgreSQL.org/pub</A> to
408 see if there is a more recent PostgreSQL version or patches.</P>
410 <H4><A name="1.14">1.14</A>) How does PostgreSQL compare to other
411 <SMALL>DBMS</SMALL>s?</H4>
413 <P>There are several ways of measuring software: features,
414 performance, reliability, support, and price.</P>
417 <DT><B>Features</B></DT>
419 <DD>PostgreSQL has most features present in large commercial
420 <SMALL>DBMS</SMALL>s, like transactions, subselects, triggers,
421 views, foreign key referential integrity, and sophisticated
422 locking. We have some features they do not have, like
423 user-defined types, inheritance, rules, and multi-version
424 concurrency control to reduce lock contention.<BR>
428 <DT><B>Performance</B></DT>
430 <DD>PostgreSQL has performance similar to other commercial and
431 open source databases. it is faster for some things, slower for
432 others. In comparison to MySQL or leaner database systems, we are
433 slower on inserts/updates because of transaction overhead. Of
434 course, MySQL does not have any of the features mentioned in the
435 <I>Features</I> section above. We are built for reliability and
436 features, though we continue to improve performance in every
437 release. There is an interesting Web page comparing PostgreSQL to
438 MySQL at <A href= "http://openacs.org/why-not-mysql.html">
440 http://openacs.org/why-not-mysql.html</A><BR>
445 <DT><B>Reliability</B></DT>
447 <DD>We realize that a <SMALL>DBMS</SMALL> must be reliable, or it
448 is worthless. We strive to release well-tested, stable code that
449 has a minimum of bugs. Each release has at least one month of
450 beta testing, and our release history shows that we can provide
451 stable, solid releases that are ready for production use. We
452 believe we compare favorably to other database software in this
457 <DT><B>Support</B></DT>
459 <DD>Our mailing list provides a large group of developers and
460 users to help resolve any problems encountered. While we can not
461 guarantee a fix, commercial <SMALL>DBMS</SMALL>s do not always
462 supply a fix either. Direct access to developers, the user
463 community, manuals, and the source code often make PostgreSQL
464 support superior to other <SMALL>DBMS</SMALL>s. There is
465 commercial per-incident support available for those who need it.
466 (See <A href="#1.6">support FAQ item</A>.)<BR>
470 <DT><B>Price</B></DT>
472 <DD>We are free for all use, both commercial and non-commercial.
473 You can add our code to your product with no limitations, except
474 those outlined in our BSD-style license stated above.<BR>
479 <H4><A name="1.15">1.15</A>) How can I financially assist
482 <P>PostgreSQL has had a first-class infrastructure since we started
483 six years ago. This is all thanks to Marc Fournier, who has created
484 and managed this infrastructure over the years.</P>
486 <P>Quality infrastructure is very important to an open-source
487 project. It prevents disruptions that can greatly delay forward
488 movement of the project.</P>
490 <P>Of course, this infrastructure is not cheap. There are a variety
491 of monthly and one-time expenses that are required to keep it
492 going. If you or your company has money it can donate to help fund
493 this effort, please go to <A href=
494 "http://www.pgsql.com/pg_goodies">http://www.pgsql.com/pg_goodies</A>
495 and make a donation.</P>
497 <P>Although the web page mentions PostgreSQL, Inc, the
498 "contributions" item is solely to support the PostgreSQL project
499 and does not fund any specific company. If you prefer, you can also
500 send a check to the contact address.</P>
503 <H2 align="center">User Client Questions</H2>
505 <H4><A name="2.1">2.1</A>) Are there <SMALL>ODBC</SMALL> drivers
508 <P>There are two <SMALL>ODBC</SMALL> drivers available, PsqlODBC
509 and OpenLink <SMALL>ODBC</SMALL>.</P>
511 <P>PsqlODBC is included in the distribution. More information about
512 it can be gotten from <A href=
513 "ftp://ftp.PostgreSQL.org/pub/odbc/">ftp://ftp.PostgreSQL.org/pub/odbc/</A>.</P>
515 <P>OpenLink <SMALL>ODBC</SMALL> can be gotten from <A href=
516 "http://www.openlinksw.com/">http://www.openlinksw.com</A>. It
517 works with their standard <SMALL>ODBC</SMALL> client software so
518 you'll have PostgreSQL <SMALL>ODBC</SMALL> available on every
519 client platform they support (Win, Mac, Unix, VMS).</P>
521 <P>They will probably be selling this product to people who need
522 commercial-quality support, but a freeware version will always be
523 available. Please send questions to <A href=
524 "mailto:postgres95@openlink.co.uk">postgres95@openlink.co.uk</A>.</P>
526 <P>See also the <A href=
527 "http://www.postgresql.org/devel-corner/docs/programmer/odbc.html">ODBC
528 chapter of the Programmer's Guide</A>.</P>
530 <H4><A name="2.2">2.2</A>) What tools are available for using
531 PostgreSQL with Web pages?</H4>
533 <P>A nice introduction to Database-backed Web pages can be seen at:
534 <A href="http://www.webreview.com">http://www.webreview.com</A></P>
536 <P>There is also one at <A href=
537 "http://www.phone.net/home/mwm/hotlist/">http://www.phone.net/home/mwm/hotlist/.</A></P>
539 <P>For Web integration, PHP is an excellent interface. It is at <A
540 href="http://www.php.net">http://www.php.net</A>.</P>
542 <P>For complex cases, many use the Perl interface and CGI.pm.</P>
544 <H4><A name="2.3">2.3</A>) Does PostgreSQL have a graphical user
545 interface? A report generator? An embedded query language
548 <P>We have a nice graphical user interface called <I>pgaccess</I>,
549 which is shipped as part of the distribution. <I>pgaccess</I> also
550 has a report generator. The Web page is <A href=
551 "http://www.flex.ro/pgaccess">http://www.flex.ro/pgaccess</A></P>
553 <P>We also include <I>ecpg</I>, which is an embedded SQL query
554 language interface for C.</P>
556 <H4><A name="2.4">2.4</A>) What languages are available to
557 communicate with PostgreSQL?</H4>
564 <LI>C++ (libpq++)</LI>
566 <LI>Embedded C (ecpg)</LI>
570 <LI>Perl (perl5)</LI>
574 <LI>Python (PyGreSQL)</LI>
576 <LI>TCL (libpgtcl)</LI>
578 <LI>C Easy API (libpgeasy)</LI>
580 <LI>Embedded <SMALL>HTML</SMALL> (<A href=
581 "http://www.php.net">PHP from http://www.php.net</A>)</LI>
585 <H2 align="center">Administrative Questions</H2>
587 <H4><A name="3.1">3.1</A>) How do I install PostgreSQL somewhere
588 other than <I>/usr/local/pgsql</I>?</H4>
590 <P>Specify the <I>--prefix</I> option when running
591 <I>configure</I>.</P>
593 <H4><A name="3.2">3.2</A>) When I start <I>postmaster</I>, I get a
594 <I>Bad System Call</I> or core dumped message. Why?</H4>
596 <P>It could be a variety of problems, but first check to see that
597 you have System V extensions installed in your kernel. PostgreSQL
598 requires kernel support for shared memory and semaphores.</P>
600 <H4><A name="3.3">3.3</A>) When I try to start <I>postmaster</I>, I
601 get <I>IpcMemoryCreate</I> errors. Why?</H4>
603 <P>You either do not have shared memory configured properly in your
604 kernel or you need to enlarge the shared memory available in the
605 kernel. The exact amount you need depends on your architecture and
606 how many buffers and backend processes you configure for
607 <I>postmaster</I>. For most systems, with default numbers of
608 buffers and processes, you need a minimum of ~1 MB. See the <A
610 "http://www.postgresql.org/idocs/index.php?kernel-resources.html">PostgreSQL
611 Administrator's Guide</A> for more detailed information about
612 shared memory and semaphores.</P>
614 <H4><A name="3.4">3.4</A>) When I try to start <I>postmaster</I>, I
615 get <I>IpcSemaphoreCreate</I> errors. Why?</H4>
617 <P>If the error message is <I>IpcSemaphoreCreate: semget failed (No
618 space left on device)</I> then your kernel is not configured with
619 enough semaphores. Postgres needs one semaphore per potential
620 backend process. A temporary solution is to start <I>postmaster</I>
621 with a smaller limit on the number of backend processes. Use
622 <I>-N</I> with a parameter less than the default of 32. A more
623 permanent solution is to increase your kernel's
624 <SMALL>SEMMNS</SMALL> and <SMALL>SEMMNI</SMALL> parameters.</P>
626 <P>Inoperative semaphores can also cause crashes during heavy
629 <P>If the error message is something else, you might not have
630 semaphore support configured in your kernel at all. See the
631 PostgreSQL Administrator's Guide for more detailed information
632 about shared memory and semaphores.</P>
634 <H4><A name="3.5">3.5</A>) How do I control connections from other
637 <P>By default, PostgreSQL only allows connections from the local
638 machine using Unix domain sockets. Other machines will not be able
639 to connect unless you add the <I>-i</I> flag to <I>postmaster</I>,
640 <B>and</B> enable host-based authentication by modifying the file
641 <I>$PGDATA/pg_hba.conf</I> accordingly. This will allow TCP/IP
644 <H4><A name="3.6">3.6</A>) How do I tune the database engine for
645 better performance?</H4>
647 <P>Certainly, indexes can speed up queries. The
648 <SMALL>EXPLAIN</SMALL> command allows you to see how PostgreSQL is
649 interpreting your query, and which indexes are being used.</P>
651 <P>If you are doing many <SMALL>INSERTs</SMALL>, consider doing
652 them in a large batch using the <SMALL>COPY</SMALL> command. This
653 is much faster than individual <SMALL>INSERTS.</SMALL> Second,
654 statements not in a <SMALL>BEGIN WORK/COMMIT</SMALL> transaction
655 block are considered to be in their own transaction. Consider
656 performing several statements in a single transaction block. This
657 reduces the transaction overhead. Also, consider dropping and
658 recreating indexes when making large data changes.</P>
660 <P>There are several tuning options. You can disable <I>fsync()</I>
661 by starting <I>postmaster</I> with a <I>-o -F</I> option. This will
662 prevent <I>fsync()</I>s from flushing to disk after every
665 <P>You can also use the <I>postmaster</I> <I>-B</I> option to
666 increase the number of shared memory buffers used by the backend
667 processes. If you make this parameter too high, the
668 <I>postmaster</I> may not start because you have exceeded your
669 kernel's limit on shared memory space. Each buffer is 8K and the
670 default is 64 buffers.</P>
672 <P>You can also use the backend <I>-S</I> option to increase the
673 maximum amount of memory used by the backend process for temporary
674 sorts. The <I>-S</I> value is measured in kilobytes, and the
675 default is 512 (i.e. 512K).</P>
677 <P>You can also use the <SMALL>CLUSTER</SMALL> command to group
678 data in tables to match an index. See the <SMALL>CLUSTER</SMALL>
679 manual page for more details.</P>
681 <H4><A name="3.7">3.7</A>) What debugging features are
684 <P>PostgreSQL has several features that report status information
685 that can be valuable for debugging purposes.</P>
687 <P>First, by running <I>configure</I> with the --enable-cassert
688 option, many <I>assert()</I>s monitor the progress of the backend
689 and halt the program when something unexpected occurs.</P>
691 <P>Both <I>postmaster</I> and <I>postgres</I> have several debug
692 options available. First, whenever you start <I>postmaster</I>,
693 make sure you send the standard output and error to a log file,
697 ./bin/postmaster >server.log 2>&1 &
700 <P>This will put a server.log file in the top-level PostgreSQL
701 directory. This file contains useful information about problems or
702 errors encountered by the server. <I>Postmaster</I> has a <I>-d</I>
703 option that allows even more detailed information to be reported.
704 The <I>-d</I> option takes a number that specifies the debug level.
705 Be warned that high debug level values generate large log
708 <P>If <I>postmaster</I> is not running, you can actually run the
709 <I>postgres</I> backend from the command line, and type your
710 <SMALL>SQL</SMALL> statement directly. This is recommended
711 <B>only</B> for debugging purposes. Note that a newline terminates
712 the query, not a semicolon. If you have compiled with debugging
713 symbols, you can use a debugger to see what is happening. Because
714 the backend was not started from <I>postmaster</I>, it is not
715 running in an identical environment and locking/backend interaction
716 problems may not be duplicated.</P>
718 <P>If <I>postmaster</I> is running, start <I>psql</I> in one
719 window, then find the <SMALL>PID</SMALL> of the <I>postgres</I>
720 process used by <I>psql</I>. Use a debugger to attach to the
721 <I>postgres</I> <SMALL>PID.</SMALL> You can set breakpoints in the
722 debugger and issue queries from <I>psql</I>. If you are debugging
723 <I>postgres</I> startup, you can set PGOPTIONS="-W n", then start
724 <I>psql</I>. This will cause startup to delay for <I>n</I> seconds
725 so you can attach to the process with the debugger, set any
726 breakpoints, and continue through the startup sequence.</P>
728 <P>The <I>postgres</I> program has <I>-s, -A</I>, and <I>-t</I>
729 options that can be very useful for debugging and performance
732 <P>You can also compile with profiling to see what functions are
733 taking execution time. The backend profile files will be deposited
734 in the <I>pgsql/data/base/dbname</I> directory. The client profile
735 file will be put in the client's current directory. Linux requires
736 a compile with <I>-DLINUX_PROFILE</I> for proper profiling.</P>
738 <H4><A name="3.8">3.8</A>) Why do I get <I>"Sorry, too many
739 clients"</I> when trying to connect?</H4>
741 <P>You need to increase <I>postmaster</I>'s limit on how many
742 concurrent backend processes it can start.</P>
744 <P>The default limit is 32 processes. You can increase it by
745 restarting <I>postmaster</I> with a suitable <I>-N</I> value or
746 modifying <I>postgresql.conf</I>.</P>
748 <P>Note that if you make <I>-N</I> larger than 32, you must also
749 increase <I>-B</I> beyond its default of 64; <I>-B</I> must be at
750 least twice <I>-N</I>, and probably should be more than that for
751 best performance. For large numbers of backend processes, you are
752 also likely to find that you need to increase various Unix kernel
753 configuration parameters. Things to check include the maximum size
754 of shared memory blocks, <SMALL>SHMMAX;</SMALL> the maximum number
755 of semaphores, <SMALL>SEMMNS</SMALL> and <SMALL>SEMMNI;</SMALL> the
756 maximum number of processes, <SMALL>NPROC;</SMALL> the maximum
757 number of processes per user, <SMALL>MAXUPRC;</SMALL> and the
758 maximum number of open files, <SMALL>NFILE</SMALL> and
759 <SMALL>NINODE.</SMALL> The reason that PostgreSQL has a limit on
760 the number of allowed backend processes is so your system won't run
761 out of resources.</P>
763 <P>In PostgreSQL versions prior to 6.5, the maximum number of
764 backends was 64, and changing it required a rebuild after altering
765 the MaxBackendId constant in
766 <I>include/storage/sinvaladt.h</I>.</P>
768 <H4><A name="3.9">3.9</A>) What are the <I>pg_sorttempNNN.NN</I>
769 files in my database directory?</H4>
771 <P>They are temporary files generated by the query executor. For
772 example, if a sort needs to be done to satisfy an <SMALL>ORDER
773 BY,</SMALL> and the sort requires more space than the backend's
774 <I>-S</I> parameter allows, then temporary files are created to
775 hold the extra data.</P>
777 <P>The temporary files should be deleted automatically, but might
778 not if a backend crashes during a sort. If you have no backends
779 running at the time, it is safe to delete the pg_tempNNN.NN
782 <H4><A name="3.10">3.10</A>) Why do I need to do a dump and restore
783 to upgrade PostgreSQL?</H4>
785 <P>The PostgreSQL team tries very heard to maintain compatability across
786 minor releases. So upgrading from 7.2 to 7.2.1 does not require a dump a
787 restore. However, new features are continuously being adding and
788 sometimes this requires new fields to be added to system tables.
790 <P>These changes may be across many tables and so maintaining backward
791 compatability would be quite difficult. Thus, restoring from a dump is
792 required to make everything work.
794 <P>Note that the actual on-disk file format does not change very often,
795 a feature the pg_upgrade script uses quite successfully. There the dump
796 is used create the necessary information in the system tables. The data
797 files are then just copied across. This method is not as guarenteed as
798 the dump/restore method but when it works it can make upgrades very
803 <H2 align="center">Operational Questions</H2>
805 <H4><A name="4.1">4.1</A>) What is the difference between binary
806 cursors and normal cursors?</H4>
808 <P>See the <SMALL>DECLARE</SMALL> manual page for a
811 <H4><A name="4.2">4.2</A>) How do I <SMALL>SELECT</SMALL> only the
812 first few rows of a query?</H4>
814 <P>See the <SMALL>FETCH</SMALL> manual page, or use
815 <SMALL>SELECT</SMALL> ... <SMALL>LIMIT</SMALL>....</P>
817 <P>The entire query may have to be evaluated, even if you only want
818 the first few rows. Consider a query that has an <SMALL>ORDER
819 BY.</SMALL> If there is an index that matches the <SMALL>ORDER
820 BY</SMALL>, PostgreSQL may be able to evaluate only the first few
821 records requested, or the entire query may have to be evaluated
822 until the desired rows have been generated.</P>
824 <H4><A name="4.3">4.3</A>) How do I get a list of tables or other
825 things I can see in <I>psql</I>?</H4>
827 <P>You can read the source code for <I>psql</I> in file
828 <I>pgsql/src/bin/psql/describe.c</I>. It contains
829 <SMALL>SQL</SMALL> commands that generate the output for psql's
830 backslash commands. You can also start <I>psql</I> with the
831 <I>-E</I> option so it will print out the queries it uses to
832 execute the commands you give.</P>
834 <H4><A name="4.4">4.4</A>) How do you remove a column from a
837 <P>We do not support <SMALL>ALTER TABLE DROP COLUMN,</SMALL> but do
841 LOCK TABLE old_table;
842 SELECT ... -- select all columns but the one you want to remove
845 DROP TABLE old_table;
846 ALTER TABLE new_table RENAME TO old_table;
850 <H4><A name="4.5">4.5</A>) What is the maximum size for a row, a
851 table, and a database?</H4>
853 <P>These are the limits:</P>
855 Maximum size for a database? unlimited (500 GB databases exist)
856 Maximum size for a table? 16 TB
857 Maximum size for a row? unlimited in 7.1 and later
858 Maximum size for a field? 1 GB in 7.1 and later
859 Maximum number of rows in a table? unlimited
860 Maximum number of columns in a table? 250-1600 depending on column types
861 Maximum number of indexes on a table? unlimited
863 Of course, these are not actually unlimited, but limited to
864 available disk space and memory/swap space. Performance may suffer
865 when these values get unusually large.
867 <P>The maximum table size of 16 TB does not require large file
868 support from the operating system. Large tables are stored as
869 multiple 1 GB files so file system size limits are not
872 <P>The maximum table size and maximum number of columns can be
873 increased if the default block size is increased to 32k.</P>
875 <H4><A name="4.6">4.6</A>) How much database disk space is required
876 to store data from a typical text file?</H4>
878 <P>A PostgreSQL database may require up to five times the disk
879 space to store data from a text file.</P>
881 <P>As an example, consider a file of 100,000 lines with an integer
882 and text description on each line. Suppose the text string
883 avergages twenty bytes in length. The flat file would be 2.8 MB.
884 The size of the PostgreSQL database file containing this data can
885 be estimated as 6.4 MB:</P>
887 36 bytes: each row header (approximate)
888 24 bytes: one int field and one text filed
889 + 4 bytes: pointer on page to tuple
890 ----------------------------------------
893 The data page size in PostgreSQL is 8192 bytes (8 KB), so:
896 ------------------- = 128 rows per database page (rounded down)
900 -------------------- = 782 database pages (rounded up)
903 782 database pages * 8192 bytes per page = 6,406,144 bytes (6.4 MB)
906 <P>Indexes do not require as much overhead, but do contain the data
907 that is being indexed, so they can be large also.</P>
909 <H4><A name="4.7">4.7</A>) How do I find out what tables, indexes,
910 databases, and users are defined?</H4>
912 <P><I>psql</I> has a variety of backslash commands to show such
913 information. Use \? to see them. There are also system tables
914 beginning with <I>pg_</I> that describe these too. Also, <I>psql
915 -l</I> will list all databases.</P>
917 <P>Also try the file <I>pgsql/src/tutorial/syscat.source</I>. It
918 illustrates many of the <SMALL>SELECT</SMALL>s needed to get
919 information from the database system tables.</P>
921 <H4><A name="4.8">4.8</A>) My queries are slow or don't make use of
922 the indexes. Why?</H4>
923 Indexes are not automatically used by every query. Indexes are only
924 used if the table is larger than a minimum size, and the query
925 selects only a small percentage of the rows in the table. This is
926 because the random disk access caused by an index scan can be
927 slower than a straight read through the table, or sequential scan.
929 <P>To determine if an index should be used, PostgreSQL must have
930 statistics about the table. These statistics are collected using
931 <SMALL>VACUUM ANALYZE</SMALL>, or simply <SMALL>ANALYZE</SMALL>.
932 Using statistics, the optimizer knows how many rows are in the
933 table, and can better determine if indexes should be used.
934 Statistics are also valuable in determining optimal join order and
935 join methods. Statistics collection should be performed
936 periodically as the contents of the table change.</P>
938 <P>Indexes are normally not used for <SMALL>ORDER BY</SMALL> or to
939 perform joins. A sequential scan followed by an explicit sort is
940 usually faster than an index scan of a large table.</P>
941 However, <SMALL>LIMIT</SMALL> combined with <SMALL>ORDER BY</SMALL>
942 often will use an index because only a small portion of the table
943 is returned. In fact, though MAX() and MIN() don't use indexes,
944 it is possible to retrieve such values using an index with ORDER BY
949 ORDER BY col [ DESC ]
953 <P>When using wild-card operators such as <SMALL>LIKE</SMALL> or
954 <I>~</I>, indexes can only be used if the default C local is used
955 during initdb and the beginning of the search
956 is anchored to the start of the string. Therefore, to use indexes,
957 <SMALL>LIKE</SMALL> patterns must not start with <I>%</I>, and
958 <I>~</I>(regular expression) patterns must start with <I>^</I>.</P>
960 <H4><A name="4.9">4.9</A>) How do I see how the query optimizer is
961 evaluating my query?</H4>
963 <P>See the <SMALL>EXPLAIN</SMALL> manual page.</P>
965 <H4><A name="4.10">4.10</A>) What is an R-tree index?</H4>
967 <P>An R-tree index is used for indexing spatial data. A hash index
968 can't handle range searches. A B-tree index only handles range
969 searches in a single dimension. R-trees can handle
970 multi-dimensional data. For example, if an R-tree index can be
971 built on an attribute of type <I>point</I>, the system can more
972 efficiently answer queries such as "select all points within a
973 bounding rectangle."</P>
975 <P>The canonical paper that describes the original R-tree design
978 <P>Guttman, A. "R-trees: A Dynamic Index Structure for Spatial
979 Searching." Proceedings of the 1984 ACM SIGMOD Int'l Conf on Mgmt
982 <P>You can also find this paper in Stonebraker's "Readings in
983 Database Systems".</P>
985 <P>Built-in R-trees can handle polygons and boxes. In theory,
986 R-trees can be extended to handle higher number of dimensions. In
987 practice, extending R-trees requires a bit of work and we don't
988 currently have any documentation on how to do it.</P>
990 <H4><A name="4.11">4.11</A>) What is the Genetic Query
993 <P>The <SMALL>GEQO</SMALL> module speeds query optimization when
994 joining many tables by means of a Genetic Algorithm (GA). It allows
995 the handling of large join queries through nonexhaustive
998 <H4><A name="4.12">4.12</A>) How do I perform regular expression
999 searches and case-insensitive regular expression searches? How do I
1000 use an index for case-insensitive searches?</H4>
1002 <P>The <I>~</I> operator does regular expression matching, and
1003 <I>~*</I> does case-insensitive regular expression matching. The
1004 case-insensitive variant of <SMALL>LIKE</SMALL> is called
1005 <SMALL>ILIKE</SMALL> in PostgreSQL 7.1 and later.</P>
1007 <P>Case-insensitive equality comparisons are normally expressed
1012 WHERE lower(col) = 'abc'
1015 This will not use an standard index. However, if you create a
1016 functional index, it will be used:
1018 CREATE INDEX tabindex on tab (lower(col));
1022 <H4><A name="4.13">4.13</A>) In a query, how do I detect if a field
1023 is <SMALL>NULL</SMALL>?</H4>
1025 <P>You test the column with <SMALL>IS NULL</SMALL> and <SMALL>IS
1026 NOT NULL</SMALL>.</P>
1028 <H4><A name="4.14">4.14</A>) What is the difference between the
1029 various character types?</H4>
1031 Type Internal Name Notes
1032 --------------------------------------------------
1033 "char" char 1 character
1034 CHAR(#) bpchar blank padded to the specified fixed length
1035 VARCHAR(#) varchar size specifies maximum length, no padding
1036 TEXT text no specific upper limit on length
1037 BYTEA bytea variable-length byte array (null-byte safe)
1040 <P>You will see the internal name when examining system catalogs
1041 and in some error messages.</P>
1043 <P>The last four types above are "varlena" types (i.e., the first
1044 four bytes on disk are the length, followed by the data). Thus the
1045 actual space used is slightly greater than the declared size.
1046 However, these data types are also subject to compression or being
1047 stored out-of-line by <SMALL>TOAST</SMALL>, so the space on disk
1048 might also be less than expected.</P>
1050 <P><SMALL>CHAR()</SMALL> is best when storing strings that are
1051 usually the same length. <SMALL>VARCHAR()</SMALL> is best when
1052 storing variable-length strings but it limits how long a string can
1053 be. <SMALL>TEXT</SMALL> is for strings of unlimited length, maximum
1054 1 gigabyte. <SMALL>BYTEA</SMALL> is for storing binary data,
1055 particularly values that include <SMALL>NULL</SMALL> bytes.</P>
1057 <H4><A name="4.15.1">4.15.1</A>) How do I create a
1058 serial/auto-incrementing field?</H4>
1060 <P>PostgreSQL supports a <SMALL>SERIAL</SMALL> data type. It
1061 auto-creates a sequence and index on the column. For example,
1064 CREATE TABLE person (
1069 is automatically translated into this:
1071 CREATE SEQUENCE person_id_seq;
1072 CREATE TABLE person (
1073 id INT4 NOT NULL DEFAULT nextval('person_id_seq'),
1076 CREATE UNIQUE INDEX person_id_key ON person ( id );
1078 See the <I>create_sequence</I> manual page for more information
1079 about sequences. You can also use each row's <I>OID</I> field as a
1080 unique value. However, if you need to dump and reload the database,
1081 you need to use <I>pg_dump</I>'s <I>-o</I> option or <SMALL>COPY
1082 WITH OIDS</SMALL> option to preserve the <SMALL>OID</SMALL>s.
1084 <H4><A name="4.15.2">4.15.2</A>) How do I get the value of a
1085 <SMALL>SERIAL</SMALL> insert?</H4>
1087 <P>One approach is to retrieve the next <SMALL>SERIAL</SMALL> value
1088 from the sequence object with the <I>nextval()</I> function
1089 <I>before</I> inserting and then insert it explicitly. Using the
1090 example table in <A href="#4.15.1">4.15.1</A>, that might look like
1093 new_id = output of "SELECT nextval('person_id_seq')"
1094 INSERT INTO person (id, name) VALUES (new_id, 'Blaise Pascal');
1096 You would then also have the new value stored in
1097 <CODE>new_id</CODE> for use in other queries (e.g., as a foreign
1098 key to the <CODE>person</CODE> table). Note that the name of the
1099 automatically created <SMALL>SEQUENCE</SMALL> object will be named
1100 <<I>table</I>>_<<I>serialcolumn</I>>_<I>seq</I>, where
1101 <I>table</I> and <I>serialcolumn</I> are the names of your table
1102 and your <SMALL>SERIAL</SMALL> column, respectively.
1104 <P>Alternatively, you could retrieve the assigned
1105 <SMALL>SERIAL</SMALL> value with the <I>currval</I>() function
1106 <I>after</I> it was inserted by default, e.g.,</P>
1108 INSERT INTO person (name) VALUES ('Blaise Pascal');
1109 new_id = output of "SELECT currval('person_id_seq')";
1111 Finally, you could use the <A href="#4.16"><SMALL>OID</SMALL></A>
1112 returned from the <SMALL>INSERT</SMALL> statement to look up the
1113 default value, though this is probably the least portable approach.
1114 In Perl, using DBI with Edmund Mergl's DBD::Pg module, the oid
1115 value is made available via <I>$sth->{pg_oid_status}</I> after
1116 <I>$sth->execute()</I>.
1118 <H4><A name="4.15.3">4.15.3</A>) Don't <I>currval()</I> and
1119 <I>nextval()</I> lead to a race condition with other users?</H4>
1121 <P>No. Currval() returns the current value assigned by your
1122 backend, not by all users.</P>
1124 <H4><A name="4.15.4">4.15.4</A>) Why aren't my sequence numbers
1125 reused on transaction abort? Why are there gaps in the numbering of
1126 my sequence/SERIAL column?</H4>
1128 <P>To improve concurrency, sequence values are given out to running
1129 transactions as needed and are not locked until the transaction
1130 completes. This causes gaps in numbering from aborted
1133 <H4><A name="4.16">4.16</A>) What is an <SMALL>OID</SMALL>? What is
1134 a <SMALL>TID</SMALL>?</H4>
1136 <P><SMALL>OID</SMALL>s are PostgreSQL's answer to unique row ids.
1137 Every row that is created in PostgreSQL gets a unique
1138 <SMALL>OID</SMALL>. All <SMALL>OID</SMALL>s generated during
1139 <I>initdb</I> are less than 16384 (from
1140 <I>backend/access/transam.h</I>). All user-created
1141 <SMALL>OID</SMALL>s are equal to or greater than this. By default,
1142 all these <SMALL>OID</SMALL>s are unique not only within a table or
1143 database, but unique within the entire PostgreSQL installation.</P>
1145 <P>PostgreSQL uses <SMALL>OID</SMALL>s in its internal system
1146 tables to link rows between tables. These <SMALL>OID</SMALL>s can
1147 be used to identify specific user rows and used in joins. It is
1148 recommended you use column type <SMALL>OID</SMALL> to store
1149 <SMALL>OID</SMALL> values. You can create an index on the
1150 <SMALL>OID</SMALL> field for faster access.</P>
1152 <P>O<SMALL>ID</SMALL>s are assigned to all new rows from a central
1153 area that is used by all databases. If you want to change the
1154 <SMALL>OID</SMALL> to something else, or if you want to make a copy
1155 of the table, with the original <SMALL>OID</SMALL>s, there is no
1156 reason you can't do it:</P>
1158 CREATE TABLE new_table(old_oid oid, mycol int);
1159 SELECT old_oid, mycol INTO new FROM old;
1160 COPY new TO '/tmp/pgtable';
1162 COPY new WITH OIDS FROM '/tmp/pgtable';
1164 CREATE TABLE new_table (mycol int);
1165 INSERT INTO new_table (oid, mycol) SELECT oid, mycol FROM old_table;
1169 <P>O<SMALL>ID</SMALL>s are stored as 4-byte integers, and will
1170 overflow at 4 billion. No one has reported this ever happening, and
1171 we plan to have the limit removed before anyone does.</P>
1173 <P>T<SMALL>ID</SMALL>s are used to identify specific physical rows
1174 with block and offset values. T<SMALL>ID</SMALL>s change after rows
1175 are modified or reloaded. They are used by index entries to point
1176 to physical rows.</P>
1178 <H4><A name="4.17">4.17</A>) What is the meaning of some of the
1179 terms used in PostgreSQL?</H4>
1181 <P>Some of the source code and older documentation use terms that
1182 have more common usage. Here are some:</P>
1185 <LI>table, relation, class</LI>
1187 <LI>row, record, tuple</LI>
1189 <LI>column, field, attribute</LI>
1191 <LI>retrieve, select</LI>
1193 <LI>replace, update</LI>
1195 <LI>append, insert</LI>
1197 <LI><SMALL>OID</SMALL>, serial value</LI>
1199 <LI>portal, cursor</LI>
1201 <LI>range variable, table name, table alias</LI>
1204 <P>A list of general database terms can be found at: <A href=
1205 "http://www.comptechnews.com/~reaster/dbdesign.html">http://www.comptechnews.com/~reaster/dbdesign.html</A></P>
1207 <H4><A name="4.18">4.18</A>) Why do I get the error <I>"ERROR:
1208 Memory exhausted in AllocSetAlloc()"</I>?</H4>
1210 <P>If you are running a version older than 7.1, an upgrade may fix
1211 the problem. Also it is possible you have run out of virtual memory
1212 on your system, or your kernel has a low limit for certain
1213 resources. Try this before starting <I>postmaster</I>:</P>
1218 Depending on your shell, only one of these may succeed, but it will
1219 set your process data segment limit much higher and perhaps allow
1220 the query to complete. This command applies to the current process,
1221 and all subprocesses created after the command is run. If you are
1222 having a problem with the <SMALL>SQL</SMALL> client because the
1223 backend is returning too much data, try it before starting the
1226 <H4><A name="4.19">4.19</A>) How do I tell what PostgreSQL version
1229 <P>From <I>psql</I>, type <CODE>select version();</CODE></P>
1231 <H4><A name="4.20">4.20</A>) Why does my large-object operations
1232 get <I>"invalid large obj descriptor"</I>?</H4>
1234 <P>You need to put <CODE>BEGIN WORK</CODE> and <CODE>COMMIT</CODE>
1235 around any use of a large object handle, that is, surrounding
1236 <CODE>lo_open</CODE> ... <CODE>lo_close.</CODE></P>
1238 <P>Currently PostgreSQL enforces the rule by closing large object
1239 handles at transaction commit. So the first attempt to do anything
1240 with the handle will draw <I>invalid large obj descriptor</I>. So
1241 code that used to work (at least most of the time) will now
1242 generate that error message if you fail to use a transaction.</P>
1244 <P>If you are using a client interface like <SMALL>ODBC</SMALL> you
1245 may need to set <CODE>auto-commit off.</CODE></P>
1247 <H4><A name="4.21">4.21</A>) How do I create a column that will
1248 default to the current time?</H4>
1250 <P>Use <I>CURRENT_TIMESTAMP</I>:</P>
1252 <CODE>CREATE TABLE test (x int, modtime timestamp DEFAULT CURRENT_TIMESTAMP );
1256 <H4><A name="4.22">4.22</A>) Why are my subqueries using
1257 <CODE><SMALL>IN</SMALL></CODE> so slow?</H4>
1259 <P>Currently, we join subqueries to outer queries by sequentially
1260 scanning the result of the subquery for each row of the outer
1261 query. A workaround is to replace <CODE>IN</CODE> with
1262 <CODE>EXISTS</CODE>:</P>
1266 WHERE col1 IN (SELECT col2 FROM TAB2)
1273 WHERE EXISTS (SELECT col2 FROM TAB2 WHERE col1 = col2)
1276 We hope to fix this limitation in a future release.
1278 <H4><A name="4.23">4.23</A>) How do I perform an outer join?</H4>
1280 <P>PostgreSQL 7.1 and later supports outer joins using the SQL
1281 standard syntax. Here are two examples:</P>
1284 FROM t1 LEFT OUTER JOIN t2 ON (t1.col = t2.col);
1289 FROM t1 LEFT OUTER JOIN t2 USING (col);
1292 <P>These identical queries join t1.col to t2.col, and also return
1293 any unjoined rows in t1 (those with no match in t2). A
1294 <SMALL>RIGHT</SMALL> join would add unjoined rows of t2. A
1295 <SMALL>FULL</SMALL> join would return the matched rows plus all
1296 unjoined rows from t1 and t2. The word <SMALL>OUTER</SMALL> is
1297 optional and is assumed in <SMALL>LEFT</SMALL>,
1298 <SMALL>RIGHT</SMALL>, and <SMALL>FULL</SMALL> joins. Ordinary joins
1299 are called <SMALL>INNER</SMALL> joins.</P>
1301 <P>In previous releases, outer joins can be simulated using
1302 <SMALL>UNION</SMALL> and <SMALL>NOT IN</SMALL>. For example, when
1303 joining <I>tab1</I> and <I>tab2</I>, the following query does an
1304 <I>outer</I> join of the two tables:<BR>
1308 SELECT tab1.col1, tab2.col2
1310 WHERE tab1.col1 = tab2.col1
1312 SELECT tab1.col1, NULL
1314 WHERE tab1.col1 NOT IN (SELECT tab2.col1 FROM tab2)
1318 <H4><A name="4.24">4.24</A>) How do I perform queries using
1319 multiple databases?</H4>
1321 <P>There is no way to query any database except the current one.
1322 Because PostgreSQL loads database-specific system catalogs, it is
1323 uncertain how a cross-database query should even behave.</P>
1325 <P>Of course, a client can make simultaneous connections to
1326 different databases and merge the information that way.</P>
1328 <H4><A name="4.25">4.25</A>) How do I return multiple rows or
1329 columns from a function?</H4>
1331 <P>You can return result sets from PL/pgSQL functions using
1332 <I>refcursors</I>. See <A href=
1333 "http://developer.postgresql.org/docs/postgres/plpgsql-cursors.html">
1334 http://developer.postgresql.org/docs/postgres/plpgsql-cursors.html,</A>
1335 section 23.7.3.3.</P>
1337 <H4><A name="4.26">4.26</A>) Why can't I reliably create/drop
1338 temporary tables in PL/PgSQL functions?</H4>
1339 PL/PgSQL caches function contents, and an unfortunate side effect
1340 is that if a PL/PgSQL function accesses a temporary table, and that
1341 table is later dropped and recreated, and the function called
1342 again, the function will fail because the cached function contents
1343 still point to the old temporary table. The solution is to use
1344 <SMALL>EXECUTE</SMALL> for temporary table access in PL/PgSQL. This
1345 will cause the query to be reparsed every time.
1349 <H2 align="center">Extending PostgreSQL</H2>
1351 <H4><A name="5.1">5.1</A>) I wrote a user-defined function. When I
1352 run it in <I>psql</I>, why does it dump core?</H4>
1354 <P>The problem could be a number of things. Try testing your
1355 user-defined function in a stand-alone test program first.</P>
1357 <H4><A name="5.2">5.2</A>) How can I contribute some nifty new
1358 types and functions to PostgreSQL?</H4>
1360 <P>Send your extensions to the <I>pgsql-hackers</I> mailing list,
1361 and they will eventually end up in the <I>contrib/</I>
1364 <H4><A name="5.3">5.3</A>) How do I write a C function to return a
1367 <P>This requires wizardry so extreme that the authors have never
1368 tried it, though in principle it can be done.</P>
1370 <H4><A name="5.4">5.4</A>) I have changed a source file. Why does
1371 the recompile not see the change?</H4>
1373 <P>The <I>Makefiles</I> do not have the proper dependencies for
1374 include files. You have to do a <I>make clean</I> and then another
1375 <I>make</I>. If you are using <SMALL>GCC</SMALL> you can use the
1376 <I>--enable-depend</I> option of <I>configure</I> to have the
1377 compiler compute the dependencies automatically.</P>