Debian Codegraph @

doc/src/sgml/glossary.sgml

raw
1 <appendix id="glossary">
2 <title>Glossary</title>
3 <para>
4 This is a list of terms and their meaning in the context of
5 <productname>PostgreSQL</productname> and relational database
6 systems in general.
7 </para>
8
9 <glosslist>
10 <glossentry id="glossary-acid">
11 <glossterm>ACID</glossterm>
12 <glossdef>
13 <para>
14 <glossterm linkend="glossary-atomicity">Atomicity</glossterm>,
15 <glossterm linkend="glossary-consistency">Consistency</glossterm>,
16 <glossterm linkend="glossary-isolation">Isolation</glossterm>, and
17 <glossterm linkend="glossary-durability">Durability</glossterm>.
18 This set of properties of database transactions is intended to
19 guarantee validity in concurrent operation and even in event of
20 errors, power failures, etc.
21 </para>
22 </glossdef>
23 </glossentry>
24
25 <glossentry id="glossary-aggregate">
26 <glossterm>Aggregate function (routine)</glossterm>
27 <glossdef>
28 <para>
29 A <glossterm linkend="glossary-function">function</glossterm> that
30 combines (<firstterm>aggregates</firstterm>) multiple input values,
31 for example by counting, averaging or adding,
32 yielding a single output value.
33 </para>
34 <para>
35 For more information, see
36 <xref linkend="functions-aggregate"/>.
37 </para>
38 <glossseealso otherterm="glossary-window-function" />
39 </glossdef>
40 </glossentry>
41
42 <glossentry id="glossary-am">
43 <glossterm>Access Method</glossterm>
44 <glossdef>
45 <para>
46 Interfaces which <productname>PostgreSQL</productname> use in order to
47 access data in tables and indexes. This abstraction allows for adding
48 support for new types of data storage.
49 </para>
50 <para>
51 For more information, see <xref linkend="tableam" /> and
52 <xref linkend="indexam" />.
53 </para>
54 </glossdef>
55 </glossentry>
56
57 <glossentry>
58 <glossterm>Analytic function</glossterm>
59 <glosssee otherterm="glossary-window-function" />
60 </glossentry>
61
62 <glossentry id="glossary-analyze">
63 <glossterm>Analyze (operation)</glossterm>
64 <glossdef>
65 <para>
66 The act of collecting statistics from data in
67 <glossterm linkend="glossary-table">tables</glossterm>
68 and other <glossterm linkend="glossary-relation">relations</glossterm>
69 to help the <glossterm linkend="glossary-planner">query planner</glossterm>
70 to make decisions about how to execute
71 <glossterm linkend="glossary-query">queries</glossterm>.
72 </para>
73 <para>
74 (Don't confuse this term with the <literal>ANALYZE</literal> option
75 to the <xref linkend="sql-explain"/> command.)
76 </para>
77 <para>
78 For more information, see
79 <xref linkend="sql-analyze"/>.
80 </para>
81 </glossdef>
82 </glossentry>
83
84 <glossentry id="glossary-aio">
85 <glossterm>Asynchronous <acronym>I/O</acronym></glossterm>
86 <acronym>AIO</acronym>
87 <indexterm>
88 <primary>Asynchronous <acronym>I/O</acronym></primary>
89 </indexterm>
90 <glossdef>
91 <para>
92 Asynchronous <acronym>I/O</acronym> (<acronym>AIO</acronym>) describes
93 performing <acronym>I/O</acronym> in a non-blocking way (asynchronously),
94 in contrast to synchronous <acronym>I/O</acronym>, which blocks for the
95 entire duration of the <acronym>I/O</acronym>.
96 </para>
97 <para>
98 With <acronym>AIO</acronym>, starting an <acronym>I/O</acronym> operation
99 is separated from waiting for the result of the operation, allowing
100 multiple <acronym>I/O</acronym> operations to be initiated concurrently,
101 as well as performing <acronym>CPU</acronym> heavy operations
102 concurrently with <acronym>I/O</acronym>. The price for that increased
103 concurrency is increased complexity.
104 </para>
105 <glossseealso otherterm="glossary-io" />
106 </glossdef>
107 </glossentry>
108
109 <glossentry id="glossary-atomic">
110 <glossterm>Atomic</glossterm>
111 <glossdef>
112 <para>
113 In reference to a <glossterm linkend="glossary-datum">datum</glossterm>:
114 the fact that its value cannot be broken down into smaller
115 components.
116 </para>
117 </glossdef>
118 <glossdef>
119 <para>
120 In reference to a
121 <glossterm linkend="glossary-transaction">database transaction</glossterm>:
122 see <glossterm linkend="glossary-atomicity">atomicity</glossterm>.
123 </para>
124 </glossdef>
125 </glossentry>
126
127 <glossentry id="glossary-atomicity">
128 <glossterm>Atomicity</glossterm>
129 <glossdef>
130 <para>
131 The property of a <glossterm linkend="glossary-transaction">transaction</glossterm>
132 that either all its operations complete as a single unit or none do.
133 In addition, if a system failure occurs during the execution of a
134 transaction, no partial results are visible after recovery.
135 This is one of the <acronym>ACID</acronym> properties.
136 </para>
137 </glossdef>
138 </glossentry>
139
140 <glossentry id="glossary-attribute">
141 <glossterm>Attribute</glossterm>
142 <glossdef>
143 <para>
144 An element with a certain name and data type found within a
145 <glossterm linkend="glossary-tuple">tuple</glossterm>.
146 </para>
147 </glossdef>
148 </glossentry>
149
150 <glossentry id="glossary-autovacuum">
151 <glossterm>Autovacuum (process)</glossterm>
152 <glossdef>
153 <para>
154 A set of background processes that routinely perform
155 <glossterm linkend="glossary-vacuum">vacuum</glossterm>
156 and <glossterm linkend="glossary-analyze">analyze</glossterm> operations.
157 The <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
158 that coordinates the work and is always present (unless autovacuum
159 is disabled) is known as the <firstterm>autovacuum launcher</firstterm>,
160 and the processes that carry out the tasks are known as the
161 <firstterm>autovacuum workers</firstterm>.
162 </para>
163 <para>
164 For more information, see
165 <xref linkend="autovacuum"/>.
166 </para>
167 </glossdef>
168 </glossentry>
169
170 <glossentry id="glossary-auxiliary-proc">
171 <glossterm>Auxiliary process</glossterm>
172 <glossdef>
173 <para>
174 A process within an <glossterm linkend="glossary-instance">instance</glossterm>
175 that is in charge of some specific background task for the instance.
176 The auxiliary processes consist of <!-- in alphabetical order -->
177 <!-- NB: In the code, the autovac launcher doesn't use the auxiliary
178 process scaffolding; however it does behave as one so we list it
179 here anyway. In addition, logger isn't connected to shared memory so
180 most code outside postmaster.c doesn't even consider it a "proc" in
181 the first place.
182 -->
183 the <glossterm linkend="glossary-autovacuum">autovacuum launcher</glossterm>
184 (but not the autovacuum workers),
185 the <glossterm linkend="glossary-background-writer">background writer</glossterm>,
186 the <glossterm linkend="glossary-checkpointer">checkpointer</glossterm>,
187 the <glossterm linkend="glossary-logger">logger</glossterm>,
188 the <glossterm linkend="glossary-startup-process">startup process</glossterm>,
189 the <glossterm linkend="glossary-wal-archiver">WAL archiver</glossterm>,
190 the <glossterm linkend="glossary-wal-receiver">WAL receiver</glossterm>
191 (but not the <glossterm linkend="glossary-wal-sender">WAL senders</glossterm>),
192 the <glossterm linkend="glossary-wal-summarizer">WAL summarizer</glossterm>,
193 and the <glossterm linkend="glossary-wal-writer">WAL writer</glossterm>.
194 </para>
195 </glossdef>
196 </glossentry>
197
198 <glossentry id="glossary-backend">
199 <glossterm>Backend (process)</glossterm>
200 <glossdef>
201 <para>
202 Process of an <glossterm linkend="glossary-instance">instance</glossterm>
203 which acts on behalf of a <glossterm linkend="glossary-session">client session</glossterm>
204 and handles its requests.
205 </para>
206 <para>
207 (Don't confuse this term with the similar terms
208 <glossterm linkend="glossary-background-worker">Background Worker</glossterm> or
209 <glossterm linkend="glossary-background-writer">Background Writer</glossterm>).
210 </para>
211 </glossdef>
212 </glossentry>
213
214 <glossentry id="glossary-background-worker">
215 <glossterm>Background worker (process)</glossterm>
216 <glossdef>
217 <para>
218 Process within an <glossterm linkend="glossary-instance">instance</glossterm>,
219 which runs system- or user-supplied code.
220 Serves as infrastructure for several features in
221 <productname>PostgreSQL</productname>, such as
222 <glossterm linkend="glossary-replication">logical replication</glossterm>
223 and <glossterm linkend="glossary-parallel-query">parallel queries</glossterm>.
224 In addition, <glossterm linkend="glossary-extension">Extensions</glossterm> can add
225 custom background worker processes.
226 </para>
227 <para>
228 For more information, see
229 <xref linkend="bgworker"/>.
230 </para>
231 </glossdef>
232 </glossentry>
233
234 <glossentry id="glossary-background-writer">
235 <glossterm>Background writer (process)</glossterm>
236 <glossdef>
237 <para>
238 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
239 that writes dirty
240 <glossterm linkend="glossary-data-page">data pages</glossterm> from
241 <glossterm linkend="glossary-shared-memory">shared memory</glossterm> to
242 the file system. It wakes up periodically, but works only for a short
243 period in order to distribute its expensive <acronym>I/O</acronym>
244 activity over time to avoid generating larger
245 <acronym>I/O</acronym> peaks which could block other processes.
246 </para>
247 <para>
248 For more information, see
249 <xref linkend="runtime-config-resource-background-writer"/>.
250 </para>
251 </glossdef>
252 </glossentry>
253
254 <glossentry id="glossary-basebackup">
255 <glossterm>Base Backup</glossterm>
256 <glossdef>
257 <para>
258 A binary copy of all
259 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>
260 files. It is generated by the tool <xref linkend="app-pgbasebackup"/>.
261 In combination with WAL files it can be used as the starting point
262 for recovery, log shipping, or streaming replication.
263 </para>
264 </glossdef>
265 </glossentry>
266
267 <glossentry id="glossary-bloat">
268 <glossterm>Bloat</glossterm>
269 <glossdef>
270 <para>
271 Space in data pages which does not contain current row versions,
272 such as unused (free) space or outdated row versions.
273 </para>
274 </glossdef>
275 </glossentry>
276
277 <glossentry id="glossary-bootstrap-superuser">
278 <glossterm>Bootstrap superuser</glossterm>
279 <glossdef>
280 <para>
281 The first <glossterm linkend="glossary-user">user</glossterm> initialized in a
282 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>.
283 </para>
284 <para>
285 This user owns all system catalog tables in each database. It is also the role
286 from which all granted permissions originate. Because of these things, this
287 role may not be dropped.
288 </para>
289 <para>
290 This role also behaves as a normal
291 <glossterm linkend="glossary-database-superuser">database superuser</glossterm>,
292 and its superuser status cannot be removed.
293 </para>
294 </glossdef>
295 </glossentry>
296
297 <glossentry id="glossary-buffer-access-strategy">
298 <glossterm>Buffer Access Strategy</glossterm>
299 <glossdef>
300 <para>
301 Some operations will access a large number of
302 <glossterm linkend="glossary-data-page">pages</glossterm>. A
303 <firstterm>Buffer Access Strategy</firstterm> helps to prevent these
304 operations from evicting too many pages from
305 <glossterm linkend="glossary-shared-memory">shared buffers</glossterm>.
306 </para>
307 <para>
308 A Buffer Access Strategy sets up references to a limited number of
309 <glossterm linkend="glossary-shared-memory">shared buffers</glossterm> and
310 reuses them circularly. When the operation requires a new page, a victim
311 buffer is chosen from the buffers in the strategy ring, which may require
312 flushing the page's dirty data and possibly also unflushed
313 <glossterm linkend="glossary-wal">WAL</glossterm> to permanent storage.
314 </para>
315 <para>
316 Buffer Access Strategies are used for various operations such as
317 sequential scans of large tables, <command>VACUUM</command>,
318 <command>COPY</command>, <command>CREATE TABLE AS SELECT</command>,
319 <command>ALTER TABLE</command>, <command>CREATE DATABASE</command>,
320 <command>CREATE INDEX</command>, and <command>CLUSTER</command>.
321 </para>
322 </glossdef>
323 </glossentry>
324
325 <glossentry id="glossary-cast">
326 <glossterm>Cast</glossterm>
327 <glossdef>
328 <para>
329 A conversion of a <glossterm linkend="glossary-datum">datum</glossterm>
330 from its current data type to another data type.
331 </para>
332 <para>
333 For more information, see
334 <xref linkend="sql-createcast"/>.
335 </para>
336 </glossdef>
337 </glossentry>
338
339 <glossentry id="glossary-catalog">
340 <glossterm>Catalog</glossterm>
341 <glossdef>
342 <para>
343 The <acronym>SQL</acronym> standard uses this term to
344 indicate what is called a
345 <glossterm linkend="glossary-database">database</glossterm> in
346 <productname>PostgreSQL</productname>'s terminology.
347 </para>
348 <para>
349 (Don't confuse this term with
350 <glossterm linkend="glossary-system-catalog">system catalog</glossterm>).
351 </para>
352 <para>
353 For more information, see
354 <xref linkend="manage-ag-overview"/>.
355 </para>
356 </glossdef>
357 </glossentry>
358
359 <glossentry id="glossary-check-constraint">
360 <glossterm>Check constraint</glossterm>
361 <glossdef>
362 <para>
363 A type of <glossterm linkend="glossary-constraint">constraint</glossterm>
364 defined on a <glossterm linkend="glossary-relation">relation</glossterm>
365 which restricts the values allowed in one or more
366 <glossterm linkend="glossary-attribute">attributes</glossterm>. The
367 check constraint can make reference to any attribute of the same row in
368 the relation, but cannot reference other rows of the same relation or
369 other relations.
370 </para>
371 <para>
372 For more information, see
373 <xref linkend="ddl-constraints"/>.
374 </para>
375 </glossdef>
376 </glossentry>
377
378 <glossentry id="glossary-checkpoint">
379 <glossterm>Checkpoint</glossterm>
380 <glossdef>
381 <para>
382 A point in the <glossterm linkend="glossary-wal">WAL</glossterm> sequence
383 at which it is guaranteed that the heap and index data files have been
384 updated with all information from
385 <glossterm linkend="glossary-shared-memory">shared memory</glossterm>
386 modified before that checkpoint;
387 a <firstterm>checkpoint record</firstterm> is written and flushed to WAL
388 to mark that point.
389 </para>
390 <para>
391 A checkpoint is also the act of carrying out all the actions that
392 are necessary to reach a checkpoint as defined above.
393 This process is initiated when predefined conditions are met,
394 such as a specified amount of time has passed, or a certain volume
395 of records has been written; or it can be invoked by the user
396 with the command <command>CHECKPOINT</command>.
397 </para>
398 <para>
399 For more information, see
400 <xref linkend="wal-configuration"/>.
401 </para>
402 </glossdef>
403 </glossentry>
404
405 <glossentry id="glossary-checkpointer">
406 <glossterm>Checkpointer (process)</glossterm>
407 <glossdef>
408 <para>
409 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
410 that is responsible for executing
411 <glossterm linkend="glossary-checkpoint">checkpoints</glossterm>.
412 </para>
413 </glossdef>
414 </glossentry>
415
416 <glossentry>
417 <glossterm>Class (archaic)</glossterm>
418 <glosssee otherterm="glossary-relation" />
419 </glossentry>
420
421 <glossentry id="glossary-client">
422 <glossterm>Client (process)</glossterm>
423 <glossdef>
424 <para>
425 Any process, possibly remote, that establishes a
426 <glossterm linkend="glossary-session">session</glossterm>
427 by <glossterm linkend="glossary-connection">connecting</glossterm> to an
428 <glossterm linkend="glossary-instance">instance</glossterm>
429 to interact with a <glossterm linkend="glossary-database">database</glossterm>.
430 </para>
431 </glossdef>
432 </glossentry>
433
434 <glossentry id="glossary-cluster-owner">
435 <glossterm>Cluster owner</glossterm>
436 <glossdef>
437 <para>
438 The operating system user that owns the
439 <glossterm linkend="glossary-data-directory">data directory</glossterm>
440 and under which the <literal>postgres</literal> process is run.
441 It is required that this user exist prior to creating a new
442 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>.
443 </para>
444 <para>
445 On operating systems with a <literal>root</literal> user,
446 said user is not allowed to be the cluster owner.
447 </para>
448 </glossdef>
449 </glossentry>
450
451 <glossentry id="glossary-column">
452 <glossterm>Column</glossterm>
453 <glossdef>
454 <para>
455 An <glossterm linkend="glossary-attribute">attribute</glossterm> found in
456 a <glossterm linkend="glossary-table">table</glossterm> or
457 <glossterm linkend="glossary-view">view</glossterm>.
458 </para>
459 </glossdef>
460 </glossentry>
461
462 <glossentry id="glossary-commit">
463 <glossterm>Commit</glossterm>
464 <glossdef>
465 <para>
466 The act of finalizing a
467 <glossterm linkend="glossary-transaction">transaction</glossterm> within
468 the <glossterm linkend="glossary-database">database</glossterm>, which
469 makes it visible to other transactions and assures its
470 <glossterm linkend="glossary-durability">durability</glossterm>.
471 </para>
472 <para>
473 For more information, see
474 <xref linkend="sql-commit"/>.
475 </para>
476 </glossdef>
477 </glossentry>
478
479 <glossentry id="glossary-concurrency">
480 <glossterm>Concurrency</glossterm>
481 <glossdef>
482 <para>
483 The concept that multiple independent operations happen within the
484 <glossterm linkend="glossary-database">database</glossterm> at the same time.
485 In <productname>PostgreSQL</productname>, concurrency is controlled by
486 the <glossterm linkend="glossary-mvcc">multiversion concurrency control</glossterm>
487 mechanism.
488 </para>
489 </glossdef>
490 </glossentry>
491
492 <glossentry id="glossary-connection">
493 <glossterm>Connection</glossterm>
494 <glossdef>
495 <para>
496 An established line of communication between a client process and a
497 <glossterm linkend="glossary-backend">backend</glossterm> process,
498 usually over a network, supporting a
499 <glossterm linkend="glossary-session">session</glossterm>. This term is
500 sometimes used as a synonym for session.
501 </para>
502 <para>
503 For more information, see
504 <xref linkend="runtime-config-connection"/>.
505 </para>
506 </glossdef>
507 </glossentry>
508
509 <glossentry id="glossary-consistency">
510 <glossterm>Consistency</glossterm>
511 <glossdef>
512 <para>
513 The property that the data in the
514 <glossterm linkend="glossary-database">database</glossterm>
515 is always in compliance with
516 <glossterm linkend="glossary-constraint">integrity constraints</glossterm>.
517 Transactions may be allowed to violate some of the constraints
518 transiently before it commits, but if such violations are not resolved
519 by the time it commits, such a transaction is automatically
520 <glossterm linkend="glossary-rollback">rolled back</glossterm>.
521 This is one of the <acronym>ACID</acronym> properties.
522 </para>
523 </glossdef>
524 </glossentry>
525
526 <glossentry id="glossary-constraint">
527 <glossterm>Constraint</glossterm>
528 <glossdef>
529 <para>
530 A restriction on the values of data allowed within a
531 <glossterm linkend="glossary-table">table</glossterm>,
532 or in attributes of a
533 <glossterm linkend="glossary-domain">domain</glossterm>.
534 </para>
535 <para>
536 For more information, see
537 <xref linkend="ddl-constraints"/>.
538 </para>
539 </glossdef>
540 </glossentry>
541
542 <glossentry id="glossary-cumulative-statistics">
543 <glossterm>Cumulative Statistics System</glossterm>
544 <glossdef>
545 <para>
546 A system which, if enabled, accumulates statistical information
547 about the <glossterm linkend="glossary-instance">instance</glossterm>'s
548 activities.
549 </para>
550 <para>
551 For more information, see
552 <xref linkend="monitoring-stats"/>.
553 </para>
554 </glossdef>
555 </glossentry>
556
557 <glossentry>
558 <glossterm>Data area</glossterm>
559 <glosssee otherterm="glossary-data-directory" />
560 </glossentry>
561
562 <glossentry id="glossary-database">
563 <glossterm>Database</glossterm>
564 <glossdef>
565 <para>
566 A named collection of
567 <glossterm linkend="glossary-sql-object">local SQL objects</glossterm>.
568 </para>
569 <para>
570 For more information, see
571 <xref linkend="manage-ag-overview"/>.
572 </para>
573 </glossdef>
574 </glossentry>
575
576 <glossentry id="glossary-db-cluster">
577 <glossterm>Database cluster</glossterm>
578 <glossdef>
579 <para>
580 A collection of databases and global SQL objects,
581 and their common static and dynamic metadata.
582 Sometimes referred to as a
583 <firstterm>cluster</firstterm>.
584 A database cluster is created using the
585 <xref linkend="app-initdb" /> program.
586 </para>
587 <para>
588 In <productname>PostgreSQL</productname>, the term
589 <firstterm>cluster</firstterm> is also sometimes used to refer to an instance.
590 (Don't confuse this term with the SQL command <command>CLUSTER</command>.)
591 </para>
592 <para>
593 See also <glossterm linkend="glossary-cluster-owner">cluster owner</glossterm>,
594 the operating-system owner of a cluster,
595 and <glossterm linkend="glossary-bootstrap-superuser">bootstrap superuser</glossterm>,
596 the <productname>PostgreSQL</productname> owner of a cluster.
597 </para>
598 </glossdef>
599 </glossentry>
600
601 <glossentry>
602 <glossterm>Database server</glossterm>
603 <glosssee otherterm="glossary-instance" />
604 </glossentry>
605
606 <glossentry id="glossary-database-superuser">
607 <glossterm>Database superuser</glossterm>
608 <glossdef>
609 <para>
610 A role having <firstterm>superuser status</firstterm>
611 (see <xref linkend="role-attributes"/>).
612 </para>
613 <para>
614 Frequently referred to as <firstterm>superuser</firstterm>.
615 </para>
616 </glossdef>
617 </glossentry>
618
619 <glossentry id="glossary-data-directory">
620 <glossterm>Data directory</glossterm>
621 <glossdef>
622 <para>
623 The base directory on the file system of a
624 <glossterm linkend="glossary-server">server</glossterm> that contains all
625 data files and subdirectories associated with a
626 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>
627 (with the exception of
628 <glossterm linkend="glossary-tablespace">tablespaces</glossterm>,
629 and optionally <glossterm linkend="glossary-wal">WAL</glossterm>).
630 The environment variable <literal>PGDATA</literal> is commonly used to
631 refer to the data directory.
632 </para>
633 <para>
634 A <glossterm linkend="glossary-db-cluster">cluster</glossterm>'s storage
635 space comprises the data directory plus any additional tablespaces.
636 </para>
637 <para>
638 For more information, see
639 <xref linkend="storage-file-layout"/>.
640 </para>
641 </glossdef>
642 </glossentry>
643
644 <glossentry id="glossary-data-page">
645 <glossterm>Data page</glossterm>
646 <glossdef>
647 <para>
648 The basic structure used to store relation data.
649 All pages are of the same size.
650 Data pages are typically stored on disk, each in a specific file,
651 and can be read to <glossterm linkend="glossary-shared-memory">shared buffers</glossterm>
652 where they can be modified, becoming
653 <firstterm>dirty</firstterm>. They become clean when written
654 to disk. New pages, which initially exist in memory only, are also
655 dirty until written.
656 </para>
657 </glossdef>
658 </glossentry>
659
660 <glossentry id="glossary-datum">
661 <glossterm>Datum</glossterm>
662 <glossdef>
663 <para>
664 The internal representation of one value of an <acronym>SQL</acronym>
665 data type.
666 </para>
667 </glossdef>
668 </glossentry>
669
670 <glossentry id="glossary-delete">
671 <glossterm>Delete</glossterm>
672 <glossdef>
673 <para>
674 An <acronym>SQL</acronym> command which removes
675 <glossterm linkend="glossary-tuple">rows</glossterm> from a given
676 <glossterm linkend="glossary-table">table</glossterm>
677 or <glossterm linkend="glossary-relation">relation</glossterm>.
678 </para>
679 <para>
680 For more information, see
681 <xref linkend="sql-delete"/>.
682 </para>
683 </glossdef>
684 </glossentry>
685
686 <glossentry id="glossary-domain">
687 <glossterm>Domain</glossterm>
688 <glossdef>
689 <para>
690 A user-defined data type that is based on another underlying data type.
691 It acts the same as the underlying type except for possibly restricting
692 the set of allowed values.
693 </para>
694 <para>
695 For more information, see <xref linkend="domains"/>.
696 </para>
697 </glossdef>
698 </glossentry>
699
700 <glossentry id="glossary-durability">
701 <glossterm>Durability</glossterm>
702 <glossdef>
703 <para>
704 The assurance that once a
705 <glossterm linkend="glossary-transaction">transaction</glossterm> has
706 been <glossterm linkend="glossary-commit">committed</glossterm>, the
707 changes remain even after a system failure or crash.
708 This is one of the <acronym>ACID</acronym> properties.
709 </para>
710 </glossdef>
711 </glossentry>
712
713 <glossentry>
714 <glossterm>Epoch</glossterm>
715 <glosssee otherterm="glossary-xid" />
716 </glossentry>
717
718 <glossentry id="glossary-extension">
719 <glossterm>Extension</glossterm>
720 <glossdef>
721 <para>
722 A software add-on package that can be installed on an
723 <glossterm linkend="glossary-instance">instance</glossterm> to
724 get extra features.
725 </para>
726 <para>
727 For more information, see
728 <xref linkend="extend-extensions" />.
729 </para>
730 </glossdef>
731 </glossentry>
732
733 <glossentry id="glossary-file-segment">
734 <glossterm>File segment</glossterm>
735 <glossdef>
736 <para>
737 A physical file which stores data for a given
738 <glossterm linkend="glossary-relation">relation</glossterm>.
739 File segments are limited in size by a configuration value
740 (typically 1 gigabyte),
741 so if a relation exceeds that size, it is split into multiple segments.
742 </para>
743 <para>
744 For more information, see
745 <xref linkend="storage-file-layout"/>.
746 </para>
747 <para>
748 (Don't confuse this term with the similar term
749 <glossterm linkend="glossary-wal-file">WAL segment</glossterm>).
750 </para>
751 </glossdef>
752 </glossentry>
753
754 <glossentry id="glossary-foreign-data-wrapper">
755 <glossterm>Foreign data wrapper</glossterm>
756 <glossdef>
757 <para>
758 A means of representing data that is not contained in the local
759 <glossterm linkend="glossary-database">database</glossterm> so that it appears as if were in local
760 <glossterm linkend="glossary-table">table(s)</glossterm>. With a foreign data wrapper it is
761 possible to define a <glossterm linkend="glossary-foreign-server">foreign server</glossterm> and
762 <glossterm linkend="glossary-foreign-table">foreign tables</glossterm>.
763 </para>
764 <para>
765 For more information, see
766 <xref linkend="sql-createforeigndatawrapper"/>.
767 </para>
768 </glossdef>
769 </glossentry>
770
771 <glossentry id="glossary-foreign-key">
772 <glossterm>Foreign key</glossterm>
773 <glossdef>
774 <para>
775 A type of <glossterm linkend="glossary-constraint">constraint</glossterm>
776 defined on one or more <glossterm linkend="glossary-column">columns</glossterm>
777 in a <glossterm linkend="glossary-table">table</glossterm> which
778 requires the value(s) in those <glossterm linkend="glossary-column">columns</glossterm> to
779 identify zero or one <glossterm linkend="glossary-tuple">row</glossterm>
780 in another (or, infrequently, the same)
781 <glossterm linkend="glossary-table">table</glossterm>.
782 </para>
783 </glossdef>
784 </glossentry>
785
786 <glossentry id="glossary-foreign-server">
787 <glossterm>Foreign server</glossterm>
788 <glossdef>
789 <para>
790 A named collection of
791 <glossterm linkend="glossary-foreign-table">foreign tables</glossterm> which
792 all use the same
793 <glossterm linkend="glossary-foreign-data-wrapper">foreign data wrapper</glossterm>
794 and have other configuration values in common.
795 </para>
796 <para>
797 For more information, see
798 <xref linkend="sql-createserver"/>.
799 </para>
800 </glossdef>
801 </glossentry>
802
803 <glossentry id="glossary-foreign-table">
804 <glossterm>Foreign table (relation)</glossterm>
805 <glossdef>
806 <para>
807 A <glossterm linkend="glossary-relation">relation</glossterm> which appears to have
808 <glossterm linkend="glossary-tuple">rows</glossterm> and
809 <glossterm linkend="glossary-column">columns</glossterm> similar to a
810 regular <glossterm linkend="glossary-table">table</glossterm>, but will forward
811 requests for data through its
812 <glossterm linkend="glossary-foreign-data-wrapper">foreign data wrapper</glossterm>,
813 which will return <glossterm linkend="glossary-result-set">result sets</glossterm>
814 structured according to the definition of the
815 <glossterm linkend="glossary-foreign-table">foreign table</glossterm>.
816 </para>
817 <para>
818 For more information, see
819 <xref linkend="sql-createforeigntable"/>.
820 </para>
821 </glossdef>
822 </glossentry>
823
824 <glossentry id="glossary-fork">
825 <glossterm>Fork</glossterm>
826 <glossdef>
827 <para>
828 Each of the separate segmented file sets in which a relation is stored.
829 The <firstterm>main fork</firstterm> is where the actual data resides.
830 There also exist two secondary forks for metadata:
831 the <glossterm linkend="glossary-fsm">free space map</glossterm>
832 and the <glossterm linkend="glossary-vm">visibility map</glossterm>.
833 <glossterm linkend="glossary-unlogged">Unlogged relations</glossterm>
834 also have an <firstterm>init fork</firstterm>.
835 </para>
836 </glossdef>
837 </glossentry>
838
839 <glossentry id="glossary-fsm">
840 <glossterm>Free space map (fork)</glossterm>
841 <glossdef>
842 <para>
843 A storage structure that keeps metadata about each data page of a table's
844 main fork. The free space map entry for each page stores the
845 amount of free space that's available for future tuples, and is structured
846 to be efficiently searched for available space for a new tuple of a given
847 size.
848 </para>
849 <para>
850 For more information, see
851 <xref linkend="storage-fsm"/>.
852 </para>
853 </glossdef>
854 </glossentry>
855
856 <glossentry id="glossary-function">
857 <glossterm>Function (routine)</glossterm>
858 <glossdef>
859 <para>
860 A type of routine that receives zero or more arguments, returns zero or more
861 output values, and is constrained to run within one transaction.
862 Functions are invoked as part of a query, for example via
863 <command>SELECT</command>.
864 Certain functions can return
865 <glossterm linkend="glossary-result-set">sets</glossterm>; those are
866 called <firstterm>set-returning functions</firstterm>.
867 </para>
868 <para>
869 Functions can also be used for
870 <glossterm linkend="glossary-trigger">triggers</glossterm> to invoke.
871 </para>
872 <para>
873 For more information, see
874 <xref linkend="sql-createfunction"/>.
875 </para>
876 </glossdef>
877 </glossentry>
878
879 <glossentry>
880 <glossterm>GMT</glossterm>
881 <glosssee otherterm="glossary-utc" />
882 </glossentry>
883
884 <glossentry id="glossary-grant">
885 <glossterm>Grant</glossterm>
886 <glossdef>
887 <para>
888 An <acronym>SQL</acronym> command that is used to allow a
889 <glossterm linkend="glossary-user">user</glossterm> or
890 <glossterm linkend="glossary-role">role</glossterm> to access
891 specific objects within the <glossterm linkend="glossary-database">database</glossterm>.
892 </para>
893 <para>
894 For more information, see
895 <xref linkend="sql-grant"/>.
896 </para>
897 </glossdef>
898 </glossentry>
899
900 <glossentry id="glossary-heap">
901 <glossterm>Heap</glossterm>
902 <glossdef>
903 <para>
904 Contains the values of <glossterm linkend="glossary-tuple">row</glossterm>
905 attributes (i.e., the data) for a
906 <glossterm linkend="glossary-relation">relation</glossterm>.
907 The heap is realized within one or more
908 <glossterm linkend="glossary-file-segment">file segments</glossterm>
909 in the relation's <glossterm linkend="glossary-fork">main fork</glossterm>.
910 </para>
911 </glossdef>
912 </glossentry>
913
914 <glossentry id="glossary-host">
915 <glossterm>Host</glossterm>
916 <glossdef>
917 <para>
918 A computer that communicates with other computers over a network.
919 This is sometimes used as a synonym for
920 <glossterm linkend="glossary-server">server</glossterm>.
921 It is also used to refer to a computer where
922 <glossterm linkend="glossary-client">client processes</glossterm> run.
923 </para>
924 </glossdef>
925 </glossentry>
926
927 <glossentry id="glossary-index">
928 <glossterm>Index (relation)</glossterm>
929 <glossdef>
930 <para>
931 A <glossterm linkend="glossary-relation">relation</glossterm> that contains
932 data derived from a <glossterm linkend="glossary-table">table</glossterm>
933 or <glossterm linkend="glossary-materialized-view">materialized view</glossterm>.
934 Its internal structure supports fast retrieval of and access to the original
935 data.
936 </para>
937 <para>
938 For more information, see
939 <xref linkend="sql-createindex"/>.
940 </para>
941 </glossdef>
942 </glossentry>
943 <glossentry id="glossary-incremental-backup">
944 <glossterm>Incremental backup</glossterm>
945 <glossdef>
946 <para>
947 A special <glossterm linkend="glossary-basebackup">base backup</glossterm>
948 that for some files may contain only those pages that were modified since
949 a previous backup, as opposed to the full contents of every file. Like
950 base backups, it is generated by the tool <xref linkend="app-pgbasebackup"/>.
951 </para>
952 <para>
953 To restore incremental backups the tool <xref linkend="app-pgcombinebackup"/>
954 is used, which combines incremental backups with a base backup.
955 Afterwards, recovery can use
956 <glossterm linkend="glossary-wal">WAL</glossterm> to bring the
957 <glossterm linkend="glossary-db-cluster">database cluster</glossterm> to
958 a consistent state.
959 </para>
960 <para>
961 For more information, see <xref linkend="backup-incremental-backup"/>.
962 </para>
963 </glossdef>
964 </glossentry>
965
966 <glossentry id="glossary-io">
967 <glossterm>Input/Output</glossterm>
968 <acronym>I/O</acronym>
969 <glossdef>
970 <para>
971 Input/Output (<acronym>I/O</acronym>) describes the communication between
972 a program and peripheral devices. In the context of database systems,
973 <acronym>I/O</acronym> commonly, but not exclusively, refers to
974 interaction with storage devices or the network.
975 </para>
976 <glossseealso otherterm="glossary-aio" />
977 </glossdef>
978 </glossentry>
979
980 <glossentry id="glossary-insert">
981 <glossterm>Insert</glossterm>
982 <glossdef>
983 <para>
984 An <acronym>SQL</acronym> command used to add new data into a
985 <glossterm linkend="glossary-table">table</glossterm>.
986 </para>
987 <para>
988 For more information, see
989 <xref linkend="sql-insert"/>.
990 </para>
991 </glossdef>
992 </glossentry>
993
994 <glossentry id="glossary-instance">
995 <glossterm>Instance</glossterm>
996 <glossdef>
997 <para>
998 A group of <glossterm linkend="glossary-backend">backend</glossterm> and
999 <glossterm linkend="glossary-auxiliary-proc">auxiliary processes</glossterm>
1000 that communicate using a common shared memory area. One
1001 <glossterm linkend="glossary-postmaster">postmaster process</glossterm>
1002 manages the instance; one instance manages exactly one
1003 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>
1004 with all its databases. Many instances can run on the same
1005 <glossterm linkend="glossary-server">server</glossterm>
1006 as long as their <acronym>TCP</acronym> ports do not conflict.
1007 </para>
1008 <para>
1009 The instance handles all key features of a <acronym>DBMS</acronym>:
1010 read and write access to files and shared memory,
1011 assurance of the <acronym>ACID</acronym> properties,
1012 <glossterm linkend="glossary-connection">connections</glossterm> to
1013 <glossterm linkend="glossary-client">client processes</glossterm>,
1014 privilege verification, crash recovery, replication, etc.
1015 </para>
1016 </glossdef>
1017 </glossentry>
1018
1019 <glossentry id="glossary-isolation">
1020 <glossterm>Isolation</glossterm>
1021 <glossdef>
1022 <para>
1023 The property that the effects of a transaction are not visible to
1024 <glossterm linkend="glossary-concurrency">concurrent transactions</glossterm>
1025 before it commits.
1026 This is one of the <acronym>ACID</acronym> properties.
1027 </para>
1028 <para>
1029 For more information, see <xref linkend="transaction-iso" />.
1030 </para>
1031 </glossdef>
1032 </glossentry>
1033
1034 <glossentry id="glossary-join">
1035 <glossterm>Join</glossterm>
1036 <glossdef>
1037 <para>
1038 An operation and <acronym>SQL</acronym> keyword used in
1039 <glossterm linkend="glossary-query">queries</glossterm>
1040 for combining data from multiple
1041 <glossterm linkend="glossary-relation">relations</glossterm>.
1042 </para>
1043 </glossdef>
1044 </glossentry>
1045
1046 <glossentry id="glossary-key">
1047 <glossterm>Key</glossterm>
1048 <glossdef>
1049 <para>
1050 A means of identifying a <glossterm linkend="glossary-tuple">row</glossterm> within a
1051 <glossterm linkend="glossary-table">table</glossterm> or
1052 other <glossterm linkend="glossary-relation">relation</glossterm> by
1053 values contained within one or more
1054 <glossterm linkend="glossary-attribute">attributes</glossterm>
1055 in that relation.
1056 </para>
1057 </glossdef>
1058 </glossentry>
1059
1060 <glossentry id="glossary-lock">
1061 <glossterm>Lock</glossterm>
1062 <glossdef>
1063 <para>
1064 A mechanism that allows a process to limit or prevent simultaneous
1065 access to a resource.
1066 </para>
1067 </glossdef>
1068 </glossentry>
1069
1070 <glossentry id="glossary-log-file">
1071 <glossterm>Log file</glossterm>
1072 <glossdef>
1073 <para>
1074 Log files contain human-readable text lines about events.
1075 Examples include login failures, long-running queries, etc.
1076 </para>
1077 <para>
1078 For more information, see
1079 <xref linkend="logfile-maintenance"/>.
1080 </para>
1081 </glossdef>
1082 </glossentry>
1083
1084 <glossentry id="glossary-logged">
1085 <glossterm>Logged</glossterm>
1086 <glossdef>
1087 <para>
1088 A <glossterm linkend="glossary-table">table</glossterm> is considered
1089 <glossterm linkend="glossary-logged">logged</glossterm> if changes to it are sent to the
1090 <glossterm linkend="glossary-wal">WAL</glossterm>. By default, all regular
1091 tables are logged. A table can be specified as
1092 <glossterm linkend="glossary-unlogged">unlogged</glossterm> either at
1093 creation time or via the <command>ALTER TABLE</command> command.
1094 </para>
1095 </glossdef>
1096 </glossentry>
1097
1098 <glossentry id="glossary-logger">
1099 <glossterm>Logger (process)</glossterm>
1100 <glossdef>
1101 <para>
1102 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
1103 which, if enabled, writes information about database events into the current
1104 <glossterm linkend="glossary-log-file">log file</glossterm>.
1105 When reaching certain time- or
1106 volume-dependent criteria, a new log file is created.
1107 Also called <firstterm>syslogger</firstterm>.
1108 </para>
1109 <para>
1110 For more information, see
1111 <xref linkend="runtime-config-logging"/>.
1112 </para>
1113 </glossdef>
1114 </glossentry>
1115
1116 <glossentry id="glossary-logical-replication-cluster">
1117 <glossterm>Logical replication cluster</glossterm>
1118 <glossdef>
1119 <para>
1120 A set of publisher and subscriber instances with the publisher instance
1121 replicating changes to the subscriber instance.
1122 </para>
1123 </glossdef>
1124 </glossentry>
1125
1126 <glossentry id="glossary-log-record">
1127 <glossterm>Log record</glossterm>
1128 <glossdef>
1129 <para>
1130 Archaic term for a <glossterm linkend="glossary-wal-record">WAL record</glossterm>.
1131 </para>
1132 </glossdef>
1133 </glossentry>
1134
1135 <glossentry id="glossary-log-sequence-number">
1136 <glossterm>Log sequence number</glossterm>
1137 <acronym>LSN</acronym>
1138 <glossdef>
1139 <para>
1140 Byte offset into the <glossterm linkend="glossary-wal">WAL</glossterm>,
1141 increasing monotonically with each new <glossterm
1142 linkend="glossary-wal-record">WAL record</glossterm>.
1143 </para>
1144 <para>
1145 For more information, see <link
1146 linkend="datatype-pg-lsn"><type>pg_lsn</type></link> and <xref
1147 linkend="wal-internals"/>.
1148 </para>
1149 </glossdef>
1150 </glossentry>
1151
1152 <glossentry>
1153 <glossterm>LSN</glossterm>
1154 <glosssee otherterm="glossary-log-sequence-number"/>
1155 </glossentry>
1156
1157 <glossentry>
1158 <glossterm>Master (server)</glossterm>
1159 <glosssee otherterm="glossary-primary-server" />
1160 </glossentry>
1161
1162 <glossentry id="glossary-materialized">
1163 <glossterm>Materialized</glossterm>
1164 <glossdef>
1165 <para>
1166 The property that some information has been pre-computed and stored
1167 for later use, rather than computing it on-the-fly.
1168 </para>
1169 <para>
1170 This term is used in
1171 <glossterm linkend="glossary-materialized-view">materialized view</glossterm>,
1172 to mean that the data derived from the view's query is stored on
1173 disk separately from the sources of that data.
1174 </para>
1175 <para>
1176 This term is also used to refer to some multi-step queries to mean that
1177 the data resulting from executing a given step is stored in memory
1178 (with the possibility of spilling to disk), so that it can be read multiple
1179 times by another step.
1180 </para>
1181 </glossdef>
1182 </glossentry>
1183
1184 <glossentry id="glossary-materialized-view">
1185 <glossterm>Materialized view (relation)</glossterm>
1186 <glossdef>
1187 <para>
1188 A <glossterm linkend="glossary-relation">relation</glossterm> that is
1189 defined by a <command>SELECT</command> statement
1190 (just like a <glossterm linkend="glossary-view">view</glossterm>),
1191 but stores data in the same way that a
1192 <glossterm linkend="glossary-table">table</glossterm> does. It cannot be
1193 modified via <command>INSERT</command>, <command>UPDATE</command>,
1194 <command>DELETE</command>, or <command>MERGE</command> operations.
1195 </para>
1196 <para>
1197 For more information, see
1198 <xref linkend="sql-creatematerializedview"/>.
1199 </para>
1200 </glossdef>
1201 </glossentry>
1202
1203 <glossentry id="glossary-merge">
1204 <glossterm>Merge</glossterm>
1205 <glossdef>
1206 <para>
1207 An <acronym>SQL</acronym> command used to conditionally add, modify,
1208 or remove <glossterm linkend="glossary-tuple">rows</glossterm>
1209 in a given <glossterm linkend="glossary-table">table</glossterm>,
1210 using data from a source
1211 <glossterm linkend="glossary-relation">relation</glossterm>.
1212 </para>
1213 <para>
1214 For more information, see
1215 <xref linkend="sql-merge"/>.
1216 </para>
1217 </glossdef>
1218 </glossentry>
1219
1220 <glossentry id="glossary-mvcc">
1221 <glossterm>Multi-version concurrency control (MVCC)</glossterm>
1222 <glossdef>
1223 <para>
1224 A mechanism designed to allow several
1225 <glossterm linkend="glossary-transaction">transactions</glossterm> to be
1226 reading and writing the same rows without one process causing other
1227 processes to stall.
1228 In <productname>PostgreSQL</productname>, MVCC is implemented by
1229 creating copies (<firstterm>versions</firstterm>) of
1230 <glossterm linkend="glossary-tuple">tuples</glossterm> as they are
1231 modified; after transactions that can see the old versions terminate,
1232 those old versions need to be removed.
1233 </para>
1234 </glossdef>
1235 </glossentry>
1236
1237 <glossentry id="glossary-null">
1238 <glossterm>Null</glossterm>
1239 <glossdef>
1240 <para>
1241 A concept of non-existence that is a central tenet of relational
1242 database theory. It represents the absence of a definite value.
1243 </para>
1244 </glossdef>
1245 </glossentry>
1246
1247 <glossentry>
1248 <glossterm>Optimizer</glossterm>
1249 <glosssee otherterm="glossary-planner" />
1250 </glossentry>
1251
1252 <glossentry id="glossary-parallel-query">
1253 <glossterm>Parallel query</glossterm>
1254 <glossdef>
1255 <para>
1256 The ability to handle parts of executing a
1257 <glossterm linkend="glossary-query">query</glossterm> to take advantage
1258 of parallel processes on servers with multiple <acronym>CPU</acronym>s.
1259 </para>
1260 </glossdef>
1261 </glossentry>
1262
1263 <glossentry id="glossary-partition">
1264 <glossterm>Partition</glossterm>
1265 <glossdef>
1266 <para>
1267 One of several disjoint (not overlapping) subsets of a larger set.
1268 </para>
1269 </glossdef>
1270 <glossdef>
1271 <para>
1272 In reference to a
1273 <glossterm linkend="glossary-partitioned-table">partitioned table</glossterm>:
1274 One of the tables that each contain part of the data of the partitioned table,
1275 which is said to be the <firstterm>parent</firstterm>.
1276 The partition is itself a table, so it can also be queried directly;
1277 at the same time, a partition can sometimes be a partitioned table,
1278 allowing hierarchies to be created.
1279 </para>
1280 </glossdef>
1281 <glossdef>
1282 <para>
1283 In reference to a <glossterm linkend="glossary-window-function">window function</glossterm>
1284 in a <glossterm linkend="glossary-query">query</glossterm>,
1285 a partition is a user-defined criterion that identifies which neighboring
1286 <glossterm linkend="glossary-tuple">rows</glossterm>
1287 of the <glossterm linkend="glossary-result-set">query's result set</glossterm>
1288 can be considered by the function.
1289 </para>
1290 </glossdef>
1291 </glossentry>
1292
1293 <glossentry id="glossary-partitioned-table">
1294 <glossterm>Partitioned table (relation)</glossterm>
1295 <glossdef>
1296 <para>
1297 A <glossterm linkend="glossary-relation">relation</glossterm> that is
1298 in semantic terms the same as a <glossterm linkend="glossary-table">table</glossterm>,
1299 but whose storage is distributed across several
1300 <glossterm linkend="glossary-partition">partitions</glossterm>.
1301 </para>
1302 </glossdef>
1303 </glossentry>
1304
1305 <glossentry id="glossary-postmaster">
1306 <glossterm>Postmaster (process)</glossterm>
1307 <glossdef>
1308 <para>
1309 The very first process of an <glossterm linkend="glossary-instance">instance</glossterm>.
1310 It starts and manages the
1311 <glossterm linkend="glossary-auxiliary-proc">auxiliary processes</glossterm>
1312 and creates <glossterm linkend="glossary-backend">backend processes</glossterm>
1313 on demand.
1314 </para>
1315 <para>
1316 For more information, see
1317 <xref linkend="server-start"/>.
1318 </para>
1319 </glossdef>
1320 </glossentry>
1321
1322 <glossentry id="glossary-primary-key">
1323 <glossterm>Primary key</glossterm>
1324 <glossdef>
1325 <para>
1326 A special case of a
1327 <glossterm linkend="glossary-unique-constraint">unique constraint</glossterm>
1328 defined on a
1329 <glossterm linkend="glossary-table">table</glossterm> or other
1330 <glossterm linkend="glossary-relation">relation</glossterm> that also
1331 guarantees that all of the
1332 <glossterm linkend="glossary-attribute">attributes</glossterm>
1333 within the <glossterm linkend="glossary-primary-key">primary key</glossterm>
1334 do not have <glossterm linkend="glossary-null">null</glossterm> values.
1335 As the name implies, there can be only one
1336 primary key per table, though it is possible to have multiple unique
1337 constraints that also have no null-capable attributes.
1338 </para>
1339 </glossdef>
1340 </glossentry>
1341
1342 <glossentry id="glossary-primary-server">
1343 <glossterm>Primary (server)</glossterm>
1344 <glossdef>
1345 <para>
1346 When two or more <glossterm linkend="glossary-database">databases</glossterm>
1347 are linked via <glossterm linkend="glossary-replication">replication</glossterm>,
1348 the <glossterm linkend="glossary-server">server</glossterm>
1349 that is considered the authoritative source of information is called
1350 the <firstterm>primary</firstterm>,
1351 also known as a <firstterm>master</firstterm>.
1352 </para>
1353 </glossdef>
1354 </glossentry>
1355
1356 <glossentry id="glossary-procedure">
1357 <glossterm>Procedure (routine)</glossterm>
1358 <glossdef>
1359 <para>
1360 A type of routine.
1361 Their distinctive qualities are that they do not return values,
1362 and that they are allowed to make transactional statements such
1363 as <command>COMMIT</command> and <command>ROLLBACK</command>.
1364 They are invoked via the <command>CALL</command> command.
1365 </para>
1366 <para>
1367 For more information, see
1368 <xref linkend="sql-createprocedure"/>.
1369 </para>
1370 </glossdef>
1371 </glossentry>
1372
1373 <glossentry id="glossary-query">
1374 <glossterm>Query</glossterm>
1375 <glossdef>
1376 <para>
1377 A request sent by a client to a <glossterm linkend="glossary-backend">backend</glossterm>,
1378 usually to return results or to modify data on the database.
1379 </para>
1380 </glossdef>
1381 </glossentry>
1382
1383 <glossentry id="glossary-planner">
1384 <glossterm>Query planner</glossterm>
1385 <glossdef>
1386 <para>
1387 The part of <productname>PostgreSQL</productname> that is devoted to
1388 determining (<firstterm>planning</firstterm>) the most efficient way to
1389 execute <glossterm linkend="glossary-query">queries</glossterm>.
1390 Also known as <firstterm>query optimizer</firstterm>,
1391 <firstterm>optimizer</firstterm>, or simply <firstterm>planner</firstterm>.
1392 </para>
1393 </glossdef>
1394 </glossentry>
1395
1396 <glossentry>
1397 <glossterm>Record</glossterm>
1398 <glosssee otherterm="glossary-tuple" />
1399 </glossentry>
1400
1401 <glossentry>
1402 <glossterm>Recycling</glossterm>
1403 <glosssee otherterm="glossary-wal-file" />
1404 </glossentry>
1405
1406 <glossentry id="glossary-referential-integrity">
1407 <glossterm>Referential integrity</glossterm>
1408 <glossdef>
1409 <para>
1410 A means of restricting data in one <glossterm linkend="glossary-relation">relation</glossterm>
1411 by a <glossterm linkend="glossary-foreign-key">foreign key</glossterm>
1412 so that it must have matching data in another
1413 <glossterm linkend="glossary-relation">relation</glossterm>.
1414 </para>
1415 </glossdef>
1416 </glossentry>
1417
1418 <glossentry id="glossary-relation">
1419 <glossterm>Relation</glossterm>
1420 <glossdef>
1421 <para>
1422 The generic term for all objects in a
1423 <glossterm linkend="glossary-database">database</glossterm>
1424 that have a name and a list of
1425 <glossterm linkend="glossary-attribute">attributes</glossterm>
1426 defined in a specific order.
1427 <glossterm linkend="glossary-table">Tables</glossterm>,
1428 <glossterm linkend="glossary-sequence">sequences</glossterm>,
1429 <glossterm linkend="glossary-view">views</glossterm>,
1430 <glossterm linkend="glossary-foreign-table">foreign tables</glossterm>,
1431 <glossterm linkend="glossary-materialized-view">materialized views</glossterm>,
1432 composite types, and
1433 <glossterm linkend="glossary-index">indexes</glossterm> are all relations.
1434 </para>
1435 <para>
1436 More generically, a relation is a set of tuples; for example,
1437 the result of a query is also a relation.
1438 </para>
1439 <para>
1440 In <productname>PostgreSQL</productname>,
1441 <firstterm>Class</firstterm> is an archaic synonym for
1442 <firstterm>relation</firstterm>.
1443 </para>
1444 </glossdef>
1445 </glossentry>
1446
1447 <glossentry id="glossary-replica">
1448 <glossterm>Replica (server)</glossterm>
1449 <glossdef>
1450 <para>
1451 A <glossterm linkend="glossary-database">database</glossterm> that is paired
1452 with a <glossterm linkend="glossary-primary-server">primary</glossterm>
1453 database and is maintaining a copy of some or all of the primary database's
1454 data. The foremost reasons for doing this are to allow for greater access
1455 to that data, and to maintain availability of the data in the event that
1456 the <glossterm linkend="glossary-primary-server">primary</glossterm>
1457 becomes unavailable.
1458 </para>
1459 </glossdef>
1460 </glossentry>
1461
1462 <glossentry id="glossary-replication">
1463 <glossterm>Replication</glossterm>
1464 <glossdef>
1465 <para>
1466 The act of reproducing data on one
1467 <glossterm linkend="glossary-server">server</glossterm> onto another
1468 server called a <glossterm linkend="glossary-replica">replica</glossterm>.
1469 This can take the form of <firstterm>physical replication</firstterm>,
1470 where all file changes from one server are copied verbatim,
1471 or <firstterm>logical replication</firstterm> where a defined subset
1472 of data changes are conveyed using a higher-level representation.
1473 </para>
1474 </glossdef>
1475 </glossentry>
1476
1477 <glossentry id="glossary-restartpoint">
1478 <glossterm>Restartpoint</glossterm>
1479 <glossdef>
1480 <para>
1481 A variant of a <glossterm
1482 linkend="glossary-checkpoint">checkpoint</glossterm> performed on a
1483 <glossterm linkend="glossary-replica">replica</glossterm>.
1484 </para>
1485 <para>
1486 For more information, see <xref linkend="wal-configuration"/>.
1487 </para>
1488 </glossdef>
1489 </glossentry>
1490
1491 <glossentry id="glossary-result-set">
1492 <glossterm>Result set</glossterm>
1493 <glossdef>
1494 <para>
1495 A <glossterm linkend="glossary-relation">relation</glossterm> transmitted
1496 from a <glossterm linkend="glossary-backend">backend process</glossterm>
1497 to a <glossterm linkend="glossary-client">client</glossterm> upon the
1498 completion of an <acronym>SQL</acronym> command, usually a
1499 <command>SELECT</command> but it can be an
1500 <command>INSERT</command>, <command>UPDATE</command>,
1501 <command>DELETE</command>, or <command>MERGE</command> command if the
1502 <literal>RETURNING</literal> clause is specified.
1503 </para>
1504 <para>
1505 The fact that a result set is a relation means that a query can be used
1506 in the definition of another query, becoming a
1507 <firstterm>subquery</firstterm>.
1508 </para>
1509 </glossdef>
1510 <glossdef>
1511 <para>
1512 </para>
1513 </glossdef>
1514 </glossentry>
1515
1516 <glossentry id="glossary-revoke">
1517 <glossterm>Revoke</glossterm>
1518 <glossdef>
1519 <para>
1520 A command to prevent access to a named set of
1521 <glossterm linkend="glossary-database">database</glossterm> objects for a
1522 named list of <glossterm linkend="glossary-role">roles</glossterm>.
1523 </para>
1524 <para>
1525 For more information, see
1526 <xref linkend="sql-revoke"/>.
1527 </para>
1528 </glossdef>
1529 </glossentry>
1530
1531 <glossentry id="glossary-role">
1532 <glossterm>Role</glossterm>
1533 <glossdef>
1534 <para>
1535 A collection of access privileges to the
1536 <glossterm linkend="glossary-database">instance</glossterm>.
1537 Roles are themselves a privilege that can be granted to other roles.
1538 This is often done for convenience or to ensure completeness
1539 when multiple <glossterm linkend="glossary-user">users</glossterm> need
1540 the same privileges.
1541 </para>
1542 <para>
1543 For more information, see
1544 <xref linkend="sql-createrole"/>.
1545 </para>
1546 </glossdef>
1547 </glossentry>
1548
1549 <glossentry id="glossary-rollback">
1550 <glossterm>Rollback</glossterm>
1551 <glossdef>
1552 <para>
1553 A command to undo all of the operations performed since the beginning
1554 of a <glossterm linkend="glossary-transaction">transaction</glossterm>.
1555 </para>
1556 <para>
1557 For more information, see
1558 <xref linkend="sql-rollback"/>.
1559 </para>
1560 </glossdef>
1561 </glossentry>
1562
1563 <glossentry id="glossary-routine">
1564 <glossterm>Routine</glossterm>
1565 <glossdef>
1566 <para>
1567 A defined set of instructions stored in the database system
1568 that can be invoked for execution.
1569 A routine can be written in a variety of programming
1570 languages. Routines can be
1571 <glossterm linkend="glossary-function">functions</glossterm>
1572 (including set-returning functions and
1573 <glossterm linkend="glossary-trigger">trigger functions</glossterm>),
1574 <glossterm linkend="glossary-aggregate">aggregate functions</glossterm>,
1575 and <glossterm linkend="glossary-procedure">procedures</glossterm>.
1576 </para>
1577 <para>
1578 Many routines are already defined within <productname>PostgreSQL</productname>
1579 itself, but user-defined ones can also be added.
1580 </para>
1581 </glossdef>
1582 </glossentry>
1583
1584 <glossentry>
1585 <glossterm>Row</glossterm>
1586 <glosssee otherterm="glossary-tuple" />
1587 </glossentry>
1588
1589 <glossentry id="glossary-savepoint">
1590 <glossterm>Savepoint</glossterm>
1591 <glossdef>
1592 <para>
1593 A special mark in the sequence of steps in a
1594 <glossterm linkend="glossary-transaction">transaction</glossterm>.
1595 Data modifications after this point in time may be reverted
1596 to the time of the savepoint.
1597 </para>
1598 <para>
1599 For more information, see
1600 <xref linkend="sql-savepoint"/>.
1601 </para>
1602 </glossdef>
1603 </glossentry>
1604
1605 <glossentry id="glossary-schema">
1606 <glossterm>Schema</glossterm>
1607 <glossdef>
1608 <para>
1609 A schema is a namespace for
1610 <glossterm linkend="glossary-sql-object">SQL objects</glossterm>,
1611 which all reside in the same
1612 <glossterm linkend="glossary-database">database</glossterm>.
1613 Each SQL object must reside in exactly one schema.
1614 </para>
1615 <para>
1616 All system-defined SQL objects reside in schema <literal>pg_catalog</literal>.
1617 </para>
1618 </glossdef>
1619 <glossdef>
1620 <para>
1621 More generically, the term <firstterm>schema</firstterm> is used to mean
1622 all data descriptions (<glossterm linkend="glossary-table">table</glossterm> definitions,
1623 <glossterm linkend="glossary-constraint">constraints</glossterm>, comments, etc.)
1624 for a given <glossterm linkend="glossary-database">database</glossterm> or
1625 subset thereof.
1626 </para>
1627 <para>
1628 For more information, see
1629 <xref linkend="ddl-schemas"/>.
1630 </para>
1631 </glossdef>
1632 </glossentry>
1633
1634 <glossentry>
1635 <glossterm>Segment</glossterm>
1636 <glosssee otherterm="glossary-file-segment" />
1637 </glossentry>
1638
1639 <glossentry id="glossary-select">
1640 <glossterm>Select</glossterm>
1641 <glossdef>
1642 <para>
1643 The <acronym>SQL</acronym> command used to request data from a
1644 <glossterm linkend="glossary-database">database</glossterm>.
1645 Normally, <command>SELECT</command> commands are not expected to modify the
1646 <glossterm linkend="glossary-database">database</glossterm> in any way,
1647 but it is possible that
1648 <glossterm linkend="glossary-function">functions</glossterm> invoked within
1649 the query could have side effects that do modify data.
1650 </para>
1651 <para>
1652 For more information, see
1653 <xref linkend="sql-select"/>.
1654 </para>
1655 </glossdef>
1656 </glossentry>
1657
1658 <glossentry id="glossary-sequence">
1659 <glossterm>Sequence (relation)</glossterm>
1660 <glossdef>
1661 <para>
1662 A type of relation that is used to generate values.
1663 Typically the generated values are sequential non-repeating numbers.
1664 They are commonly used to generate surrogate
1665 <glossterm linkend="glossary-primary-key">primary key</glossterm>
1666 values.
1667 </para>
1668 </glossdef>
1669 </glossentry>
1670
1671 <!-- XXX should define all other isolation levels (and improve this definition)
1672 <glossentry id="glossary-serializable">
1673 <glossterm>Serializable (isolation level)</glossterm>
1674 <glossdef>
1675 <para>
1676 Transactions defined as <literal>SERIALIZABLE</literal> are unable to
1677 see changes made within other transactions. In effect, for the
1678 initializing session the entire <glossterm linkend="glossary-database">database</glossterm>
1679 appears to be frozen for the duration of the
1680 <glossterm linkend="glossary-transaction">Transaction</glossterm>.
1681 </para>
1682 </glossdef>
1683 </glossentry>
1684 -->
1685
1686 <glossentry id="glossary-server">
1687 <glossterm>Server</glossterm>
1688 <glossdef>
1689 <para>
1690 A computer on which <productname>PostgreSQL</productname>
1691 <glossterm linkend="glossary-instance">instances</glossterm> run.
1692 The term <firstterm>server</firstterm> denotes real hardware, a
1693 container, or a <firstterm>virtual machine</firstterm>.
1694 </para>
1695 <para>
1696 This term is sometimes used to refer to an instance or to a host.
1697 </para>
1698 </glossdef>
1699 </glossentry>
1700
1701 <glossentry id="glossary-session">
1702 <glossterm>Session</glossterm>
1703 <glossdef>
1704 <para>
1705 A state that allows a client and a backend to interact,
1706 communicating over a <glossterm linkend="glossary-connection">connection</glossterm>.
1707 </para>
1708 </glossdef>
1709 </glossentry>
1710
1711 <glossentry id="glossary-shared-memory">
1712 <glossterm>Shared memory</glossterm>
1713 <glossdef>
1714 <para>
1715 <acronym>RAM</acronym> which is used by the processes common to an
1716 <glossterm linkend="glossary-instance">instance</glossterm>.
1717 It mirrors parts of <glossterm linkend="glossary-database">database</glossterm>
1718 files, provides a transient area for
1719 <glossterm linkend="glossary-wal-record">WAL records</glossterm>,
1720 and stores additional common information.
1721 Note that shared memory belongs to the complete instance, not to a single
1722 database.
1723 </para>
1724 <para>
1725 The largest part of shared memory is known as <firstterm>shared buffers</firstterm>
1726 and is used to mirror part of data files, organized into pages.
1727 When a page is modified, it is called a dirty page until it is
1728 written back to the file system.
1729 </para>
1730 <para>
1731 For more information, see
1732 <xref linkend="runtime-config-resource-memory"/>.
1733 </para>
1734 </glossdef>
1735 </glossentry>
1736
1737 <glossentry id="glossary-sql-object">
1738 <glossterm>SQL object</glossterm>
1739 <glossdef>
1740 <para>
1741 Any object that can be created with a <command>CREATE</command>
1742 command. Most objects are specific to one database, and are commonly
1743 known as <firstterm>local objects</firstterm>.
1744 </para>
1745 <para>
1746 Most local objects reside in a specific
1747 <glossterm linkend="glossary-schema">schema</glossterm> in their
1748 containing database, such as
1749 <glossterm linkend="glossary-relation">relations</glossterm> (all types),
1750 <glossterm linkend="glossary-function">routines</glossterm> (all types),
1751 data types, etc.
1752 The names of such objects of the same type in the same schema
1753 are enforced to be unique.
1754 </para>
1755 <para>
1756 There also exist local objects that do not reside in schemas; some examples are
1757 <glossterm linkend="glossary-extension">extensions</glossterm>,
1758 <glossterm linkend="glossary-cast">data type casts</glossterm>, and
1759 <glossterm linkend="glossary-foreign-data-wrapper">foreign data wrappers</glossterm>.
1760 The names of such objects of the same type are enforced to be unique
1761 within the database.
1762 </para>
1763 <para>
1764 Other object types, such as
1765 <glossterm linkend="glossary-role">roles</glossterm>,
1766 <glossterm linkend="glossary-tablespace">tablespaces</glossterm>,
1767 replication origins, subscriptions for logical replication, and
1768 databases themselves are not local SQL objects since they exist
1769 entirely outside of any specific database;
1770 they are called <firstterm>global objects</firstterm>.
1771 The names of such objects are enforced to be unique within the whole
1772 database cluster.
1773 </para>
1774 <para>
1775 For more information, see
1776 <xref linkend="manage-ag-overview"/>.
1777 </para>
1778 </glossdef>
1779 </glossentry>
1780
1781 <glossentry id="glossary-sql-standard">
1782 <glossterm>SQL standard</glossterm>
1783 <glossdef>
1784 <para>
1785 A series of documents that define the <acronym>SQL</acronym> language.
1786 </para>
1787 </glossdef>
1788 </glossentry>
1789
1790 <glossentry>
1791 <glossterm>Standby (server)</glossterm>
1792 <glosssee otherterm="glossary-replica" />
1793 </glossentry>
1794
1795 <glossentry id="glossary-startup-process">
1796 <glossterm>Startup process</glossterm>
1797 <glossdef>
1798 <para>
1799 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
1800 that replays WAL during crash recovery and in a
1801 <glossterm linkend="glossary-replication">physical replica</glossterm>.
1802 </para>
1803 <para>
1804 (The name is historical: the startup process was named before
1805 replication was implemented; the name refers to its task as it
1806 relates to the server startup following a crash.)
1807 </para>
1808 </glossdef>
1809 </glossentry>
1810
1811 <glossentry id="glossary-superuser">
1812 <glossterm>Superuser</glossterm>
1813 <glossdef>
1814 <para>
1815 As used in this documentation, it is a synonym for
1816 <glossterm linkend="glossary-database-superuser">database superuser</glossterm>.
1817 </para>
1818 </glossdef>
1819 </glossentry>
1820
1821 <glossentry id="glossary-system-catalog">
1822 <glossterm>System catalog</glossterm>
1823 <glossdef>
1824 <para>
1825 A collection of <glossterm linkend="glossary-table">tables</glossterm>
1826 which describe the structure of all
1827 <glossterm linkend="glossary-sql-object">SQL objects</glossterm>
1828 of the instance.
1829 The system catalog resides in the schema <literal>pg_catalog</literal>.
1830 These tables contain data in internal representation and are
1831 not typically considered useful for user examination;
1832 a number of user-friendlier <glossterm linkend="glossary-view">views</glossterm>,
1833 also in schema <literal>pg_catalog</literal>, offer more convenient access to
1834 some of that information, while additional tables and views
1835 exist in schema <literal>information_schema</literal>
1836 (see <xref linkend="information-schema" />) that expose some
1837 of the same and additional information as mandated by the
1838 <glossterm linkend="glossary-sql-standard">SQL standard</glossterm>.
1839 </para>
1840 <para>
1841 For more information, see
1842 <xref linkend="ddl-schemas"/>.
1843 </para>
1844 </glossdef>
1845 </glossentry>
1846
1847 <glossentry id="glossary-table">
1848 <glossterm>Table</glossterm>
1849 <glossdef>
1850 <para>
1851 A collection of <glossterm linkend="glossary-tuple">tuples</glossterm> having
1852 a common data structure (the same number of
1853 <glossterm linkend="glossary-attribute">attributes</glossterm>, in the same
1854 order, having the same name and type per position).
1855 A table is the most common form of
1856 <glossterm linkend="glossary-relation">relation</glossterm> in
1857 <productname>PostgreSQL</productname>.
1858 </para>
1859 <para>
1860 For more information, see
1861 <xref linkend="sql-createtable"/>.
1862 </para>
1863 </glossdef>
1864 </glossentry>
1865
1866 <glossentry id="glossary-tablespace">
1867 <glossterm>Tablespace</glossterm>
1868 <glossdef>
1869 <para>
1870 A named location on the server file system.
1871 All <glossterm linkend="glossary-sql-object">SQL objects</glossterm>
1872 which require storage beyond their definition in the
1873 <glossterm linkend="glossary-system-catalog">system catalog</glossterm>
1874 must belong to a single tablespace.
1875 Initially, a database cluster contains a single usable tablespace which is
1876 used as the default for all SQL objects, called <literal>pg_default</literal>.
1877 </para>
1878 <para>
1879 For more information, see
1880 <xref linkend="manage-ag-tablespaces"/>.
1881 </para>
1882 </glossdef>
1883 </glossentry>
1884
1885 <glossentry id="glossary-temporary-table">
1886 <glossterm>Temporary table</glossterm>
1887 <glossdef>
1888 <para>
1889 <glossterm linkend="glossary-table">Tables</glossterm> that exist either
1890 for the lifetime of a
1891 <glossterm linkend="glossary-session">session</glossterm> or a
1892 <glossterm linkend="glossary-transaction">transaction</glossterm>, as
1893 specified at the time of creation.
1894 The data in them is not visible to other sessions, and is not
1895 <glossterm linkend="glossary-logged">logged</glossterm>.
1896 Temporary tables are often used to store intermediate data for a
1897 multi-step operation.
1898 </para>
1899 <para>
1900 For more information, see
1901 <xref linkend="sql-createtable"/>.
1902 </para>
1903 </glossdef>
1904 </glossentry>
1905
1906 <glossentry id="glossary-toast">
1907 <glossterm>TOAST</glossterm>
1908 <glossdef>
1909 <para>
1910 A mechanism by which large attributes of table rows are split and
1911 stored in a secondary table, called the <firstterm>TOAST table</firstterm>.
1912 Each relation with large attributes has its own TOAST table.
1913 </para>
1914 <para>
1915 For more information, see
1916 <xref linkend="storage-toast" />.
1917 </para>
1918 </glossdef>
1919 </glossentry>
1920
1921 <glossentry id="glossary-transaction">
1922 <glossterm>Transaction</glossterm>
1923 <glossdef>
1924 <para>
1925 A combination of commands that must act as a single
1926 <glossterm linkend="glossary-atomic">atomic</glossterm> command: they all
1927 succeed or all fail as a single unit, and their effects are not visible to
1928 other <glossterm linkend="glossary-session">sessions</glossterm> until
1929 the transaction is complete, and possibly even later, depending on the
1930 isolation level.
1931 </para>
1932 <para>
1933 For more information, see
1934 <xref linkend="transaction-iso"/>.
1935 </para>
1936 </glossdef>
1937 </glossentry>
1938
1939 <glossentry id="glossary-xid">
1940 <glossterm>Transaction ID</glossterm>
1941 <glossdef>
1942 <para>
1943 The numerical, unique, sequentially-assigned identifier that each
1944 transaction receives when it first causes a database modification.
1945 Frequently abbreviated as <firstterm>xid</firstterm>.
1946 When stored on disk, xids are only 32-bits wide, so only
1947 approximately four billion write transaction IDs can be generated;
1948 to permit the system to run for longer than that,
1949 <firstterm>epochs</firstterm> are used, also 32 bits wide.
1950 When the counter reaches the maximum xid value, it starts over at
1951 <literal>3</literal> (values under that are reserved) and the
1952 epoch value is incremented by one.
1953 In some contexts, the epoch and xid values are
1954 considered together as a single 64-bit value; see <xref
1955 linkend="transaction-id"/> for more details.
1956 </para>
1957 <para>
1958 For more information, see
1959 <xref linkend="datatype-oid"/>.
1960 </para>
1961 </glossdef>
1962 </glossentry>
1963
1964 <glossentry id="glossary-tps">
1965 <glossterm>Transactions per second (TPS)</glossterm>
1966 <glossdef>
1967 <para>
1968 Average number of transactions that are executed per second,
1969 totaled across all sessions active for a measured run.
1970 This is used as a measure of the performance characteristics of
1971 an instance.
1972 </para>
1973 </glossdef>
1974 </glossentry>
1975
1976 <glossentry id="glossary-trigger">
1977 <glossterm>Trigger</glossterm>
1978 <glossdef>
1979 <para>
1980 A <glossterm linkend="glossary-function">function</glossterm> which can
1981 be defined to execute whenever a certain operation (<command>INSERT</command>,
1982 <command>UPDATE</command>, <command>DELETE</command>,
1983 <command>TRUNCATE</command>) is applied to a
1984 <glossterm linkend="glossary-relation">relation</glossterm>.
1985 A trigger executes within the same
1986 <glossterm linkend="glossary-transaction">transaction</glossterm> as the
1987 statement which invoked it, and if the function fails, then the invoking
1988 statement also fails.
1989 </para>
1990 <para>
1991 For more information, see
1992 <xref linkend="sql-createtrigger"/>.
1993 </para>
1994 </glossdef>
1995 </glossentry>
1996
1997 <glossentry id="glossary-tuple">
1998 <glossterm>Tuple</glossterm>
1999 <glossdef>
2000 <para>
2001 A collection of <glossterm linkend="glossary-attribute">attributes</glossterm>
2002 in a fixed order.
2003 That order may be defined by the <glossterm linkend="glossary-table">table</glossterm>
2004 (or other <glossterm linkend="glossary-relation">relation</glossterm>)
2005 where the tuple is contained, in which case the tuple is often called a
2006 <firstterm>row</firstterm>. It may also be defined by the structure of a
2007 result set, in which case it is sometimes called a <firstterm>record</firstterm>.
2008 </para>
2009 </glossdef>
2010 </glossentry>
2011
2012 <glossentry id="glossary-unique-constraint">
2013 <glossterm>Unique constraint</glossterm>
2014 <glossdef>
2015 <para>
2016 A type of <glossterm linkend="glossary-constraint">constraint</glossterm>
2017 defined on a <glossterm linkend="glossary-relation">relation</glossterm>
2018 which restricts the values allowed in one or a combination of columns
2019 so that each value or combination of values can only appear once in the
2020 relation &mdash; that is, no other row in the relation contains values
2021 that are equal to those.
2022 </para>
2023 <para>
2024 Because <glossterm linkend="glossary-null">null values</glossterm> are
2025 not considered equal to each other, multiple rows with null values are
2026 allowed to exist without violating the unique constraint.
2027 </para>
2028 </glossdef>
2029 </glossentry>
2030
2031 <glossentry id="glossary-unlogged">
2032 <glossterm>Unlogged</glossterm>
2033 <glossdef>
2034 <para>
2035 The property of certain <glossterm linkend="glossary-relation">relations</glossterm>
2036 that the changes to them are not reflected in the
2037 <glossterm linkend="glossary-wal">WAL</glossterm>.
2038 This disables replication and crash recovery for these relations.
2039 </para>
2040 <para>
2041 The primary use of unlogged tables is for storing
2042 transient work data that must be shared across processes.
2043 </para>
2044 <para>
2045 <glossterm linkend="glossary-temporary-table">Temporary tables</glossterm>
2046 are always unlogged.
2047 </para>
2048 </glossdef>
2049 </glossentry>
2050
2051 <glossentry id="glossary-update">
2052 <glossterm>Update</glossterm>
2053 <glossdef>
2054 <para>
2055 An <acronym>SQL</acronym> command used to modify
2056 <glossterm linkend="glossary-tuple">rows</glossterm>
2057 that may already exist in a specified <glossterm linkend="glossary-table">table</glossterm>.
2058 It cannot create or remove rows.
2059 </para>
2060 <para>
2061 For more information, see
2062 <xref linkend="sql-update"/>.
2063 </para>
2064 </glossdef>
2065 </glossentry>
2066
2067 <glossentry id="glossary-user">
2068 <glossterm>User</glossterm>
2069 <glossdef>
2070 <para>
2071 A <glossterm linkend="glossary-role">role</glossterm> that has the
2072 <firstterm>login privilege</firstterm>
2073 (see <xref linkend="role-attributes"/>).
2074 </para>
2075 </glossdef>
2076 </glossentry>
2077
2078 <glossentry id="glossary-user-mapping">
2079 <glossterm>User mapping</glossterm>
2080 <glossdef>
2081 <para>
2082 The translation of login credentials in the local
2083 <glossterm linkend="glossary-database">database</glossterm> to credentials
2084 in a remote data system defined by a
2085 <glossterm linkend="glossary-foreign-data-wrapper">foreign data wrapper</glossterm>.
2086 </para>
2087 <para>
2088 For more information, see
2089 <xref linkend="sql-createusermapping"/>.
2090 </para>
2091 </glossdef>
2092 </glossentry>
2093
2094 <glossentry id="glossary-utc">
2095 <glossterm>UTC</glossterm>
2096 <glossdef>
2097 <para>
2098 Universal Coordinated Time, the primary global time reference,
2099 approximately the time prevailing at the zero meridian of longitude.
2100 Often but inaccurately referred to as GMT (Greenwich Mean Time).
2101 </para>
2102 </glossdef>
2103 </glossentry>
2104
2105 <glossentry id="glossary-vacuum">
2106 <glossterm>Vacuum</glossterm>
2107 <glossdef>
2108 <para>
2109 The process of removing outdated
2110 <glossterm linkend="glossary-tuple">tuple versions</glossterm>
2111 from tables or materialized views, and other closely related
2112 processing required by <productname>PostgreSQL</productname>'s
2113 implementation of <glossterm linkend="glossary-mvcc">MVCC</glossterm>.
2114 This can be initiated through the use of
2115 the <command>VACUUM</command> command, but can also be handled automatically
2116 via <glossterm linkend="glossary-autovacuum">autovacuum</glossterm> processes.
2117 </para>
2118 <para>
2119 For more information, see
2120 <xref linkend="routine-vacuuming"/> .
2121 </para>
2122 </glossdef>
2123 </glossentry>
2124
2125 <glossentry id="glossary-view">
2126 <glossterm>View</glossterm>
2127 <glossdef>
2128 <para>
2129 A <glossterm linkend="glossary-relation">relation</glossterm> that is defined by a
2130 <command>SELECT</command> statement, but has no storage of its own.
2131 Any time a query references a view, the definition of the view is
2132 substituted into the query as if the user had typed it as a subquery
2133 instead of the name of the view.
2134 </para>
2135 <para>
2136 For more information, see
2137 <xref linkend="sql-createview"/>.
2138 </para>
2139 </glossdef>
2140 </glossentry>
2141
2142 <glossentry id="glossary-vm">
2143 <glossterm>Visibility map (fork)</glossterm>
2144 <glossdef>
2145 <para>
2146 A storage structure that keeps metadata about each data page
2147 of a table's main fork. The visibility map entry for
2148 each page stores two bits: the first one
2149 (<literal>all-visible</literal>) indicates that all tuples
2150 in the page are visible to all transactions. The second one
2151 (<literal>all-frozen</literal>) indicates that all tuples
2152 in the page are marked frozen.
2153 </para>
2154 </glossdef>
2155 </glossentry>
2156
2157 <glossentry>
2158 <glossterm>WAL</glossterm>
2159 <glosssee otherterm="glossary-wal" />
2160 </glossentry>
2161
2162 <glossentry id="glossary-wal-archiver">
2163 <glossterm>WAL archiver (process)</glossterm>
2164 <glossdef>
2165 <para>
2166 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
2167 which, if enabled, saves copies of
2168 <glossterm linkend="glossary-wal-file">WAL files</glossterm>
2169 for the purpose of creating backups or keeping
2170 <glossterm linkend="glossary-replica">replicas</glossterm> current.
2171 </para>
2172 <para>
2173 For more information, see
2174 <xref linkend="continuous-archiving"/>.
2175 </para>
2176 </glossdef>
2177 </glossentry>
2178
2179 <glossentry id="glossary-wal-file">
2180 <glossterm>WAL file</glossterm>
2181 <glossdef>
2182 <para>
2183 Also known as <firstterm>WAL segment</firstterm> or
2184 <firstterm>WAL segment file</firstterm>.
2185 Each of the sequentially-numbered files that provide storage space for
2186 <glossterm linkend="glossary-wal">WAL</glossterm>.
2187 The files are all of the same predefined size
2188 and are written in sequential order, interspersing changes
2189 as they occur in multiple simultaneous sessions.
2190 If the system crashes, the files are read in order, and each of the
2191 changes is replayed to restore the system to the state it was in
2192 before the crash.
2193 </para>
2194 <para>
2195 Each WAL file can be released after a
2196 <glossterm linkend="glossary-checkpoint">checkpoint</glossterm>
2197 writes all the changes in it to the corresponding data files.
2198 Releasing the file can be done either by deleting it, or by changing its
2199 name so that it will be used in the future, which is called
2200 <firstterm>recycling</firstterm>.
2201 </para>
2202 <para>
2203 For more information, see
2204 <xref linkend="wal-internals"/>.
2205 </para>
2206 </glossdef>
2207 </glossentry>
2208
2209 <glossentry id="glossary-wal-record">
2210 <glossterm>WAL record</glossterm>
2211 <glossdef>
2212 <para>
2213 A low-level description of an individual data change.
2214 It contains sufficient information for the data change to be
2215 re-executed (<firstterm>replayed</firstterm>) in case a system failure
2216 causes the change to be lost.
2217 WAL records use a non-printable binary format.
2218 </para>
2219 <para>
2220 For more information, see
2221 <xref linkend="wal-internals"/>.
2222 </para>
2223 </glossdef>
2224 </glossentry>
2225
2226 <glossentry id="glossary-wal-receiver">
2227 <glossterm>WAL receiver (process)</glossterm>
2228 <glossdef>
2229 <para>
2230 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
2231 that runs on a <glossterm linkend="glossary-replica">replica</glossterm>
2232 to receive WAL from the
2233 <glossterm linkend="glossary-primary-server">primary server</glossterm>
2234 for replay by the
2235 <glossterm linkend="glossary-startup-process">startup process</glossterm>.
2236 </para>
2237
2238 <para>
2239 For more information, see
2240 <xref linkend="warm-standby"/>.
2241 </para>
2242 </glossdef>
2243 </glossentry>
2244
2245 <glossentry>
2246 <glossterm>WAL segment</glossterm>
2247 <glosssee otherterm="glossary-wal-file" />
2248 </glossentry>
2249
2250 <glossentry id="glossary-wal-sender">
2251 <glossterm>WAL sender (process)</glossterm>
2252 <glossdef>
2253 <para>
2254 A special <glossterm linkend="glossary-backend">backend process</glossterm>
2255 that streams WAL over a network. The receiving end can be a
2256 <glossterm linkend="glossary-wal-receiver">WAL receiver</glossterm>
2257 in a <glossterm linkend="glossary-replica">replica</glossterm>,
2258 <xref linkend="app-pgreceivewal"/>, or any other client program
2259 that speaks the replication protocol.
2260 </para>
2261 </glossdef>
2262 </glossentry>
2263
2264 <glossentry id="glossary-wal-summarizer">
2265 <glossterm>WAL summarizer (process)</glossterm>
2266 <glossdef>
2267 <para>
2268 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
2269 that summarizes WAL data for
2270 <glossterm linkend="glossary-incremental-backup">incremental backups</glossterm>.
2271 </para>
2272 <para>
2273 For more information, see <xref linkend="runtime-config-wal-summarization"/>.
2274 </para>
2275 </glossdef>
2276 </glossentry>
2277
2278 <glossentry id="glossary-wal-writer">
2279 <glossterm>WAL writer (process)</glossterm>
2280 <glossdef>
2281 <para>
2282 An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
2283 that writes <glossterm linkend="glossary-wal-record">WAL records</glossterm>
2284 from <glossterm linkend="glossary-shared-memory">shared memory</glossterm> to
2285 <glossterm linkend="glossary-wal-file">WAL files</glossterm>.
2286 </para>
2287 <para>
2288 For more information, see
2289 <xref linkend="runtime-config-wal"/>.
2290 </para>
2291 </glossdef>
2292 </glossentry>
2293
2294 <glossentry id="glossary-window-function">
2295 <glossterm>Window function (routine)</glossterm>
2296 <glossdef>
2297 <para>
2298 A type of <glossterm linkend="glossary-function">function</glossterm>
2299 used in a <glossterm linkend="glossary-query">query</glossterm>
2300 that applies to a <glossterm linkend="glossary-partition">partition</glossterm>
2301 of the query's <glossterm linkend="glossary-result-set">result set</glossterm>;
2302 the function's result is based on values found in
2303 <glossterm linkend="glossary-tuple">rows</glossterm> of the same partition or frame.
2304 </para>
2305 <para>
2306 All <glossterm linkend="glossary-aggregate">aggregate functions</glossterm>
2307 can be used as window functions, but window functions can also be
2308 used to, for example, give ranks to each of the rows in the partition.
2309 Also known as <firstterm>analytic functions</firstterm>.
2310 </para>
2311 <para>
2312 For more information, see
2313 <xref linkend="tutorial-window"/>.
2314 </para>
2315 </glossdef>
2316 </glossentry>
2317
2318 <glossentry id="glossary-wal">
2319 <glossterm>Write-ahead log</glossterm>
2320 <glossdef>
2321 <para>
2322 The journal that keeps track of the changes in the
2323 <glossterm linkend="glossary-db-cluster">database cluster</glossterm>
2324 as user- and system-invoked operations take place.
2325 It comprises many individual
2326 <glossterm linkend="glossary-wal-record">WAL records</glossterm> written
2327 sequentially to <glossterm linkend="glossary-wal-file">WAL files</glossterm>.
2328 </para>
2329 </glossdef>
2330 </glossentry>
2331 </glosslist>
2332 </appendix>