Changeset 747
- Timestamp:
- 10/11/07 09:30:08 (1 year ago)
- Files:
-
- traduc/trunk/manuel/backup.xml (modified) (2 diffs)
- traduc/trunk/manuel/catalogs.xml (modified) (7 diffs)
- traduc/trunk/manuel/config.xml (modified) (5 diffs)
- traduc/trunk/manuel/datatype.xml (modified) (1 diff)
- traduc/trunk/manuel/dfunc.xml (modified) (2 diffs)
- traduc/trunk/manuel/dml.xml (modified) (2 diffs)
- traduc/trunk/manuel/docguide.xml (modified) (1 diff)
- traduc/trunk/manuel/features.xml (modified) (1 diff)
- traduc/trunk/manuel/func.xml (modified) (3 diffs)
- traduc/trunk/manuel/indices.xml (modified) (1 diff)
- traduc/trunk/manuel/installation.xml (modified) (2 diffs)
- traduc/trunk/manuel/libpq.xml (modified) (1 diff)
- traduc/trunk/manuel/mvcc.xml (modified) (2 diffs)
- traduc/trunk/manuel/plpgsql.xml (modified) (6 diffs)
- traduc/trunk/manuel/queries.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/alter_language.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/analyze.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/create_cast.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/create_constraint.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/create_opclass.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/create_sequence.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/deallocate.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/delete.xml (modified) (1 diff)
- traduc/trunk/manuel/ref/psql-ref.xml (modified) (1 diff)
- traduc/trunk/manuel/spi.xml (modified) (1 diff)
- traduc/trunk/manuel/textsearch.xml (modified) (5 diffs)
- traduc/trunk/manuel/typeconv.xml (modified) (2 diffs)
- traduc/trunk/manuel/version.xml (modified) (1 diff)
- traduc/trunk/manuel/wal.xml (modified) (3 diffs)
Legend:
- Unmodified
- Added
- Removed
- Modified
- Copied
- Moved
traduc/trunk/manuel/backup.xml
r737 r747 1562 1562 <para> 1563 1563 A working example of a waiting <varname>restore_command</varname> is provided 1564 as a contrib module, named <application>pg_standby</ varname>. This can be1564 as a contrib module, named <application>pg_standby</application>. This can be 1565 1565 extended as needed to support specific configurations or environments. 1566 1566 </para> … … 1819 1819 Ce programme inspecte le fichier de contrôle pour déterminer 1820 1820 l'emplacement du point de contrôle WAL actuel, or by using the 1821 <varname>log_restartpoints</ > option to print values to the server log.1821 <varname>log_restartpoints</varname> option to print values to the server log. 1822 1822 </para> 1823 1823 </sect2> traduc/trunk/manuel/catalogs.xml
r735 r747 2054 2054 <entry><structfield>conpfeqop</structfield></entry> 2055 2055 <entry><type>oid[]</type></entry> 2056 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</ ></entry>2056 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</literal></entry> 2057 2057 <entry>If a foreign key, list of the equality operators for PK = FK comparisons</entry> 2058 2058 </row> … … 2061 2061 <entry><structfield>conppeqop</structfield></entry> 2062 2062 <entry><type>oid[]</type></entry> 2063 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</ ></entry>2063 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</literal></entry> 2064 2064 <entry>If a foreign key, list of the equality operators for PK = PK comparisons</entry> 2065 2065 </row> … … 2068 2068 <entry><structfield>conffeqop</structfield></entry> 2069 2069 <entry><type>oid[]</type></entry> 2070 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</ ></entry>2070 <entry><literal><link linkend="catalog-pg-operator"><structname>pg_operator</structname></link>.oid</literal></entry> 2071 2071 <entry>If a foreign key, list of the equality operators for FK = FK comparisons</entry> 2072 2072 </row> … … 2149 2149 <entry><type>oid</type></entry> 2150 2150 <entry><literal><link linkend="catalog-pg-type"><structname>pg_type</structname></link>.oid</literal></entry> 2151 <entry>The OID of the <structname>pg_type</ > entry owning this enum value</entry>2151 <entry>The OID of the <structname>pg_type</structname> entry owning this enum value</entry> 2152 2152 </row> 2153 2153 … … 3783 3783 <entry><type>float4</type></entry> 3784 3784 <entry></entry> 3785 <entry>Estimated number of result rows (zero if not <structfield>proretset</ >)</entry>3785 <entry>Estimated number of result rows (zero if not <structfield>proretset</structfield>)</entry> 3786 3786 </row> 3787 3787 … … 4627 4627 <entry></entry> 4628 4628 <entry> 4629 Controls in which <xref linkend="guc-session-replication-role" > modes4629 Controls in which <xref linkend="guc-session-replication-role"/> modes 4630 4630 the trigger fires. 4631 4631 <literal>O</literal> = trigger fires in <quote>origin</quote> and <quote>local</quote> modes, … … 4893 4893 parameters supported by the template. This division of labor allows 4894 4894 dictionaries to be created by unprivileged users. The parameters 4895 are specified by a text string <structfield>dictinitoption</ >,4895 are specified by a text string <structfield>dictinitoption</structfield>, 4896 4896 whose format and meaning vary depending on the template. 4897 4897 </para> traduc/trunk/manuel/config.xml
r735 r747 1368 1368 <term><varname>bgwriter_lru_multiplier</varname> (<type>floating point</type>)</term> 1369 1369 <indexterm> 1370 <primary><varname>bgwriter_lru_multiplier</ > configuration parameter</primary>1370 <primary><varname>bgwriter_lru_multiplier</varname> configuration parameter</primary> 1371 1371 </indexterm> 1372 1372 <listitem> 1373 1373 <para> 1374 Unless limited by <varname>bgwriter_lru_maxpages</ >, the number1374 Unless limited by <varname>bgwriter_lru_maxpages</varname>, the number 1375 1375 of dirty buffers written in each round is determined by reference 1376 1376 to the number of new buffers that have been needed by server 1377 1377 processes during recent rounds. This number is multiplied by 1378 <varname>bgwriter_lru_multiplier</ > to arrive at the estimate1378 <varname>bgwriter_lru_multiplier</varname> to arrive at the estimate 1379 1379 of the number of buffers that will be needed during the next round. 1380 Thus, a setting of 1.0 represents a <quote>just in time</ > policy1380 Thus, a setting of 1.0 represents a <quote>just in time</quote> policy 1381 1381 of writing exactly the number of buffers predicted to be needed. 1382 1382 Larger values provide some cushion against spikes in demand, … … 1384 1384 server processes. 1385 1385 The default is 2.0. 1386 This parameter can only be set in the <filename>postgresql.conf</ >1386 This parameter can only be set in the <filename>postgresql.conf</filename> 1387 1387 file or on the server command line. 1388 1388 </para> … … 1461 1461 1462 1462 <para> 1463 In many situations, turning off <xref linkend="guc-synchronous-commit" >1463 In many situations, turning off <xref linkend="guc-synchronous-commit"/> 1464 1464 for noncritical transactions can provide much of the potential 1465 1465 performance benefit of turning off <varname>fsync</varname>, without … … 1784 1784 <title>Archivage</title> 1785 1785 1786 <variablelist>1787 1786 <variablelist> 1788 1787 <varlistentry id="guc-archive-mode" xreflabel="archive_mode"> … … 1822 1821 Ce paramètre ne peut qu'être configuré dans le fichier 1823 1822 <filename>postgresql.conf</filename> ou indiqué sur la ligne de commande. It is ignored unless 1824 <varname>archive_mode</ > was enabled at server start.1825 If <varname>archive_command</ > is an empty string (the default) while1826 <varname>archive_mode</ > is enabled, then WAL archiving is temporarily1823 <varname>archive_mode</varname> was enabled at server start. 1824 If <varname>archive_command</varname> is an empty string (the default) while 1825 <varname>archive_mode</varname> is enabled, then WAL archiving is temporarily 1827 1826 disabled, but the server continues to accumulate WAL segment files in 1828 1827 the expectation that a command will soon be provided. traduc/trunk/manuel/datatype.xml
r735 r747 2423 2423 <productname>PostgreSQL</productname> ; 2424 2424 (These rules are also used to choose the default value of 2425 <xref linkend="guc-log-timezone" >, if it is not specified.)2425 <xref linkend="guc-log-timezone"/>, if it is not specified.) 2426 2426 </para> 2427 2427 </listitem> traduc/trunk/manuel/dfunc.xml
r735 r747 104 104 commutateur de l'éditeur de liens pour les bibliothèques partagées est 105 105 <option>-b</option>. 106 Ainsi&nb psp;:106 Ainsi : 107 107 <programlisting>cc +z -c foo.c 108 108 </programlisting> 109 ou&nb psp;:110 <programlisting>gcc -fpic -c foo.c 111 </programlisting> 112 puis&nb psp;:109 ou : 110 <programlisting>gcc -fpic -c foo.c 111 </programlisting> 112 puis : 113 113 <programlisting>ld -b -o foo.sl foo.o 114 114 </programlisting> … … 127 127 <acronym>PIC</acronym> est l'option par défaut. Aucune option de compilation 128 128 particulière n'est nécessaire. Le commutateur de l'éditeur de liens pour 129 produire des bibliothèques partagées est <option>-shared</option>&nb psp;:129 produire des bibliothèques partagées est <option>-shared</option> : 130 130 <programlisting>cc -c foo.c 131 131 ld -shared -o foo.so foo.o traduc/trunk/manuel/dml.xml
r735 r747 80 80 Si les valeurs de certaines colonnes ne sont pas connues, 81 81 elles peuvent être omises. Dans ce cas, elles sont remplies avec leur 82 valeur par défaut. Par exemple&nb psp;:82 valeur par défaut. Par exemple : 83 83 <programlisting>INSERT INTO produits (no_produit, nom) VALUES (1, 'Fromage'); 84 84 INSERT INTO produits VALUES (1, 'Fromage');</programlisting> … … 267 267 268 268 <para> 269 En indiquant simplement&nb psp;:269 En indiquant simplement : 270 270 <programlisting>DELETE FROM produits;</programlisting> 271 271 on supprime toutes les lignes de la table. Attention aux mauvaises traduc/trunk/manuel/docguide.xml
r735 r747 568 568 To create a proper index, the build might process several identical 569 569 stages. If you do not care about the index, and just want to 570 proof-read the output, use <literal>draft</ >:570 proof-read the output, use <literal>draft</literal>: 571 571 <screen> 572 572 <prompt>doc/src/sgml$ </prompt><userinput>gmake draft</userinput> traduc/trunk/manuel/features.xml
r735 r747 74 74 <listitem><para>ISO/IEC 9075-13 Routines and Types using the Java Language (SQL/JRT)</para></listitem> 75 75 <listitem><para>ISO/IEC 9075-14 XML-related specifications (SQL/XML)</para></listitem> 76 </itemizedlist> 76 77 </para> 77 78 traduc/trunk/manuel/func.xml
r746 r747 3479 3479 3480 3480 <para> 3481 The <function>regexp_split_to_table< function/> function splits a string using a POSIX3481 The <function>regexp_split_to_table</function> function splits a string using a POSIX 3482 3482 regular expression pattern as a delimiter. It has the syntax 3483 3483 <function>regexp_split_to_table</function>(<replaceable>string</replaceable>, <replaceable>pattern</replaceable> … … 3733 3733 antislash, il faut écrire deux antislashs dans l'instruction, en supposant 3734 3734 que la syntaxe de chaîne d'échappement est utilisée (voir <xref 3735 linkend="sql-syntax-strings" >).3735 linkend="sql-syntax-strings"/>). 3736 3736 </para> 3737 3737 </note> … … 8227 8227 given as the second trigger argument. It must be schema-qualified as 8228 8228 shown above, so that the trigger behavior will not change with changes 8229 in <varname>search_path</ function>. For8229 in <varname>search_path</varname>. For 8230 8230 <function>tsvector_update_trigger_column</function>, the second trigger argument 8231 8231 is the name of another table column, which must be of type traduc/trunk/manuel/indices.xml
r735 r747 264 264 Voir <xref linkend="functions-array"/> pour la signification de ces 265 265 opérateurs. 266 GIN indexes cannot use <literal>IS NULL</ > as a search condition.266 GIN indexes cannot use <literal>IS NULL</literal> as a search condition. 267 267 Les autres classes d'opérateurs GIN sont disponibles dans les modules 268 268 <literal>contrib</literal> <literal>tsearch2</literal> et traduc/trunk/manuel/installation.xml
r735 r747 427 427 account (which you already have if you are upgrading). 428 428 <programlisting> 429 <userinput>/usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data</ >429 <userinput>/usr/local/pgsql/bin/initdb -D /usr/local/pgsql/data</userinput> 430 430 </programlisting> 431 431 </para> … … 453 453 Finally, restore your data from backup with 454 454 <screen> 455 <userinput>/usr/local/pgsql/bin/psql -d postgres -f <replaceable>outputfile</ ></userinput>455 <userinput>/usr/local/pgsql/bin/psql -d postgres -f <replaceable>outputfile</replaceable></userinput> 456 456 </screen> 457 457 using the <emphasis>new</emphasis> <application>psql</application>. traduc/trunk/manuel/libpq.xml
r744 r747 287 287 Kerberos 5 et GSSAPI. Il doit correspondre avec le nom du service spécifié dans 288 288 la configuration du serveur pour que l'authentification Kerberos puisse 289 réussir (voir aussi la <xref linkend="kerberos-auth"/> and <xref linkend="gssapi-auth" >.)289 réussir (voir aussi la <xref linkend="kerberos-auth"/> and <xref linkend="gssapi-auth"/>.) 290 290 </para> 291 291 </listitem> traduc/trunk/manuel/mvcc.xml
r735 r747 756 756 </para> 757 757 758 <!-- 758 759 <table tocentry="1" id="table-lock-compatibility"> 759 760 <title>Modes de verrou conflictuels</title> … … 869 870 </tbody> 870 871 </tgroup> 871 </table> 872 </table>--> 872 873 </sect2> 873 874 traduc/trunk/manuel/plpgsql.xml
r735 r747 682 682 what happens behind the scenes is 683 683 <programlisting> 684 PREPARE <replaceable>statement_name</ >(integer, integer) AS SELECT $1 < $2;684 PREPARE <replaceable>statement_name</replaceable>(integer, integer) AS SELECT $1 < $2; 685 685 </programlisting> 686 686 and then this prepared statement is <command>EXECUTE</command>d for each … … 757 757 puis la valeur actuelle de la variable est fournie comme valeur du 758 758 paramètre à l'exécution. This is exactly like the processing described earlier 759 for expressions; for details see <xref linkend="plpgsql-var-subst" >.759 for expressions; for details see <xref linkend="plpgsql-var-subst"/>. 760 760 Par exemple, si vous écrivez 761 761 <programlisting>DECLARE … … 812 812 <literal>FOUND</literal> est configurée à true si la requête a produit 813 813 au moins une ligne, false dans le cas contraire (voir 814 <xref linkend="plpgsql-statements-diagnostics" >).814 <xref linkend="plpgsql-statements-diagnostics"/>). 815 815 </para> 816 816 … … 975 975 <application>PL/pgSQL</application> pour garder en cache les planifications des 976 976 commandes (as discussed in 977 <xref linkend="plpgsql-plan-caching" >) ne fonctionneront pas dans de tels scénarios. Pour gérer ce type977 <xref linkend="plpgsql-plan-caching"/>) ne fonctionneront pas dans de tels scénarios. Pour gérer ce type 978 978 de problème, l'instruction <command>EXECUTE</command> est fournie : 979 979 … … 3027 3027 The <command>INSERT</command> statement will effectively be processed as 3028 3028 <programlisting> 3029 PREPARE <replaceable>statement_name</ >(text, timestamp) AS3029 PREPARE <replaceable>statement_name</replaceable>(text, timestamp) AS 3030 3030 INSERT INTO logtable VALUES ($1, $2); 3031 3031 </programlisting> … … 3082 3082 SELECT $1 FROM table WHERE key = $2 3083 3083 </programlisting> 3084 and thus it was just an expensive way of assigning <literal>val</ >'s3084 and thus it was just an expensive way of assigning <literal>val</literal>'s 3085 3085 current value back to itself for each row. 3086 3086 </para> traduc/trunk/manuel/queries.xml
r735 r747 1188 1188 ORDER BY <replaceable>expression_tri1</replaceable> <optional>ASC | DESC</optional> <optional>NULLS { FIRST | LAST }</optional> 1189 1189 <optional>, <replaceable>expression_tri2</replaceable> <optional>ASC | DESC</optional> <optional>NULLS { FIRST | LAST }</optional> ...</optional> 1190 </synopsis> 1190 1191 1191 1192 Les expressions de tri peuvent être toute expression qui serait valide dans traduc/trunk/manuel/ref/alter_language.xml
r740 r747 19 19 <refsynopsisdiv> 20 20 <synopsis>ALTER LANGUAGE <replaceable>nom</replaceable> RENAME TO <replaceable>nouveau_nom</replaceable> 21 </synopsis> 21 22 <synopsis>ALTER LANGUAGE <replaceable>nom</replaceable> OWNER TO <replaceable>nouveau_proprietaire</replaceable> 22 23 </synopsis> traduc/trunk/manuel/ref/analyze.xml
r740 r747 131 131 qu'<command>ANALYZE</command> est lancé, même si le contenu réel de la table 132 132 n'a pas changé. Cela peut résulter en de petites modifications dans les 133 coûts estimés par l'optimiseur affichés par <xref linkend="sql-explain" endterm="sql-explain-title" >. Dans133 coûts estimés par l'optimiseur affichés par <xref linkend="sql-explain" endterm="sql-explain-title"/>. Dans 134 134 de rares situations, ce non-déterminisme entraîne le choix par l'optimiseur 135 135 d'un plan de requête différent entre deux lancements traduc/trunk/manuel/ref/create_cast.xml
r740 r747 99 99 between the above interpretation and the alternative of casting the 100 100 <type>numeric</type> constant to <type>integer</type> and applying the 101 <type>integer</type> <literal>+</ type> <type>integer</type> operator. Lacking any101 <type>integer</type> <literal>+</literal> <type>integer</type> operator. Lacking any 102 102 knowledge of which choice to prefer, it would give up and declare the 103 103 query ambiguous. The fact that only one of the two casts is traduc/trunk/manuel/ref/create_constraint.xml
r740 r747 151 151 152 152 <simplelist type="inline"> 153 <member><xref linkend="sql-createtrigger" endterm="sql-createtrigger-title" ></member>154 <member><xref linkend="sql-droptrigger" endterm="sql-droptrigger-title" ></member>155 <member><xref linkend="sql-set-constraints" endterm="sql-set-constraints-title" ></member>153 <member><xref linkend="sql-createtrigger" endterm="sql-createtrigger-title"/></member> 154 <member><xref linkend="sql-droptrigger" endterm="sql-droptrigger-title"/></member> 155 <member><xref linkend="sql-set-constraints" endterm="sql-set-constraints-title"/></member> 156 156 </simplelist> 157 157 </refsect1> traduc/trunk/manuel/ref/create_opclass.xml
r740 r747 303 303 <member><xref linkend="sql-alteropclass" endterm="sql-alteropclass-title"/></member> 304 304 <member><xref linkend="sql-dropopclass" endterm="sql-dropopclass-title"/></member> 305 <member><xref linkend="sql-createopfamily" endterm="sql-createopfamily-title" ></member>306 <member><xref linkend="sql-alteropfamily" endterm="sql-alteropfamily-title" ></member>305 <member><xref linkend="sql-createopfamily" endterm="sql-createopfamily-title"/></member> 306 <member><xref linkend="sql-alteropfamily" endterm="sql-alteropfamily-title"/></member> 307 307 </simplelist> 308 308 </refsect1> traduc/trunk/manuel/ref/create_sequence.xml
r740 r747 305 305 102 306 306 </programlisting> 307 </para> 307 308 308 309 <para> traduc/trunk/manuel/ref/deallocate.xml
r740 r747 67 67 </varlistentry> 68 68 69 <variablelist>70 69 <varlistentry> 71 70 <term><literal>ALL</literal></term> traduc/trunk/manuel/ref/delete.xml
r740 r747 137 137 Une expression retournant une valeur de type 138 138 <type>boolean</type>. 139 Only rows for which this expression returns <literal>true</ >139 Only rows for which this expression returns <literal>true</literal> 140 140 will be deleted. 141 141 </para> traduc/trunk/manuel/ref/psql-ref.xml
r740 r747 2084 2084 <literal>(<replaceable class="parameter">R</replaceable>|)</literal> pour 2085 2085 <literal><replaceable class="parameter">R</replaceable>?</literal>. 2086 <literal>$</ > is not needed as a regular-expression character since2086 <literal>$</literal> is not needed as a regular-expression character since 2087 2087 the pattern must match the whole name, unlike the usual 2088 2088 interpretation of regular expressions (in other words, <literal>$</literal> traduc/trunk/manuel/spi.xml
r735 r747 825 825 of the planner's <quote>cursor options</quote> parameter. This is a bitmask 826 826 having the values shown in <filename>nodes/parsenodes.h</filename> 827 for the <structfield>options</structfield> field of <structname>DeclareCursorStmt</struct field>.827 for the <structfield>options</structfield> field of <structname>DeclareCursorStmt</structname>. 828 828 <function>SPI_prepare</function> always takes these options as zero. 829 829 </para> traduc/trunk/manuel/textsearch.xml
r735 r747 556 556 557 557 <para> 558 Because <function>to_tsvector</function>(< LITERAL>NULL</LITERAL>) can559 return < LITERAL>NULL</LITERAL>, it is recommended to use558 Because <function>to_tsvector</function>(<literal>NULL</literal>) can 559 return <literal>NULL</literal>, it is recommended to use 560 560 <function>coalesce</function>. Here is the safe method for creating a 561 561 <type>tsvector</type> from a structured document: … … 2403 2403 <listitem> 2404 2404 <para> 2405 The < LITERAL>MAXLEN</literal> parameter specifies the maximum length of the2405 The <literal>MAXLEN</literal> parameter specifies the maximum length of the 2406 2406 number considered as a 'good' integer. The default value is 6. 2407 2407 </para> … … 2410 2410 <listitem> 2411 2411 <para> 2412 The < LITERAL>REJECTLONG</LITERAL> parameter specifies if a 'long' integer2412 The <literal>REJECTLONG</literal> parameter specifies if a 'long' integer 2413 2413 should be indexed or treated as a stop word. If 2414 <literal>REJECTLONG</literal>=< LITERAL>FALSE</LITERAL> (default),2414 <literal>REJECTLONG</literal>=<literal>FALSE</literal> (default), 2415 2415 the dictionary returns the prefixed part of the integer with length 2416 < LITERAL>MAXLEN</literal>. If2417 < LITERAL>REJECTLONG</LITERAL>=<LITERAL>TRUE</LITERAL>, the dictionary2416 <literal>MAXLEN</literal>. If 2417 <literal>REJECTLONG</literal>=<literal>TRUE</literal>, the dictionary 2418 2418 considers a long integer as a stop word. 2419 2419 </para> … … 2475 2475 <programlisting> 2476 2476 $ make && make install 2477 $ psql DBNAME <dict_intdict.sql2477 $ psql DBNAME < dict_intdict.sql 2478 2478 </programlisting> 2479 2479 </para> … … 2770 2770 $ make 2771 2771 $ make install 2772 $ psql regression <test_parser.sql2772 $ psql regression < test_parser.sql 2773 2773 </programlisting> 2774 2774 </para> traduc/trunk/manuel/typeconv.xml
r735 r747 156 156 <firstterm>casts</firstterm> en anglais. Des conversions de types 157 157 supplémentaires peuvent être ajoutées par l'utilisateur avec la commande 158 <xref linkend="sql-createcast" endterm="sql-createcast-title" > (c'est158 <xref linkend="sql-createcast" endterm="sql-createcast-title"/> (c'est 159 159 habituellement réalisé en 160 160 conjonction avec la définition de nouveaux types de données. … … 372 372 Quelques exemples suivent. 373 373 </para> 374 375 <example>376 <title>Résolution de types pour l'opérateur factoriel</title>377 374 378 375 <example> traduc/trunk/manuel/version.xml
r708 r747 1 <!ENTITY version "8. 2.5">2 <!ENTITY majorversion "8. 2">1 <!ENTITY version "8.3.0-beta1"> 2 <!ENTITY majorversion "8.3"> traduc/trunk/manuel/wal.xml
r735 r747 375 375 writing dirty buffers during a checkpoint is spread over a period of time. 376 376 That period is controlled by 377 <xref linkend="guc-checkpoint-completion-target" >, which is377 <xref linkend="guc-checkpoint-completion-target"/>, which is 378 378 given as a fraction of the checkpoint interval. 379 379 The I/O rate is adjusted so that the checkpoint finishes when the … … 382 382 <varname>checkpoint_timeout</varname> seconds have elapsed, 383 383 whichever is sooner. With the default value of 0.5, 384 <productname>PostgreSQL</ > can be expected to complete each checkpoint384 <productname>PostgreSQL</productname> can be expected to complete each checkpoint 385 385 in about half the time before the next checkpoint starts. On a system 386 386 that's very close to maximum I/O throughput during normal operation, … … 397 397 </para> 398 398 399 <para>400 There will always be at least one WAL segment file, and will normally401 not be more than (2 + <varname>checkpoint_completion_target</varname>) * <varname>checkpoint_segments</varname> + 1402 files. Each segment file is normally 16 MB (though this size can be403 399 <para> 404 400 Il y aura toujours au moins un fichier segment WAL et normalement

