@@ -332,7 +332,7 @@ NET STOP postgresql-9.0
332332 <para>
333333 Also, if upgrading standby servers, change <varname>wal_level</>
334334 to <literal>replica</> in the <filename>postgresql.conf</> file on
335- the new master cluster.
335+ the new primary cluster.
336336 </para>
337337 </step>
338338
@@ -425,8 +425,8 @@ pg_upgrade.exe
425425 linkend="streaming-replication">) or Log-Shipping (see <xref
426426 linkend="warm-standby">) standby servers, follow these steps to
427427 upgrade them. You will not be running <application>pg_upgrade</>
428- on the standby servers, but rather <application>rsync</>. Do not
429- start any servers yet.
428+ on the standby servers, but rather <application>rsync</> on the
429+ primary. Do not start any servers yet.
430430 </para>
431431
432432 <substeps>
@@ -455,7 +455,7 @@ pg_upgrade.exe
455455
456456 <para>
457457 Install the same custom shared object files on the new standbys
458- that you installed in the new master cluster.
458+ that you installed in the new primary cluster.
459459 </para>
460460 </step>
461461
@@ -482,25 +482,33 @@ pg_upgrade.exe
482482 <title>Run <application>rsync</></title>
483483
484484 <para>
485- From a directory that is above the old and new database cluster
486- directories, run this for each slave:
485+ From a directory on the primary server that is above the old and
486+ new database cluster directories, run this on the
487+ <emphasis>primary</> for each standby server:
487488
488489<programlisting>
489490rsync --archive --delete --hard-links --size-only old_pgdata new_pgdata remote_dir
490491</programlisting>
491492
492493 where <option>old_pgdata</> and <option>new_pgdata</> are relative
493- to the current directory, and <option>remote_dir</> is
494- <emphasis>above</> the old and new cluster directories on
495- the standby server . The old and new relative cluster paths
496- must match on the master and standby server. Consult the
494+ to the current directory on the primary , and <option>remote_dir</>
495+ is <emphasis>above</> the old and new cluster directories on
496+ the standby. The old and new relative cluster paths
497+ must match on the primary and standby server. Consult the
497498 <application>rsync</> manual page for details on specifying the
498499 remote directory, e.g. <literal>standbyhost:/opt/PostgreSQL/</>.
499- <application>rsync</> will be fast when <application>pg_upgrade</>'s
500- <option>--link</> mode is used because it will create hard links
501- on the remote server rather than transferring user data.
502- Unfortunately, <application>rsync</> needlessly copies the
503- files associated with temporary and unlogged tables.
500+ </para>
501+
502+ <para>
503+ What <application>rsync</> does is to copy files from the
504+ primary to the standby, and, if <application>pg_upgrade</>'s
505+ <option>--link</> mode was used, link files from the old to
506+ new clusters on the standby. It links the same files that
507+ <application>pg_upgrade</> linked in the primary old and new
508+ clusters. (Of course, linking speeds up <application>rsync</>.)
509+ Unfortunately, <application>rsync</> needlessly copies files
510+ associated with temporary and unlogged tables because these files
511+ don't normally exist on standby servers.
504512 </para>
505513
506514 <para>
@@ -517,8 +525,8 @@ rsync --archive --delete --hard-links --size-only old_pgdata new_pgdata remote_d
517525 <para>
518526 Configure the servers for log shipping. (You do not need to run
519527 <function>pg_start_backup()</> and <function>pg_stop_backup()</>
520- or take a file system backup as the slaves are still synchronized
521- with the master .)
528+ or take a file system backup as the standbys are still synchronized
529+ with the primary .)
522530 </para>
523531 </step>
524532
0 commit comments