On Fri, 2008-03-28 at 00:54 -0400, Tom Lane wrote:
> Greg Smith <[EMAIL PROTECTED]> writes:
> > ... That was a small change in a utility that should never be run on a
> > production system. You're trying to get a change made to the code path
> > people rely on for their *backups*. Good luck with that.
>
> While I quite agree with Greg's comments about not changing stable
> release branches unnecessarily, it seems that there's another
> consideration in this case. If we don't back-patch %r then users
> will have to rely on hacky scripts like the one posted upthread.
> Is that really a net gain in reliability?
>
> (I'm honestly not sure of the answer; I'm just thinking it might
> be open to debate. In particular I don't remember how complicated
> the patch to add %r was.)
Here's the original patch, edited to remove pg_standby changes.
I've not even checked whether it will apply, but it seems fairly simple.
Gurjeet, would you like to freshen and test that up for apply to 8.2?
--
Simon Riggs
2ndQuadrant http://www.2ndQuadrant.com
PostgreSQL UK 2008 Conference: http://www.postgresql.org.uk
Index: doc/src/sgml/backup.sgml
===================================================================
RCS file: /projects/cvsroot/pgsql/doc/src/sgml/backup.sgml,v
retrieving revision 2.97
diff -c -r2.97 backup.sgml
*** doc/src/sgml/backup.sgml 1 Feb 2007 00:28:16 -0000 2.97
--- doc/src/sgml/backup.sgml 9 Feb 2007 23:05:26 -0000
***************
*** 526,531 ****
--- 526,536 ----
archive, while any <literal>%f</> is replaced by the file name only.
(The path name is relative to the working directory of the server,
i.e., the cluster's data directory.)
+ A <literal>%r</> will be replaced by the name of the file containing
+ the last valid restartpoint. This is the earliest file that must be
+ kept to allow a restore to be restartable, so this information can
+ allow if the archive to be truncated to just the minimum required
+ to support restart of the current restore.
Write <literal>%%</> if you need to embed an actual <literal>%</>
character in the command. The simplest useful command is something
like:
***************
*** 922,927 ****
--- 927,937 ----
which is replaced by the path name to copy the log file to.
(The path name is relative to the working directory of the server,
i.e., the cluster's data directory.)
+ A <literal>%r</> will be replaced by the name of the file containing
+ the last valid restartpoint. This is the earliest file that must be
+ kept to allow a restore to be restartable, so this information can
+ allow if the archive to be truncated to just the minimum required
+ to support restart of the current restore.
Write <literal>%%</> if you need to embed an actual <literal>%</>
character in the command. The simplest useful command is
something like:
***************
*** 1008,1020 ****
<listitem>
<para>
The shell command to execute to retrieve an archived segment of
! the WAL file series. This parameter is required.
! Any <literal>%f</> in the string is
! replaced by the name of the file to retrieve from the archive,
! and any <literal>%p</> is replaced by the path name to copy
! it to on the server.
! (The path name is relative to the working directory of the server,
! i.e., the cluster's data directory.)
Write <literal>%%</> to embed an actual <literal>%</> character
in the command.
</para>
--- 1018,1033 ----
<listitem>
<para>
The shell command to execute to retrieve an archived segment of
! the WAL file series. This parameter is required. Any <literal>%f</>
! in the string is replaced by the name of the file to retrieve from
! the archive, and any <literal>%p</> is replaced by the path name to
! copy it to on the server. (The path name is relative to the working
! directory of the server, i.e., the cluster's data directory.)
! A <literal>%r</> will be replaced by the name of the file containing
! the last valid restartpoint. This is the earliest file that must be
! kept to allow a restore to be restartable, so this information can
! allow if the archive to be truncated to just the minimum required
! to support restart of the current restore.
Write <literal>%%</> to embed an actual <literal>%</> character
in the command.
</para>
***************
*** 1024,1031 ****
names that are not present in the archive; it must return nonzero
when so asked. Examples:
<programlisting>
! restore_command = 'cp /mnt/server/archivedir/%f "%p"'
! restore_command = 'copy /mnt/server/archivedir/%f "%p"' # Windows
</programlisting>
</para>
</listitem>
--- 1037,1044 ----
names that are not present in the archive; it must return nonzero
when so asked. Examples:
<programlisting>
! restore_command = 'cp /mnt/server/archivedir/%f "%p" "%r"'
! restore_command = 'copy /mnt/server/archivedir/%f "%p" "%r"' # Windows
</programlisting>
</para>
</listitem>
***************
*** 1374,1380 ****
contact between the two database servers is the archive of WAL files
that both share: primary writing to the archive, standby reading from
the archive. Care must be taken to ensure that WAL archives for separate
! primary servers do not become mixed together or confused.
</para>
<para>
--- 1387,1394 ----
contact between the two database servers is the archive of WAL files
that both share: primary writing to the archive, standby reading from
the archive. Care must be taken to ensure that WAL archives for separate
! primary servers do not become mixed together or confused. The archive
! need not be large, if it is only required to the standby operation.
</para>
<para>
***************
*** 1434,1439 ****
--- 1448,1466 ----
as the explicit creation of a trigger file is less error prone, if
this can be arranged.
</para>
+
+ <para>
+ A working version of a wait-for <varname>restore_command</> is provided
+ as a contrib module, known as <applications>pg_standby</>. This can be
+ extended as needed to support specific configurations or environments.
+ </para>
+
+ <para>
+ The size of the WAL archive can be minimised by using the <literal>%r</>
+ option of the <varname>restore_command</>. This option specifies the
+ last archive filename that needs to be kept to allow the recovery to
+ restart correctly. This can be used to truncate the archive once
+ files are no longer required.
</sect2>
<sect2 id="warm-standby-config">
Index: src/backend/access/transam/xlog.c
===================================================================
RCS file: /projects/cvsroot/pgsql/src/backend/access/transam/xlog.c,v
retrieving revision 1.263
diff -c -r1.263 xlog.c
*** src/backend/access/transam/xlog.c 8 Feb 2007 11:10:27 -0000 1.263
--- src/backend/access/transam/xlog.c 9 Feb 2007 23:05:34 -0000
***************
*** 190,195 ****
--- 190,196 ----
static TransactionId recoveryStopXid;
static time_t recoveryStopTime;
static bool recoveryStopAfter;
+ static char lastRestorepointFname[MAXFNAMELEN];
/*
* During normal operation, the only timeline we care about is ThisTimeLineID.
***************
*** 2359,2364 ****
--- 2360,2366 ----
{
char xlogpath[MAXPGPATH];
char xlogRestoreCmd[MAXPGPATH];
+ char lastRestartPointFname[MAXPGPATH];
char *dp;
char *endp;
const char *sp;
***************
*** 2437,2442 ****
--- 2439,2454 ----
StrNCpy(dp, xlogfname, endp - dp);
dp += strlen(dp);
break;
+ case 'r':
+ /* %r: filename of last restartpoint */
+ sp++;
+ XLogFileName(lastRestartPointFname,
+ ControlFile->checkPointCopy->ThisTimelineId,
+ ControlFile->checkPointCopy->redo.xlogid,
+ ControlFile->checkPointCopy->redo.xrecoff);
+ StrNCpy(dp, lastRestartPointFname, endp - dp);
+ dp += strlen(dp);
+ break;
case '%':
/* convert %% to a single % */
sp++;
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
