Backend support for streaming base backups

Add BASE_BACKUP command to walsender, allowing it to stream a
base backup to the client (in tar format). The syntax is still
far from ideal, that will be fixed in the switch to use a proper
grammar for walsender.

No client included yet, will come as a separate commit.

Magnus Hagander and Heikki Linnakangas

doc/src/sgml/protocol.sgml

@@ -1458,6 +1458,98 @@ The commands accepted in walsender mode are:
      </para>
     </listitem>
   </varlistentry>
+
+  <varlistentry>
+    <term>BASE_BACKUP <replaceable>options</><literal>;</><replaceable>label</></term>
+    <listitem>
+     <para>
+      Instructs the server to start streaming a base backup.
+      The system will automatically be put in backup mode with the label
+      specified in <replaceable>label</> before the backup is started, and
+      taken out of it when the backup is complete. The following options
+      are accepted:
+      <variablelist>
+       <varlistentry>
+        <term><literal>PROGRESS</></term>
+        <listitem>
+         <para>
+          Request information required to generate a progress report. This will
+          send back an approximate size in the header of each tablespace, which
+          can be used to calculate how far along the stream is done. This is
+          calculated by enumerating all the file sizes once before the transfer
+          is even started, and may as such have a negative impact on the
+          performance - in particular it may take longer before the first data
+          is streamed. Since the database files can change during the backup,
+          the size is only approximate and may both grow and shrink between
+          the time of approximation and the sending of the actual files.
+         </para>
+        </listitem>
+       </varlistentry>
+      </variablelist>
+     </para>
+     <para>
+      When the backup is started, the server will first send a header in
+      ordinary result set format, followed by one or more CopyResponse
+      results, one for PGDATA and one for each additional tablespace other
+      than <literal>pg_default</> and <literal>pg_global</>. The data in
+      the CopyResponse results will be a tar format (using ustar00
+      extensions) dump of the tablespace contents.
+     </para>
+     <para>
+      The header is an ordinary resultset with one row for each tablespace.
+      The fields in this row are:
+      <variablelist>
+       <varlistentry>
+        <term>spcoid</term>
+        <listitem>
+         <para>
+          The oid of the tablespace, or <literal>NULL</> if it's the base
+          directory.
+         </para>
+        </listitem>
+       </varlistentry>
+       <varlistentry>
+        <term>spclocation</term>
+        <listitem>
+         <para>
+          The full path of the tablespace directory, or <literal>NULL</>
+          if it's the base directory.
+         </para>
+        </listitem>
+       </varlistentry>
+       <varlistentry>
+        <term>size</term>
+        <listitem>
+         <para>
+          The approximate size of the tablespace, if progress report has
+          been requested; otherwise it's <literal>NULL</>.
+         </para>
+        </listitem>
+       </varlistentry>
+      </variablelist>
+     </para>
+     <para>
+      The tar archive for the data directory and each tablespace will contain
+      all files in the directories, regardless of whether they are
+      <productname>PostgreSQL</> files or other files added to the same
+      directory. The only excluded files are:
+      <itemizedlist spacing="compact" mark="bullet">
+       <listitem>
+        <para>
+         <filename>postmaster.pid</>
+        </para>
+       </listitem>
+       <listitem>
+        <para>
+         <filename>pg_xlog</> (including subdirectories)
+        </para>
+       </listitem>
+      </itemizedlist>
+      Owner, group and file mode are set if the underlying filesystem on
+      the server supports it.
+     </para>
+    </listitem>
+  </varlistentry>
 </variablelist>
 
 </para>

src/backend/replication/Makefile

@@ -12,6 +12,6 @@ subdir = src/backend/replication
 top_builddir = ../../..
 include $(top_builddir)/src/Makefile.global
 
-OBJS = walsender.o walreceiverfuncs.o walreceiver.o
+OBJS = walsender.o walreceiverfuncs.o walreceiver.o basebackup.o
 
 include $(top_srcdir)/src/backend/common.mk

src/backend/replication/walsender.c

@@ -44,6 +44,7 @@
 #include "libpq/pqformat.h"
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
+#include "replication/basebackup.h"
 #include "replication/walprotocol.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
@@ -54,6 +55,7 @@
 #include "utils/guc.h"
 #include "utils/memutils.h"
 #include "utils/ps_status.h"
+#include "utils/resowner.h"
 
 
 /* Array of WalSnds in shared memory */
@@ -136,6 +138,9 @@ WalSenderMain(void)
 										   ALLOCSET_DEFAULT_MAXSIZE);
 	MemoryContextSwitchTo(walsnd_context);
 
+	/* Set up resource owner */
+	CurrentResourceOwner = ResourceOwnerCreate(NULL, "walsender top-level resource owner");
+
 	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
@@ -305,6 +310,15 @@ WalSndHandshake(void)
 						/* break out of the loop */
 						replication_started = true;
 					}
+					else if (strncmp(query_string, "BASE_BACKUP ", 12) == 0)
+					{
+						/* Command is BASE_BACKUP <options>;<label> */
+						SendBaseBackup(query_string + strlen("BASE_BACKUP "));
+						/* Send CommandComplete and ReadyForQuery messages */
+						EndCommand("SELECT", DestRemote);
+						ReadyForQuery(DestRemote);
+						/* ReadyForQuery did pq_flush for us */
+					}
 					else
 					{
 						ereport(FATAL,

src/include/replication/basebackup.h

+/*-------------------------------------------------------------------------
+ *
+ * basebackup.h
+ *	  Exports from replication/basebackup.c.
+ *
+ * Portions Copyright (c) 2010-2010, PostgreSQL Global Development Group
+ *
+ * src/include/replication/walsender.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _BASEBACKUP_H
+#define _BASEBACKUP_H
+
+extern void SendBaseBackup(const char *options);
+
+#endif   /* _BASEBACKUP_H */