Add support for zstd with compression of full-page writes in WAL

michaelpq · michaelpq · commit e9537321a74a · 2022-03-11T12:18:53.000+09:00
wal_compression gains a new value, "zstd", to allow the compression of full-page images using the compression method of the same name. Compression is done using the default level recommended by the library, as of ZSTD_CLEVEL_DEFAULT = 3. Some benchmarking has shown that it could make sense to use a level lower for the FPI compression, like 1 or 2, as the compression rate did not change much with a bit less CPU consumed, but any tests done would only cover few scenarios so it is hard to come to a clear conclusion. Anyway, there is no reason to not use the default level instead, which is the level recommended by the library so it should be fine for most cases. zstd outclasses easily pglz, and is better than LZ4 where one wants to have more compression at the cost of extra CPU but both are good enough in their own scenarios, so the choice between one or the other of these comes to a study of the workload patterns and the schema involved, mainly. This commit relies heavily on 4035cd5, that reshaped the code creating and restoring full-page writes to be aware of the compression type, making this integration straight-forward. This patch borrows some early work from Andrey Borodin, though the patch got a complete rewrite. Author: Justin Pryzby Discussion: https://postgr.es/m/20220222231948.GJ9008@telsasoft.com
diff --git a/doc/src/sgml/config.sgml b/doc/src/sgml/config.sgml
@@ -3154,10 +3154,13 @@ include_dir 'conf.d'
         server compresses full page images written to WAL when
         <xref linkend="guc-full-page-writes"/> is on or during a base backup.
         A compressed page image will be decompressed during WAL replay.
-        The supported methods are <literal>pglz</literal> and
-        <literal>lz4</literal> (if <productname>PostgreSQL</productname> was
-        compiled with <option>--with-lz4</option>). The default value is
-        <literal>off</literal>. Only superusers can change this setting.
+        The supported methods are <literal>pglz</literal>,
+        <literal>lz4</literal> (if <productname>PostgreSQL</productname>
+        was compiled with <option>--with-lz4</option>) and
+        <literal>zstd</literal> (if <productname>PostgreSQL</productname>
+        was compiled with <option>--with-zstd</option>) and
+        The default value is <literal>off</literal>.
+        Only superusers can change this setting.
        </para>
 
        <para>
diff --git a/doc/src/sgml/installation.sgml b/doc/src/sgml/installation.sgml
@@ -271,6 +271,14 @@ su - postgres
      </para>
     </listitem>
 
+    <listitem>
+     <para>
+      You need <productname>zstd</productname>, if you want to support
+      compression of data with this method; see
+      <xref linkend="guc-wal-compression"/>.
+     </para>
+    </listitem>
+
     <listitem>
      <para>
       To build the <productname>PostgreSQL</productname> documentation,
diff --git a/src/backend/access/transam/xloginsert.c b/src/backend/access/transam/xloginsert.c
@@ -23,6 +23,10 @@
 #include <lz4.h>
 #endif
 
+#ifdef USE_ZSTD
+#include <zstd.h>
+#endif
+
 #include "access/xact.h"
 #include "access/xlog.h"
 #include "access/xlog_internal.h"
@@ -47,9 +51,16 @@
 #define LZ4_MAX_BLCKSZ		0
 #endif
 
+#ifdef USE_ZSTD
+#define ZSTD_MAX_BLCKSZ		ZSTD_COMPRESSBOUND(BLCKSZ)
+#else
+#define ZSTD_MAX_BLCKSZ		0
+#endif
+
 #define PGLZ_MAX_BLCKSZ		PGLZ_MAX_OUTPUT(BLCKSZ)
 
-#define COMPRESS_BUFSIZE	Max(PGLZ_MAX_BLCKSZ, LZ4_MAX_BLCKSZ)
+/* Buffer size required to store a compressed version of backup block image */
+#define COMPRESS_BUFSIZE	Max(Max(PGLZ_MAX_BLCKSZ, LZ4_MAX_BLCKSZ), ZSTD_MAX_BLCKSZ)
 
 /*
  * For each block reference registered with XLogRegisterBuffer, we fill in
@@ -698,6 +709,14 @@ XLogRecordAssemble(RmgrId rmid, uint8 info,
 #endif
 						break;
 
+					case WAL_COMPRESSION_ZSTD:
+#ifdef USE_ZSTD
+						bimg.bimg_info |= BKPIMAGE_COMPRESS_ZSTD;
+#else
+						elog(ERROR, "zstd is not supported by this build");
+#endif
+						break;
+
 					case WAL_COMPRESSION_NONE:
 						Assert(false);	/* cannot happen */
 						break;
@@ -906,6 +925,17 @@ XLogCompressBackupBlock(char *page, uint16 hole_offset, uint16 hole_length,
 #endif
 			break;
 
+		case WAL_COMPRESSION_ZSTD:
+#ifdef USE_ZSTD
+			len = ZSTD_compress(dest, COMPRESS_BUFSIZE, source, orig_len,
+								ZSTD_CLEVEL_DEFAULT);
+			if (ZSTD_isError(len))
+				len = -1;		/* failure */
+#else
+			elog(ERROR, "zstd is not supported by this build");
+#endif
+			break;
+
 		case WAL_COMPRESSION_NONE:
 			Assert(false);		/* cannot happen */
 			break;
diff --git a/src/backend/access/transam/xlogreader.c b/src/backend/access/transam/xlogreader.c
@@ -21,6 +21,9 @@
 #ifdef USE_LZ4
 #include <lz4.h>
 #endif
+#ifdef USE_ZSTD
+#include <zstd.h>
+#endif
 
 #include "access/transam.h"
 #include "access/xlog_internal.h"
@@ -1618,6 +1621,23 @@ RestoreBlockImage(XLogReaderState *record, uint8 block_id, char *page)
 								  "LZ4",
 								  block_id);
 			return false;
+#endif
+		}
+		else if ((bkpb->bimg_info & BKPIMAGE_COMPRESS_ZSTD) != 0)
+		{
+#ifdef USE_ZSTD
+			size_t		decomp_result = ZSTD_decompress(tmp.data,
+														BLCKSZ - bkpb->hole_length,
+														ptr, bkpb->bimg_len);
+
+			if (ZSTD_isError(decomp_result))
+				decomp_success = false;
+#else
+			report_invalid_record(record, "image at %X/%X compressed with %s not supported by build, block %d",
+								  LSN_FORMAT_ARGS(record->ReadRecPtr),
+								  "zstd",
+								  block_id);
+			return false;
 #endif
 		}
 		else
diff --git a/src/backend/utils/misc/guc.c b/src/backend/utils/misc/guc.c
@@ -550,6 +550,9 @@ static const struct config_enum_entry wal_compression_options[] = {
 	{"pglz", WAL_COMPRESSION_PGLZ, false},
 #ifdef USE_LZ4
 	{"lz4", WAL_COMPRESSION_LZ4, false},
+#endif
+#ifdef USE_ZSTD
+	{"zstd", WAL_COMPRESSION_ZSTD, false},
 #endif
 	{"on", WAL_COMPRESSION_PGLZ, false},
 	{"off", WAL_COMPRESSION_NONE, false},
diff --git a/src/backend/utils/misc/postgresql.conf.sample b/src/backend/utils/misc/postgresql.conf.sample
@@ -220,7 +220,7 @@
 #wal_log_hints = off			# also do full page writes of non-critical updates
 					# (change requires restart)
 #wal_compression = off			# enables compression of full-page writes;
-					# off, pglz, lz4, or on
+					# off, pglz, lz4, zstd, or on
 #wal_init_zero = on			# zero-fill new WAL files
 #wal_recycle = on			# recycle WAL files
 #wal_buffers = -1			# min 32kB, -1 sets based on shared_buffers
diff --git a/src/bin/pg_waldump/pg_waldump.c b/src/bin/pg_waldump/pg_waldump.c
@@ -562,6 +562,8 @@ XLogDumpDisplayRecord(XLogDumpConfig *config, XLogReaderState *record)
 						method = "pglz";
 					else if ((bimg_info & BKPIMAGE_COMPRESS_LZ4) != 0)
 						method = "lz4";
+					else if ((bimg_info & BKPIMAGE_COMPRESS_ZSTD) != 0)
+						method = "zstd";
 					else
 						method = "unknown";
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
@@ -75,7 +75,8 @@ typedef enum WalCompression
 {
 	WAL_COMPRESSION_NONE = 0,
 	WAL_COMPRESSION_PGLZ,
-	WAL_COMPRESSION_LZ4
+	WAL_COMPRESSION_LZ4,
+	WAL_COMPRESSION_ZSTD
 } WalCompression;
 
 /* Recovery states */
diff --git a/src/include/access/xlogrecord.h b/src/include/access/xlogrecord.h
@@ -149,8 +149,11 @@ typedef struct XLogRecordBlockImageHeader
 /* compression methods supported */
 #define BKPIMAGE_COMPRESS_PGLZ	0x04
 #define BKPIMAGE_COMPRESS_LZ4	0x08
+#define BKPIMAGE_COMPRESS_ZSTD	0x10
+
 #define	BKPIMAGE_COMPRESSED(info) \
-	((info & (BKPIMAGE_COMPRESS_PGLZ | BKPIMAGE_COMPRESS_LZ4)) != 0)
+	((info & (BKPIMAGE_COMPRESS_PGLZ | BKPIMAGE_COMPRESS_LZ4 | \
+			  BKPIMAGE_COMPRESS_ZSTD)) != 0)
 
 /*
  * Extra header information used when page image has "hole" and

Original file line number	Diff line number	Diff line change
`@@ -75,7 +75,8 @@ typedef enum WalCompression`
`75`	`75`	`{`
`76`	`76`	`WAL_COMPRESSION_NONE = 0,`
`77`	`77`	`WAL_COMPRESSION_PGLZ,`
`78`		`- WAL_COMPRESSION_LZ4`
	`78`	`+ WAL_COMPRESSION_LZ4,`
	`79`	`+ WAL_COMPRESSION_ZSTD`
`79`	`80`	`} WalCompression;`
`80`	`81`
`81`	`82`	`/* Recovery states */`