From: Andrey Borodin Date: Sat, 27 Feb 2021 09:03:50 +0500 Subject: [PATCH 01/10] Allow alternate compression methods for wal_compression TODO: bump XLOG_PAGE_MAGIC --- doc/src/sgml/config.sgml | 17 +++++ src/backend/Makefile | 2 +- src/backend/access/transam/xlog.c | 10 +++ src/backend/access/transam/xloginsert.c | 67 ++++++++++++++++--- src/backend/access/transam/xlogreader.c | 64 +++++++++++++++++- src/backend/utils/misc/guc.c | 11 +++ src/backend/utils/misc/postgresql.conf.sample | 1 + src/include/access/xlog.h | 1 + src/include/access/xlog_internal.h | 16 +++++ src/include/access/xlogrecord.h | 11 ++- 10 files changed, 187 insertions(+), 13 deletions(-) diff --git a/doc/src/sgml/config.sgml b/doc/src/sgml/config.sgml index a218d78bef..7fb2a84626 100644 --- a/doc/src/sgml/config.sgml +++ b/doc/src/sgml/config.sgml @@ -3072,6 +3072,23 @@ include_dir 'conf.d' + + wal_compressionion_method (enum) + + wal_compression_method configuration parameter + + + + + This parameter selects the compression method used to compress WAL when + wal_compression is enabled. + The supported methods are pglz and zlib. + The default value is pglz. + Only superusers can change this setting. + + + + wal_init_zero (boolean) diff --git a/src/backend/Makefile b/src/backend/Makefile index 0da848b1fd..3af216ddfc 100644 --- a/src/backend/Makefile +++ b/src/backend/Makefile @@ -48,7 +48,7 @@ OBJS = \ LIBS := $(filter-out -lpgport -lpgcommon, $(LIBS)) $(LDAP_LIBS_BE) $(ICU_LIBS) # The backend doesn't need everything that's in LIBS, however -LIBS := $(filter-out -lz -lreadline -ledit -ltermcap -lncurses -lcurses, $(LIBS)) +LIBS := $(filter-out -lreadline -ledit -ltermcap -lncurses -lcurses, $(LIBS)) ifeq ($(with_systemd),yes) LIBS += -lsystemd diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c index f4d1ce5dea..15da91a8dd 100644 --- a/src/backend/access/transam/xlog.c +++ b/src/backend/access/transam/xlog.c @@ -99,6 +99,7 @@ bool EnableHotStandby = false; bool fullPageWrites = true; bool wal_log_hints = false; bool wal_compression = false; +int wal_compression_method = WAL_COMPRESSION_PGLZ; char *wal_consistency_checking_string = NULL; bool *wal_consistency_checking = NULL; bool wal_init_zero = true; @@ -180,6 +181,15 @@ const struct config_enum_entry recovery_target_action_options[] = { {NULL, 0, false} }; +/* Note that due to conditional compilation, offsets within the array are not static */ +const struct config_enum_entry wal_compression_options[] = { + {"pglz", WAL_COMPRESSION_PGLZ, false}, +#ifdef HAVE_LIBZ + {"zlib", WAL_COMPRESSION_ZLIB, false}, +#endif + {NULL, 0, false} +}; + /* * Statistics for current checkpoint are collected in this global struct. * Because only the checkpointer or a stand-alone backend can perform diff --git a/src/backend/access/transam/xloginsert.c b/src/backend/access/transam/xloginsert.c index 7052dc245e..a93b33464f 100644 --- a/src/backend/access/transam/xloginsert.c +++ b/src/backend/access/transam/xloginsert.c @@ -33,8 +33,18 @@ #include "storage/proc.h" #include "utils/memutils.h" +#ifdef HAVE_LIBZ +#include +/* zlib compressBound is not a macro */ +#define ZLIB_MAX_BLCKSZ BLCKSZ + (BLCKSZ>>12) + (BLCKSZ>>14) + (BLCKSZ>>25) + 13 +#else +#define ZLIB_MAX_BLCKSZ 0 +#endif + /* Buffer size required to store a compressed version of backup block image */ -#define PGLZ_MAX_BLCKSZ PGLZ_MAX_OUTPUT(BLCKSZ) +#define PGLZ_MAX_BLCKSZ PGLZ_MAX_OUTPUT(BLCKSZ) + +#define COMPRESS_BUFSIZE Max(PGLZ_MAX_BLCKSZ, ZLIB_MAX_BLCKSZ) /* * For each block reference registered with XLogRegisterBuffer, we fill in @@ -58,7 +68,7 @@ typedef struct * backup block data in XLogRecordAssemble() */ /* buffer to store a compressed version of backup block image */ - char compressed_page[PGLZ_MAX_BLCKSZ]; + char compressed_page[COMPRESS_BUFSIZE]; } registered_buffer; static registered_buffer *registered_buffers; @@ -113,7 +123,8 @@ static XLogRecData *XLogRecordAssemble(RmgrId rmid, uint8 info, XLogRecPtr RedoRecPtr, bool doPageWrites, XLogRecPtr *fpw_lsn, int *num_fpi); static bool XLogCompressBackupBlock(char *page, uint16 hole_offset, - uint16 hole_length, char *dest, uint16 *dlen); + uint16 hole_length, char *dest, + uint16 *dlen, WalCompression compression); /* * Begin constructing a WAL record. This must be called before the @@ -625,16 +636,26 @@ XLogRecordAssemble(RmgrId rmid, uint8 info, cbimg.hole_length = 0; } + bimg.bimg_info = (cbimg.hole_length == 0) ? 0 : BKPIMAGE_HAS_HOLE; + /* * Try to compress a block image if wal_compression is enabled */ if (wal_compression) { + int compression; + /* The current compression is stored in the WAL record */ + wal_compression_name(wal_compression_method); /* Range check */ + compression = walmethods[wal_compression_method].walmethod; + Assert(compression < (1 << BKPIMAGE_COMPRESS_BITS)); + bimg.bimg_info |= + compression << BKPIMAGE_COMPRESS_OFFSET_BITS; is_compressed = XLogCompressBackupBlock(page, bimg.hole_offset, cbimg.hole_length, regbuf->compressed_page, - &compressed_len); + &compressed_len, + wal_compression_method); } /* @@ -652,8 +673,6 @@ XLogRecordAssemble(RmgrId rmid, uint8 info, rdt_datas_last->next = ®buf->bkp_rdatas[0]; rdt_datas_last = rdt_datas_last->next; - bimg.bimg_info = (cbimg.hole_length == 0) ? 0 : BKPIMAGE_HAS_HOLE; - /* * If WAL consistency checking is enabled for the resource manager * of this WAL record, a full-page image is included in the record @@ -827,7 +846,7 @@ XLogRecordAssemble(RmgrId rmid, uint8 info, */ static bool XLogCompressBackupBlock(char *page, uint16 hole_offset, uint16 hole_length, - char *dest, uint16 *dlen) + char *dest, uint16 *dlen, WalCompression compression) { int32 orig_len = BLCKSZ - hole_length; int32 len; @@ -853,12 +872,42 @@ XLogCompressBackupBlock(char *page, uint16 hole_offset, uint16 hole_length, else source = page; + switch (compression) + { + case WAL_COMPRESSION_PGLZ: + len = pglz_compress(source, orig_len, dest, PGLZ_strategy_default); + break; + +#ifdef HAVE_LIBZ + case WAL_COMPRESSION_ZLIB: + { + unsigned long len_l = COMPRESS_BUFSIZE; + int ret; + ret = compress2((Bytef*)dest, &len_l, (Bytef*)source, orig_len, 1); + if (ret != Z_OK) + len_l = -1; + len = len_l; + break; + } +#endif + + default: + /* + * It should be impossible to get here for unsupported algorithms, + * which cannot be assigned if they're not enabled at compile time. + */ + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("unknown compression method requested: %d(%s)", + compression, wal_compression_name(compression)))); + + } + /* - * We recheck the actual size even if pglz_compress() reports success and + * We recheck the actual size even if compression reports success and * see if the number of bytes saved by compression is larger than the * length of extra data needed for the compressed version of block image. */ - len = pglz_compress(source, orig_len, dest, PGLZ_strategy_default); if (len >= 0 && len + extra_bytes < orig_len) { diff --git a/src/backend/access/transam/xlogreader.c b/src/backend/access/transam/xlogreader.c index 42738eb940..0d8830fc50 100644 --- a/src/backend/access/transam/xlogreader.c +++ b/src/backend/access/transam/xlogreader.c @@ -33,6 +33,10 @@ #include "utils/memutils.h" #endif +#ifdef HAVE_LIBZ +#include +#endif + static void report_invalid_record(XLogReaderState *state, const char *fmt,...) pg_attribute_printf(2, 3); static bool allocate_recordbuf(XLogReaderState *state, uint32 reclength); @@ -1535,6 +1539,30 @@ XLogRecGetBlockData(XLogReaderState *record, uint8 block_id, Size *len) } } +/* This is a mapping indexed by wal_compression */ +// XXX: maybe this is better done as a GUC hook to assign the 1) method; and 2) level +struct walcompression walmethods[] = { + {"pglz", WAL_COMPRESSION_PGLZ}, + {"zlib", WAL_COMPRESSION_ZLIB}, +}; + +/* + * Return a statically allocated string associated with the given compression + * method. + * This is here to be visible to frontend tools like pg_rewind. + */ +const char * +wal_compression_name(WalCompression compression) +{ + /* + * This could index into the guc array, except that it's compiled + * conditionally and unsupported methods are elided. + */ + if (compression < sizeof(walmethods)/sizeof(*walmethods)) + return walmethods[compression].name; + return "???"; +} + /* * Restore a full-page image from a backup block attached to an XLOG record. * @@ -1557,9 +1585,41 @@ RestoreBlockImage(XLogReaderState *record, uint8 block_id, char *page) if (bkpb->bimg_info & BKPIMAGE_IS_COMPRESSED) { + int compression_method = BKPIMAGE_COMPRESSION(bkpb->bimg_info); /* If a backup block image is compressed, decompress it */ - if (pglz_decompress(ptr, bkpb->bimg_len, tmp.data, - BLCKSZ - bkpb->hole_length, true) < 0) + int32 decomp_result = -1; + switch (compression_method) + { + case WAL_COMPRESSION_PGLZ: + decomp_result = pglz_decompress(ptr, bkpb->bimg_len, tmp.data, + BLCKSZ - bkpb->hole_length, true); + break; + +#ifdef HAVE_LIBZ + case WAL_COMPRESSION_ZLIB: + { + unsigned long decomp_result_l; + decomp_result_l = BLCKSZ - bkpb->hole_length; + if (uncompress((Bytef*)tmp.data, &decomp_result_l, + (Bytef*)ptr, bkpb->bimg_len) == Z_OK) + decomp_result = decomp_result_l; + else + decomp_result = -1; + break; + } +#endif + + default: + report_invalid_record(record, "image at %X/%X is compressed with unsupported codec, block %d (%d/%s)", + (uint32) (record->ReadRecPtr >> 32), + (uint32) record->ReadRecPtr, + block_id, + compression_method, + wal_compression_name(compression_method)); + return false; + } + + if (decomp_result < 0) { report_invalid_record(record, "invalid compressed image at %X/%X, block %d", LSN_FORMAT_ARGS(record->ReadRecPtr), diff --git a/src/backend/utils/misc/guc.c b/src/backend/utils/misc/guc.c index 855076b1fd..8084027465 100644 --- a/src/backend/utils/misc/guc.c +++ b/src/backend/utils/misc/guc.c @@ -508,6 +508,7 @@ extern const struct config_enum_entry archive_mode_options[]; extern const struct config_enum_entry recovery_target_action_options[]; extern const struct config_enum_entry sync_method_options[]; extern const struct config_enum_entry dynamic_shared_memory_options[]; +extern const struct config_enum_entry wal_compression_options[]; /* * GUC option variables that are exported from this module @@ -4721,6 +4722,16 @@ static struct config_enum ConfigureNamesEnum[] = NULL, NULL, NULL }, + { + {"wal_compression_method", PGC_SIGHUP, WAL_SETTINGS, + gettext_noop("Set the method used to compress full page images in the WAL."), + NULL + }, + &wal_compression_method, + WAL_COMPRESSION_PGLZ, wal_compression_options, + NULL, NULL, NULL + }, + { {"dynamic_shared_memory_type", PGC_POSTMASTER, RESOURCES_MEM, gettext_noop("Selects the dynamic shared memory implementation used."), diff --git a/src/backend/utils/misc/postgresql.conf.sample b/src/backend/utils/misc/postgresql.conf.sample index f46c2dd7a8..ef69a94492 100644 --- a/src/backend/utils/misc/postgresql.conf.sample +++ b/src/backend/utils/misc/postgresql.conf.sample @@ -213,6 +213,7 @@ # open_sync #full_page_writes = on # recover from partial page writes #wal_compression = off # enable compression of full-page writes +#wal_compression_method = pglz # pglz, zlib #wal_log_hints = off # also do full page writes of non-critical updates # (change requires restart) #wal_init_zero = on # zero-fill new WAL files diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h index 6d384d3ce6..fa2e5c611f 100644 --- a/src/include/access/xlog.h +++ b/src/include/access/xlog.h @@ -117,6 +117,7 @@ extern bool EnableHotStandby; extern bool fullPageWrites; extern bool wal_log_hints; extern bool wal_compression; +extern int wal_compression_method; extern bool wal_init_zero; extern bool wal_recycle; extern bool *wal_consistency_checking; diff --git a/src/include/access/xlog_internal.h b/src/include/access/xlog_internal.h index b23e286406..b000a21557 100644 --- a/src/include/access/xlog_internal.h +++ b/src/include/access/xlog_internal.h @@ -324,4 +324,20 @@ extern bool InArchiveRecovery; extern bool StandbyMode; extern char *recoveryRestoreCommand; +struct walcompression +{ + char *name; + int walmethod; /* Compression method to be stored in WAL */ +}; + +extern struct walcompression walmethods[]; + +typedef enum WalCompression +{ + WAL_COMPRESSION_PGLZ, + WAL_COMPRESSION_ZLIB, +} WalCompression; + +extern const char *wal_compression_name(WalCompression compression); + #endif /* XLOG_INTERNAL_H */ diff --git a/src/include/access/xlogrecord.h b/src/include/access/xlogrecord.h index 80c92a2498..7107cf6186 100644 --- a/src/include/access/xlogrecord.h +++ b/src/include/access/xlogrecord.h @@ -114,7 +114,7 @@ typedef struct XLogRecordBlockHeader * present is (BLCKSZ - ). * * Additionally, when wal_compression is enabled, we will try to compress full - * page images using the PGLZ compression algorithm, after removing the "hole". + * page images, after removing the "hole". * This can reduce the WAL volume, but at some extra cost of CPU spent * on the compression during WAL logging. In this case, since the "hole" * length cannot be calculated by subtracting the number of page image bytes @@ -147,6 +147,15 @@ typedef struct XLogRecordBlockImageHeader #define BKPIMAGE_IS_COMPRESSED 0x02 /* page image is compressed */ #define BKPIMAGE_APPLY 0x04 /* page image should be restored during * replay */ +#define BKPIMAGE_COMPRESS_METHOD1 0x08 /* bits to encode compression method */ +#define BKPIMAGE_COMPRESS_METHOD2 0x10 /* 0=pglz; 1=zlib; */ + +/* How many bits to shift to extract compression */ +#define BKPIMAGE_COMPRESS_OFFSET_BITS 3 +/* How many bits are for compression */ +#define BKPIMAGE_COMPRESS_BITS 2 +/* Extract the compression from the bimg_info */ +#define BKPIMAGE_COMPRESSION(info) ((info >> BKPIMAGE_COMPRESS_OFFSET_BITS) & ((1<