bitmap feature merge

redis.conf

@@ -1716,6 +1716,11 @@ notify-keyspace-events ""
 # gopher-enabled no
 
 ############################### ADVANCED CONFIG ###############################
+# Bitmap object are processed as subkey during swap-in or swap-out in ROR. Different size of subkey
+# may have an impact on performance. Which size is best may depends on the hardware. It's recommended to 
+# keep same as the minimum read/write unit of SSD or HDD. Default size is 4096 bytes. Do not modify it 
+# during the server running!!!
+bitmap-subkey-size 4096
 
 # Hashes are encoded using a memory efficient data structure when they have a
 # small number of entries, and the biggest entry does not exceed a given

src/Makefile

@@ -309,7 +309,7 @@ endif
 
 REDIS_SERVER_NAME=redis-server$(PROG_SUFFIX)
 REDIS_SENTINEL_NAME=redis-sentinel$(PROG_SUFFIX)
-REDIS_SERVER_OBJ=adlist.o quicklist.o ae.o anet.o dict.o server.o sds.o zmalloc.o lzf_c.o lzf_d.o pqsort.o zipmap.o sha1.o ziplist.o release.o networking.o util.o object.o db.o replication.o rdb.o t_string.o t_list.o t_set.o t_zset.o t_hash.o config.o aof.o pubsub.o multi.o debug.o sort.o intset.o syncio.o cluster.o crc16.o endianconv.o slowlog.o scripting.o bio.o rio.o rand.o memtest.o crcspeed.o crc64.o bitops.o sentinel.o notify.o setproctitle.o blocked.o hyperloglog.o latency.o sparkline.o redis-check-rdb.o redis-check-aof.o geo.o lazyfree.o module.o evict.o expire.o geohash.o geohash_helper.o childinfo.o defrag.o siphash.o rax.o t_stream.o listpack.o localtime.o lolwut.o lolwut5.o lolwut6.o acl.o gopher.o tracking.o connection.o tls.o sha256.o timeout.o setcpuaffinity.o monotonic.o mt19937-64.o ctrip.o ctrip_swap.o ctrip_swap_adlist.o ctrip_lru_cache.o ctrip_swap_async.o ctrip_swap_batch.o ctrip_swap_cmd.o ctrip_swap_data.o ctrip_swap_debug.o ctrip_swap_evict.o ctrip_swap_exec.o ctrip_swap_expire.o ctrip_swap_hash.o ctrip_swap_set.o ctrip_swap_list.o ctrip_swap_iter.o ctrip_swap_zset.o ctrip_swap_meta.o ctrip_swap_object.o ctrip_swap_rdb.o ctrip_swap_repl.o ctrip_swap_rio.o ctrip_swap_rocks.o ctrip_swap_stat.o ctrip_swap_sync.o ctrip_swap_thread.o ctrip_swap_util.o ctrip_swap_lock.o ctrip_swap_string.o  ctrip_swap_compact.o  ctrip_swap_slowlog.o ctrip_swap_blocked.o ../deps/xredis-gtid/xredis_gtid.o ctrip_cuckoo_hash.o ctrip_cuckoo_filter.o ctrip_swap_filter.o ctrip_absent_cache.o ctrip_swap_load.o ctrip_swap_dirty.o ctrip_swap_persist.o ctrip_swap_rordb.o
+REDIS_SERVER_OBJ=adlist.o quicklist.o ae.o anet.o dict.o server.o sds.o zmalloc.o lzf_c.o lzf_d.o pqsort.o zipmap.o sha1.o ziplist.o release.o networking.o util.o object.o db.o replication.o rdb.o t_string.o t_list.o t_set.o t_zset.o t_hash.o config.o aof.o pubsub.o multi.o debug.o sort.o intset.o syncio.o cluster.o crc16.o endianconv.o slowlog.o scripting.o bio.o rio.o rand.o memtest.o crcspeed.o crc64.o bitops.o sentinel.o notify.o setproctitle.o blocked.o hyperloglog.o latency.o sparkline.o redis-check-rdb.o redis-check-aof.o geo.o lazyfree.o module.o evict.o expire.o geohash.o geohash_helper.o childinfo.o defrag.o siphash.o rax.o t_stream.o listpack.o localtime.o lolwut.o lolwut5.o lolwut6.o acl.o gopher.o tracking.o connection.o tls.o sha256.o timeout.o setcpuaffinity.o monotonic.o mt19937-64.o ctrip.o ctrip_swap.o ctrip_swap_adlist.o ctrip_lru_cache.o ctrip_swap_async.o ctrip_swap_batch.o ctrip_swap_cmd.o ctrip_swap_data.o ctrip_swap_debug.o ctrip_swap_evict.o ctrip_swap_exec.o ctrip_swap_expire.o ctrip_swap_hash.o ctrip_swap_set.o ctrip_swap_list.o ctrip_swap_iter.o ctrip_swap_zset.o ctrip_swap_meta.o ctrip_swap_object.o ctrip_swap_rdb.o ctrip_swap_repl.o ctrip_swap_rio.o ctrip_swap_rocks.o ctrip_swap_stat.o ctrip_swap_sync.o ctrip_swap_thread.o ctrip_swap_util.o ctrip_swap_lock.o ctrip_swap_string.o ctrip_swap_bitmap.o ctrip_swap_compact.o  ctrip_swap_slowlog.o ctrip_swap_blocked.o ../deps/xredis-gtid/xredis_gtid.o ctrip_cuckoo_hash.o ctrip_cuckoo_filter.o ctrip_swap_filter.o ctrip_absent_cache.o ctrip_swap_load.o ctrip_swap_dirty.o ctrip_swap_persist.o ctrip_roaring_bitmap.o ctrip_swap_rordb.o
 REDIS_CLI_NAME=redis-cli$(PROG_SUFFIX)
 REDIS_CLI_OBJ=anet.o adlist.o dict.o redis-cli.o zmalloc.o release.o ae.o crcspeed.o crc64.o siphash.o crc16.o monotonic.o cli_common.o mt19937-64.o
 REDIS_BENCHMARK_NAME=redis-benchmark$(PROG_SUFFIX)

src/bitops.c

@@ -486,9 +486,21 @@ robj *lookupStringForBitCommand(client *c, uint64_t maxbit) {
     if (o == NULL) {
         o = createObject(OBJ_STRING,sdsnewlen(NULL, byte+1));
         dbAdd(c->db,c->argv[1],o);
+        dbAddMeta(c->db,c->argv[1],createBitmapObjectMarker()); /* add meta to distinguish between bitmap and string types. */
     } else {
         o = dbUnshareStringValue(c->db,c->argv[1],o);
-        o->ptr = sdsgrowzero(o->ptr,byte+1);
+        objectMeta *om = lookupMeta(c->db,c->argv[1]);
+        metaBitmap meta_bitmap;
+        if (server.swap_mode == SWAP_MODE_DISK) {
+            serverAssert(om->swap_type == SWAP_TYPE_BITMAP);
+            metaBitmapInit(&meta_bitmap, objectMetaGetPtr(om), o);
+        } else {
+            //TODO confirm fixed how?
+            /* maybe it is a empty string object. */
+            /* it is never processed as bitmap in ror. */
+            metaBitmapInit(&meta_bitmap, NULL, o);
+        }
+        metaBitmapGrow(&meta_bitmap, byte+1);
     }
     return o;
 }
@@ -768,7 +780,9 @@ void bitopCommand(client *c) {
     /* Store the computed value into the target key */
     if (maxlen) {
         o = createObject(OBJ_STRING,res);
+        bitmapClearObjectMarkerIfNeeded(c->db,targetkey);
         setKey(c,c->db,targetkey,o);
+        bitmapSetObjectMarkerIfNeeded(c->db,targetkey);
         notifyKeyspaceEventDirty(NOTIFY_STRING,"set",targetkey,c->db->id,o,NULL);
         decrRefCount(o);
         server.dirty++;
@@ -780,18 +794,18 @@ void bitopCommand(client *c) {
     addReplyLongLong(c,maxlen); /* Return the output string length in bytes. */
 }
 
-/* BITCOUNT key [start end] */
-void bitcountCommand(client *c) {
-    robj *o;
+void metaBitmapBitcount(metaBitmap *meta_bitmap, client *c)
+{
     long start, end, strlen;
     unsigned char *p;
     char llbuf[LONG_STR_SIZE];
+    robj *o = meta_bitmap->bitmap;
 
-    /* Lookup, check for type, and return 0 for non existing keys. */
-    if ((o = lookupKeyReadOrReply(c,c->argv[1],shared.czero)) == NULL ||
-        checkType(c,o,OBJ_STRING)) return;
     p = getObjectReadOnlyString(o,&strlen,llbuf);
 
+    /* maybe it is no hole in object. */
+    long bitmap_size = meta_bitmap->meta == NULL? strlen:(long)metaBitmapGetSize(meta_bitmap);
+
     /* Parse start/end range if any. */
     if (c->argc == 4) {
         if (getLongFromObjectOrReply(c,c->argv[2],&start,NULL) != C_OK)
@@ -803,15 +817,15 @@ void bitcountCommand(client *c) {
             addReply(c,shared.czero);
             return;
         }
-        if (start < 0) start = strlen+start;
-        if (end < 0) end = strlen+end;
+        if (start < 0) start = bitmap_size+start;
+        if (end < 0) end = bitmap_size+end;
         if (start < 0) start = 0;
         if (end < 0) end = 0;
-        if (end >= strlen) end = strlen-1;
+        if (end >= bitmap_size) end = bitmap_size-1;
     } else if (c->argc == 2) {
         /* The whole string. */
         start = 0;
-        end = strlen-1;
+        end = bitmap_size-1;
     } else {
         /* Syntax error. */
         addReplyErrorObject(c,shared.syntaxerr);
@@ -823,39 +837,49 @@ void bitcountCommand(client *c) {
     if (start > end) {
         addReply(c,shared.czero);
     } else {
+        unsigned long cold_subkeys_size = metaBitmapGetColdSubkeysSize(meta_bitmap, start);
+
+        start -= cold_subkeys_size;
+        end -= cold_subkeys_size;
+
         long bytes = end-start+1;
 
         addReplyLongLong(c,redisPopcount(p+start,bytes));
     }
 }
 
-/* BITPOS key bit [start [end]] */
-void bitposCommand(client *c) {
+/* BITCOUNT key [start end] */
+void bitcountCommand(client *c) {
     robj *o;
-    long bit, start, end, strlen;
+
+    /* Lookup, check for type, and return 0 for non existing keys. */
+    if ((o = lookupKeyReadOrReply(c,c->argv[1],shared.czero)) == NULL ||
+        checkType(c,o,OBJ_STRING)) return;
+
+    objectMeta *om = lookupMeta(c->db,c->argv[1]);
+    metaBitmap meta_bitmap;
+    if (server.swap_mode == SWAP_MODE_DISK) {
+        serverAssert(om->swap_type == SWAP_TYPE_BITMAP);
+        metaBitmapInit(&meta_bitmap, objectMetaGetPtr(om), o);
+    } else {
+        /* it is never processed as bitmap in ror. */
+        metaBitmapInit(&meta_bitmap, NULL, o);
+    }
+    metaBitmapBitcount(&meta_bitmap, c);
+}
+
+void metaBitmapBitpos(metaBitmap *meta_bitmap, client *c, unsigned long bit)
+{
+    long start, end, strlen;
     unsigned char *p;
     char llbuf[LONG_STR_SIZE];
     int end_given = 0;
+    robj *o = meta_bitmap->bitmap;
 
-    /* Parse the bit argument to understand what we are looking for, set
-     * or clear bits. */
-    if (getLongFromObjectOrReply(c,c->argv[2],&bit,NULL) != C_OK)
-        return;
-    if (bit != 0 && bit != 1) {
-        addReplyError(c, "The bit argument must be 1 or 0.");
-        return;
-    }
-
-    /* If the key does not exist, from our point of view it is an infinite
-     * array of 0 bits. If the user is looking for the fist clear bit return 0,
-     * If the user is looking for the first set bit, return -1. */
-    if ((o = lookupKeyRead(c->db,c->argv[1])) == NULL) {
-        addReplyLongLong(c, bit ? -1 : 0);
-        return;
-    }
-    if (checkType(c,o,OBJ_STRING)) return;
     p = getObjectReadOnlyString(o,&strlen,llbuf);
 
+    /* maybe it is no hole in object. */
+    long bitmap_size = meta_bitmap->meta == NULL? strlen:(long)metaBitmapGetSize(meta_bitmap);
     /* Parse start/end range if any. */
     if (c->argc == 4 || c->argc == 5) {
         if (getLongFromObjectOrReply(c,c->argv[3],&start,NULL) != C_OK)
@@ -865,18 +889,18 @@ void bitposCommand(client *c) {
                 return;
             end_given = 1;
         } else {
-            end = strlen-1;
+            end = bitmap_size-1;
         }
         /* Convert negative indexes */
-        if (start < 0) start = strlen+start;
-        if (end < 0) end = strlen+end;
+        if (start < 0) start = bitmap_size+start;
+        if (end < 0) end = bitmap_size+end;
         if (start < 0) start = 0;
         if (end < 0) end = 0;
-        if (end >= strlen) end = strlen-1;
+        if (end >= bitmap_size) end = bitmap_size-1;
     } else if (c->argc == 3) {
         /* The whole string. */
         start = 0;
-        end = strlen-1;
+        end = bitmap_size-1;
     } else {
         /* Syntax error. */
         addReplyErrorObject(c,shared.syntaxerr);
@@ -888,6 +912,12 @@ void bitposCommand(client *c) {
     if (start > end) {
         addReplyLongLong(c, -1);
     } else {
+
+        unsigned long cold_subkeys_size = metaBitmapGetColdSubkeysSize(meta_bitmap, start);
+
+        start -= cold_subkeys_size;
+        end -= cold_subkeys_size;
+
         long bytes = end-start+1;
         long long pos = redisBitpos(p+start,bytes,bit);
 
@@ -903,8 +933,47 @@ void bitposCommand(client *c) {
             return;
         }
         if (pos != -1) pos += (long long)start<<3; /* Adjust for the bytes we skipped. */
+
+        pos += cold_subkeys_size * 8;
         addReplyLongLong(c,pos);
     }
+    return;
+}
+
+/* BITPOS key bit [start [end]] */
+void bitposCommand(client *c) {
+    robj *o;
+    long bit;
+
+    /* Parse the bit argument to understand what we are looking for, set
+     * or clear bits. */
+    if (getLongFromObjectOrReply(c,c->argv[2],&bit,NULL) != C_OK)
+        return;
+    if (bit != 0 && bit != 1) {
+        addReplyError(c, "The bit argument must be 1 or 0.");
+        return;
+    }
+
+    /* If the key does not exist, from our point of view it is an infinite
+     * array of 0 bits. If the user is looking for the fist clear bit return 0,
+     * If the user is looking for the first set bit, return -1. */
+    if ((o = lookupKeyRead(c->db,c->argv[1])) == NULL) {
+        addReplyLongLong(c, bit ? -1 : 0);
+        return;
+    }
+    if (checkType(c,o,OBJ_STRING)) return;
+
+    objectMeta *om = lookupMeta(c->db,c->argv[1]);
+    metaBitmap meta_bitmap;
+    if (server.swap_mode == SWAP_MODE_DISK) {
+        serverAssert(om->swap_type == SWAP_TYPE_BITMAP);
+        metaBitmapInit(&meta_bitmap, objectMetaGetPtr(om), o);
+    } else {
+        /* it is never processed as bitmap in ror. */
+        metaBitmapInit(&meta_bitmap, NULL, o);
+    }
+
+    metaBitmapBitpos(&meta_bitmap, c, bit);
 }
 
 /* BITFIELD key subcommmand-1 arg ... subcommand-2 arg ... subcommand-N ...

src/config.c

@@ -2896,6 +2896,7 @@ standardConfig configs[] = {
     createBoolConfig("swap-dirty-subkeys-enabled", NULL, MODIFIABLE_CONFIG, server.swap_dirty_subkeys_enabled, 0, NULL, NULL),
     createBoolConfig("swap-persist-enabled", NULL, IMMUTABLE_CONFIG, server.swap_persist_enabled, 0, NULL, NULL),
     createBoolConfig("swap-repl-rordb-sync", NULL, MODIFIABLE_CONFIG, server.swap_repl_rordb_sync, 1, NULL, NULL),
+    createBoolConfig("swap-rdb-bitmap-encode-enabled", NULL, MODIFIABLE_CONFIG, server.swap_rdb_bitmap_encode_enabled, 1, NULL, NULL),
     createBoolConfig("rocksdb.data.cache_index_and_filter_blocks", "rocksdb.cache_index_and_filter_blocks", IMMUTABLE_CONFIG, server.rocksdb_data_cache_index_and_filter_blocks, 0, NULL, NULL),
     createBoolConfig("rocksdb.meta.cache_index_and_filter_blocks", NULL, IMMUTABLE_CONFIG, server.rocksdb_meta_cache_index_and_filter_blocks, 0, NULL, NULL),
     createBoolConfig("rocksdb.enable_pipelined_write", NULL, IMMUTABLE_CONFIG, server.rocksdb_enable_pipelined_write, 0, NULL, NULL),
@@ -3095,6 +3096,7 @@ standardConfig configs[] = {
     createSizeTConfig("hll-sparse-max-bytes", NULL, MODIFIABLE_CONFIG, 0, LONG_MAX, server.hll_sparse_max_bytes, 3000, MEMORY_CONFIG, NULL, NULL),
     createSizeTConfig("tracking-table-max-keys", NULL, MODIFIABLE_CONFIG, 0, LONG_MAX, server.tracking_table_max_keys, 1000000, INTEGER_CONFIG, NULL, NULL), /* Default: 1 million keys max. */
     createSizeTConfig("client-query-buffer-limit", NULL, MODIFIABLE_CONFIG, 1024*1024, LONG_MAX, server.client_max_querybuf_len, 1024*1024*1024, MEMORY_CONFIG, NULL, NULL), /* Default: 1GB max query buffer. */
+    createSizeTConfig("bitmap-subkey-size", NULL, MODIFIABLE_CONFIG, 256, 16*1024, server.swap_bitmap_subkey_size, 4*1024, MEMORY_CONFIG, NULL, NULL), /* Default: 4096 bytes. */
 
     /* Other configs */
     createTimeTConfig("repl-backlog-ttl", NULL, MODIFIABLE_CONFIG, 0, LONG_MAX, server.repl_backlog_time_limit, 60*60, INTEGER_CONFIG, NULL, NULL), /* Default: 1 hour */