X-Git-Url: https://git.saurik.com/redis.git/blobdiff_plain/e2641e09cc0daf44f63f654230f72d22acf3a9af..f0d962c0eca870e3fd2a4529de321db217c2e9fe:/src/db.c?ds=inline diff --git a/src/db.c b/src/db.c index e1e82cb2..e65106a5 100644 --- a/src/db.c +++ b/src/db.c @@ -1,6 +1,10 @@ #include "redis.h" #include +#include + +void SlotToKeyAdd(robj *key); +void SlotToKeyDel(robj *key); /*----------------------------------------------------------------------------- * C-level DB API @@ -9,30 +13,13 @@ robj *lookupKey(redisDb *db, robj *key) { dictEntry *de = dictFind(db->dict,key->ptr); if (de) { - robj *val = dictGetEntryVal(de); - - if (server.vm_enabled) { - if (val->storage == REDIS_VM_MEMORY || - val->storage == REDIS_VM_SWAPPING) - { - /* If we were swapping the object out, cancel the operation */ - if (val->storage == REDIS_VM_SWAPPING) - vmCancelThreadedIOJob(val); - /* Update the access time for the aging algorithm. */ - val->lru = server.lruclock; - } else { - int notify = (val->storage == REDIS_VM_LOADING); - - /* Our value was swapped on disk. Bring it at home. */ - redisAssert(val->type == REDIS_VMPOINTER); - val = vmLoadObject(val); - dictGetEntryVal(de) = val; - - /* Clients blocked by the VM subsystem may be waiting for - * this key... */ - if (notify) handleClientsBlockedOnSwappedKey(db,key); - } - } + robj *val = dictGetVal(de); + + /* Update the access time for the aging algorithm. + * Don't do it if we have a saving child, as this will trigger + * a copy on write madness. */ + if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) + val->lru = server.lruclock; return val; } else { return NULL; @@ -40,13 +27,19 @@ robj *lookupKey(redisDb *db, robj *key) { } robj *lookupKeyRead(redisDb *db, robj *key) { + robj *val; + expireIfNeeded(db,key); - return lookupKey(db,key); + val = lookupKey(db,key); + if (val == NULL) + server.stat_keyspace_misses++; + else + server.stat_keyspace_hits++; + return val; } robj *lookupKeyWrite(redisDb *db, robj *key) { - deleteIfVolatile(db,key); - touchWatchedKey(db,key); + expireIfNeeded(db,key); return lookupKey(db,key); } @@ -62,34 +55,44 @@ robj *lookupKeyWriteOrReply(redisClient *c, robj *key, robj *reply) { return o; } -/* Add the key to the DB. If the key already exists REDIS_ERR is returned, - * otherwise REDIS_OK is returned, and the caller should increment the - * refcount of 'val'. */ -int dbAdd(redisDb *db, robj *key, robj *val) { - /* Perform a lookup before adding the key, as we need to copy the - * key value. */ - if (dictFind(db->dict, key->ptr) != NULL) { - return REDIS_ERR; - } else { - sds copy = sdsdup(key->ptr); - dictAdd(db->dict, copy, val); - return REDIS_OK; - } +/* Add the key to the DB. It's up to the caller to increment the reference + * counte of the value if needed. + * + * The program is aborted if the key already exists. */ +void dbAdd(redisDb *db, robj *key, robj *val) { + sds copy = sdsdup(key->ptr); + int retval = dictAdd(db->dict, copy, val); + + redisAssertWithInfo(NULL,key,retval == REDIS_OK); + } + +/* Overwrite an existing key with a new value. Incrementing the reference + * count of the new value is up to the caller. + * This function does not modify the expire time of the existing key. + * + * The program is aborted if the key was not already present. */ +void dbOverwrite(redisDb *db, robj *key, robj *val) { + struct dictEntry *de = dictFind(db->dict,key->ptr); + + redisAssertWithInfo(NULL,key,de != NULL); + dictReplace(db->dict, key->ptr, val); } -/* If the key does not exist, this is just like dbAdd(). Otherwise - * the value associated to the key is replaced with the new one. +/* High level Set operation. This function can be used in order to set + * a key, whatever it was existing or not, to a new object. * - * On update (key already existed) 0 is returned. Otherwise 1. */ -int dbReplace(redisDb *db, robj *key, robj *val) { - if (dictFind(db->dict,key->ptr) == NULL) { - sds copy = sdsdup(key->ptr); - dictAdd(db->dict, copy, val); - return 1; + * 1) The ref count of the value object is incremented. + * 2) clients WATCHing for the destination key notified. + * 3) The expire time of the key is reset (the key is made persistent). */ +void setKey(redisDb *db, robj *key, robj *val) { + if (lookupKeyWrite(db,key) == NULL) { + dbAdd(db,key,val); } else { - dictReplace(db->dict, key->ptr, val); - return 0; + dbOverwrite(db,key,val); } + incrRefCount(val); + removeExpire(db,key); + signalModifiedKey(db,key); } int dbExists(redisDb *db, robj *key) { @@ -110,7 +113,7 @@ robj *dbRandomKey(redisDb *db) { de = dictGetRandomKey(db->dict); if (de == NULL) return NULL; - key = dictGetEntryKey(de); + key = dictGetKey(de); keyobj = createStringObject(key,sdslen(key)); if (dictFind(db->expires,key)) { if (expireIfNeeded(db,keyobj)) { @@ -127,10 +130,13 @@ int dbDelete(redisDb *db, robj *key) { /* Deleting an entry from the expires dict will not free the sds of * the key, because it is shared with the main dictionary. */ if (dictSize(db->expires) > 0) dictDelete(db->expires,key->ptr); - return dictDelete(db->dict,key->ptr) == DICT_OK; + if (dictDelete(db->dict,key->ptr) == DICT_OK) { + return 1; + } else { + return 0; + } } -/* Empty the whole database */ long long emptyDb() { int j; long long removed = 0; @@ -150,27 +156,50 @@ int selectDb(redisClient *c, int id) { return REDIS_OK; } +/*----------------------------------------------------------------------------- + * Hooks for key space changes. + * + * Every time a key in the database is modified the function + * signalModifiedKey() is called. + * + * Every time a DB is flushed the function signalFlushDb() is called. + *----------------------------------------------------------------------------*/ + +void signalModifiedKey(redisDb *db, robj *key) { + touchWatchedKey(db,key); +} + +void signalFlushedDb(int dbid) { + touchWatchedKeysOnFlush(dbid); +} + /*----------------------------------------------------------------------------- * Type agnostic commands operating on the key space *----------------------------------------------------------------------------*/ void flushdbCommand(redisClient *c) { server.dirty += dictSize(c->db->dict); - touchWatchedKeysOnFlush(c->db->id); + signalFlushedDb(c->db->id); dictEmpty(c->db->dict); dictEmpty(c->db->expires); addReply(c,shared.ok); } void flushallCommand(redisClient *c) { - touchWatchedKeysOnFlush(-1); + signalFlushedDb(-1); server.dirty += emptyDb(); addReply(c,shared.ok); - if (server.bgsavechildpid != -1) { - kill(server.bgsavechildpid,SIGKILL); - rdbRemoveTempFile(server.bgsavechildpid); + if (server.rdb_child_pid != -1) { + kill(server.rdb_child_pid,SIGKILL); + rdbRemoveTempFile(server.rdb_child_pid); + } + if (server.saveparamslen > 0) { + /* Normally rdbSave() will reset dirty, but we don't want this here + * as otherwise FLUSHALL will not be replicated nor put into the AOF. */ + int saved_dirty = server.dirty; + rdbSave(server.rdb_filename); + server.dirty = saved_dirty; } - rdbSave(server.dbfilename); server.dirty++; } @@ -179,7 +208,7 @@ void delCommand(redisClient *c) { for (j = 1; j < c->argc; j++) { if (dbDelete(c->db,c->argv[j])) { - touchWatchedKey(c->db,c->argv[j]); + signalModifiedKey(c->db,c->argv[j]); server.dirty++; deleted++; } @@ -200,7 +229,7 @@ void selectCommand(redisClient *c) { int id = atoi(c->argv[1]->ptr); if (selectDb(c,id) == REDIS_ERR) { - addReplySds(c,sdsnew("-ERR invalid DB index\r\n")); + addReplyError(c,"invalid DB index"); } else { addReply(c,shared.ok); } @@ -222,19 +251,17 @@ void keysCommand(redisClient *c) { dictIterator *di; dictEntry *de; sds pattern = c->argv[1]->ptr; - int plen = sdslen(pattern); + int plen = sdslen(pattern), allkeys; unsigned long numkeys = 0; - robj *lenobj = createObject(REDIS_STRING,NULL); + void *replylen = addDeferredMultiBulkLength(c); - di = dictGetIterator(c->db->dict); - addReply(c,lenobj); - decrRefCount(lenobj); + di = dictGetSafeIterator(c->db->dict); + allkeys = (pattern[0] == '*' && pattern[1] == '\0'); while((de = dictNext(di)) != NULL) { - sds key = dictGetEntryKey(de); + sds key = dictGetKey(de); robj *keyobj; - if ((pattern[0] == '*' && pattern[1] == '\0') || - stringmatchlen(pattern,plen,key,sdslen(key),0)) { + if (allkeys || stringmatchlen(pattern,plen,key,sdslen(key),0)) { keyobj = createStringObject(key,sdslen(key)); if (expireIfNeeded(c->db,keyobj) == 0) { addReplyBulk(c,keyobj); @@ -244,17 +271,15 @@ void keysCommand(redisClient *c) { } } dictReleaseIterator(di); - lenobj->ptr = sdscatprintf(sdsempty(),"*%lu\r\n",numkeys); + setDeferredMultiBulkLength(c,replylen,numkeys); } void dbsizeCommand(redisClient *c) { - addReplySds(c, - sdscatprintf(sdsempty(),":%lu\r\n",dictSize(c->db->dict))); + addReplyLongLong(c,dictSize(c->db->dict)); } void lastsaveCommand(redisClient *c) { - addReplySds(c, - sdscatprintf(sdsempty(),":%lu\r\n",server.lastsave)); + addReplyLongLong(c,server.lastsave); } void typeCommand(redisClient *c) { @@ -263,54 +288,43 @@ void typeCommand(redisClient *c) { o = lookupKeyRead(c->db,c->argv[1]); if (o == NULL) { - type = "+none"; + type = "none"; } else { switch(o->type) { - case REDIS_STRING: type = "+string"; break; - case REDIS_LIST: type = "+list"; break; - case REDIS_SET: type = "+set"; break; - case REDIS_ZSET: type = "+zset"; break; - case REDIS_HASH: type = "+hash"; break; - default: type = "+unknown"; break; + case REDIS_STRING: type = "string"; break; + case REDIS_LIST: type = "list"; break; + case REDIS_SET: type = "set"; break; + case REDIS_ZSET: type = "zset"; break; + case REDIS_HASH: type = "hash"; break; + default: type = "unknown"; break; } } - addReplySds(c,sdsnew(type)); - addReply(c,shared.crlf); + addReplyStatus(c,type); } -void saveCommand(redisClient *c) { - if (server.bgsavechildpid != -1) { - addReplySds(c,sdsnew("-ERR background save in progress\r\n")); - return; - } - if (rdbSave(server.dbfilename) == REDIS_OK) { - addReply(c,shared.ok); - } else { - addReply(c,shared.err); - } -} +void shutdownCommand(redisClient *c) { + int flags = 0; -void bgsaveCommand(redisClient *c) { - if (server.bgsavechildpid != -1) { - addReplySds(c,sdsnew("-ERR background save already in progress\r\n")); + if (c->argc > 2) { + addReply(c,shared.syntaxerr); return; + } else if (c->argc == 2) { + if (!strcasecmp(c->argv[1]->ptr,"nosave")) { + flags |= REDIS_SHUTDOWN_NOSAVE; + } else if (!strcasecmp(c->argv[1]->ptr,"save")) { + flags |= REDIS_SHUTDOWN_SAVE; + } else { + addReply(c,shared.syntaxerr); + return; + } } - if (rdbSaveBackground(server.dbfilename) == REDIS_OK) { - char *status = "+Background saving started\r\n"; - addReplySds(c,sdsnew(status)); - } else { - addReply(c,shared.err); - } -} - -void shutdownCommand(redisClient *c) { - if (prepareForShutdown() == REDIS_OK) - exit(0); - addReplySds(c, sdsnew("-ERR Errors trying to SHUTDOWN. Check logs.\r\n")); + if (prepareForShutdown(flags) == REDIS_OK) exit(0); + addReplyError(c,"Errors trying to SHUTDOWN. Check logs."); } void renameGenericCommand(redisClient *c, int nx) { robj *o; + long long expire; /* To use the same key as src and dst is probably an error */ if (sdscmp(c->argv[1]->ptr,c->argv[2]->ptr) == 0) { @@ -322,17 +336,21 @@ void renameGenericCommand(redisClient *c, int nx) { return; incrRefCount(o); - deleteIfVolatile(c->db,c->argv[2]); - if (dbAdd(c->db,c->argv[2],o) == REDIS_ERR) { + expire = getExpire(c->db,c->argv[1]); + if (lookupKeyWrite(c->db,c->argv[2]) != NULL) { if (nx) { decrRefCount(o); addReply(c,shared.czero); return; } - dbReplace(c->db,c->argv[2],o); + /* Overwrite: delete the old key before creating the new one with the same name. */ + dbDelete(c->db,c->argv[2]); } + dbAdd(c->db,c->argv[2],o); + if (expire != -1) setExpire(c->db,c->argv[2],expire); dbDelete(c->db,c->argv[1]); - touchWatchedKey(c->db,c->argv[2]); + signalModifiedKey(c->db,c->argv[1]); + signalModifiedKey(c->db,c->argv[2]); server.dirty++; addReply(c,nx ? shared.cone : shared.ok); } @@ -374,12 +392,12 @@ void moveCommand(redisClient *c) { return; } - /* Try to add the element to the target DB */ - deleteIfVolatile(dst,c->argv[1]); - if (dbAdd(dst,c->argv[1],o) == REDIS_ERR) { + /* Return zero if the key already exists in the target DB */ + if (lookupKeyWrite(dst,c->argv[1]) != NULL) { addReply(c,shared.czero); return; } + dbAdd(dst,c->argv[1],o); incrRefCount(o); /* OK! key moved, free the entry in the source DB */ @@ -395,29 +413,23 @@ void moveCommand(redisClient *c) { int removeExpire(redisDb *db, robj *key) { /* An expire may only be removed if there is a corresponding entry in the * main dict. Otherwise, the key will never be freed. */ - redisAssert(dictFind(db->dict,key->ptr) != NULL); - if (dictDelete(db->expires,key->ptr) == DICT_OK) { - return 1; - } else { - return 0; - } + redisAssertWithInfo(NULL,key,dictFind(db->dict,key->ptr) != NULL); + return dictDelete(db->expires,key->ptr) == DICT_OK; } -int setExpire(redisDb *db, robj *key, time_t when) { - dictEntry *de; +void setExpire(redisDb *db, robj *key, long long when) { + dictEntry *kde, *de; /* Reuse the sds from the main dict in the expire dict */ - redisAssert((de = dictFind(db->dict,key->ptr)) != NULL); - if (dictAdd(db->expires,dictGetEntryKey(de),(void*)when) == DICT_ERR) { - return 0; - } else { - return 1; - } + kde = dictFind(db->dict,key->ptr); + redisAssertWithInfo(NULL,key,kde != NULL); + de = dictReplaceRaw(db->expires,dictGetKey(kde)); + dictSetSignedIntegerVal(de,when); } /* Return the expire time of the specified key, or -1 if no expire * is associated with this key (i.e. the key is non volatile) */ -time_t getExpire(redisDb *db, robj *key) { +long long getExpire(redisDb *db, robj *key) { dictEntry *de; /* No expire? return ASAP */ @@ -426,29 +438,60 @@ time_t getExpire(redisDb *db, robj *key) { /* The entry was found in the expire dict, this means it should also * be present in the main dict (safety check). */ - redisAssert(dictFind(db->dict,key->ptr) != NULL); - return (time_t) dictGetEntryVal(de); + redisAssertWithInfo(NULL,key,dictFind(db->dict,key->ptr) != NULL); + return dictGetSignedIntegerVal(de); } -int expireIfNeeded(redisDb *db, robj *key) { - time_t when = getExpire(db,key); - if (when < 0) return 0; +/* Propagate expires into slaves and the AOF file. + * When a key expires in the master, a DEL operation for this key is sent + * to all the slaves and the AOF file if enabled. + * + * This way the key expiry is centralized in one place, and since both + * AOF and the master->slave link guarantee operation ordering, everything + * will be consistent even if we allow write operations against expiring + * keys. */ +void propagateExpire(redisDb *db, robj *key) { + robj *argv[2]; - /* Return when this key has not expired */ - if (time(NULL) <= when) return 0; + argv[0] = shared.del; + argv[1] = key; + incrRefCount(argv[0]); + incrRefCount(argv[1]); - /* Delete the key */ - server.stat_expiredkeys++; - server.dirty++; - return dbDelete(db,key); + if (server.aof_state != REDIS_AOF_OFF) + feedAppendOnlyFile(server.delCommand,db->id,argv,2); + if (listLength(server.slaves)) + replicationFeedSlaves(server.slaves,db->id,argv,2); + + decrRefCount(argv[0]); + decrRefCount(argv[1]); } -int deleteIfVolatile(redisDb *db, robj *key) { - if (getExpire(db,key) < 0) return 0; +int expireIfNeeded(redisDb *db, robj *key) { + long long when = getExpire(db,key); + + if (when < 0) return 0; /* No expire for this key */ + + /* Don't expire anything while loading. It will be done later. */ + if (server.loading) return 0; + + /* If we are running in the context of a slave, return ASAP: + * the slave key expiration is controlled by the master that will + * send us synthesized DEL operations for expired keys. + * + * Still we try to return the right information to the caller, + * that is, 0 if we think the key should be still valid, 1 if + * we think the key is expired at this time. */ + if (server.masterhost != NULL) { + return mstime() > when; + } + + /* Return when this key has not expired */ + if (mstime() <= when) return 0; /* Delete the key */ server.stat_expiredkeys++; - server.dirty++; + propagateExpire(db,key); return dbDelete(db,key); } @@ -456,53 +499,181 @@ int deleteIfVolatile(redisDb *db, robj *key) { * Expires Commands *----------------------------------------------------------------------------*/ -void expireGenericCommand(redisClient *c, robj *key, robj *param, long offset) { +/* This is the generic command implementation for EXPIRE, PEXPIRE, EXPIREAT + * and PEXPIREAT. Because the commad second argument may be relative or absolute + * the "basetime" argument is used to signal what the base time is (either 0 + * for *AT variants of the command, or the current time for relative expires). + * + * unit is either UNIT_SECONDS or UNIT_MILLISECONDS, and is only used for + * the argv[2] parameter. The basetime is always specified in milliesconds. */ +void expireGenericCommand(redisClient *c, long long basetime, int unit) { dictEntry *de; - time_t seconds; + robj *key = c->argv[1], *param = c->argv[2]; + long long when; /* unix time in milliseconds when the key will expire. */ - if (getLongFromObjectOrReply(c, param, &seconds, NULL) != REDIS_OK) return; + if (getLongLongFromObjectOrReply(c, param, &when, NULL) != REDIS_OK) + return; - seconds -= offset; + if (unit == UNIT_SECONDS) when *= 1000; + when += basetime; de = dictFind(c->db->dict,key->ptr); if (de == NULL) { addReply(c,shared.czero); return; } - if (seconds <= 0) { - if (dbDelete(c->db,key)) server.dirty++; + /* EXPIRE with negative TTL, or EXPIREAT with a timestamp into the past + * should never be executed as a DEL when load the AOF or in the context + * of a slave instance. + * + * Instead we take the other branch of the IF statement setting an expire + * (possibly in the past) and wait for an explicit DEL from the master. */ + if (when <= mstime() && !server.loading && !server.masterhost) { + robj *aux; + + redisAssertWithInfo(c,key,dbDelete(c->db,key)); + server.dirty++; + + /* Replicate/AOF this as an explicit DEL. */ + aux = createStringObject("DEL",3); + rewriteClientCommandVector(c,2,aux,key); + decrRefCount(aux); + signalModifiedKey(c->db,key); addReply(c, shared.cone); return; } else { - time_t when = time(NULL)+seconds; - if (setExpire(c->db,key,when)) { - addReply(c,shared.cone); - server.dirty++; - } else { - addReply(c,shared.czero); - } + setExpire(c->db,key,when); + addReply(c,shared.cone); + signalModifiedKey(c->db,key); + server.dirty++; return; } } void expireCommand(redisClient *c) { - expireGenericCommand(c,c->argv[1],c->argv[2],0); + expireGenericCommand(c,mstime(),UNIT_SECONDS); } void expireatCommand(redisClient *c) { - expireGenericCommand(c,c->argv[1],c->argv[2],time(NULL)); + expireGenericCommand(c,0,UNIT_SECONDS); } -void ttlCommand(redisClient *c) { - time_t expire; - int ttl = -1; +void pexpireCommand(redisClient *c) { + expireGenericCommand(c,mstime(),UNIT_MILLISECONDS); +} + +void pexpireatCommand(redisClient *c) { + expireGenericCommand(c,0,UNIT_MILLISECONDS); +} + +void ttlGenericCommand(redisClient *c, int output_ms) { + long long expire, ttl = -1; expire = getExpire(c->db,c->argv[1]); if (expire != -1) { - ttl = (int) (expire-time(NULL)); + ttl = expire-mstime(); if (ttl < 0) ttl = -1; } - addReplySds(c,sdscatprintf(sdsempty(),":%d\r\n",ttl)); + if (ttl == -1) { + addReplyLongLong(c,-1); + } else { + addReplyLongLong(c,output_ms ? ttl : ((ttl+500)/1000)); + } +} + +void ttlCommand(redisClient *c) { + ttlGenericCommand(c, 0); +} + +void pttlCommand(redisClient *c) { + ttlGenericCommand(c, 1); } +void persistCommand(redisClient *c) { + dictEntry *de; + de = dictFind(c->db->dict,c->argv[1]->ptr); + if (de == NULL) { + addReply(c,shared.czero); + } else { + if (removeExpire(c->db,c->argv[1])) { + addReply(c,shared.cone); + server.dirty++; + } else { + addReply(c,shared.czero); + } + } +} + +/* ----------------------------------------------------------------------------- + * API to get key arguments from commands + * ---------------------------------------------------------------------------*/ + +int *getKeysUsingCommandTable(struct redisCommand *cmd,robj **argv, int argc, int *numkeys) { + int j, i = 0, last, *keys; + REDIS_NOTUSED(argv); + + if (cmd->firstkey == 0) { + *numkeys = 0; + return NULL; + } + last = cmd->lastkey; + if (last < 0) last = argc+last; + keys = zmalloc(sizeof(int)*((last - cmd->firstkey)+1)); + for (j = cmd->firstkey; j <= last; j += cmd->keystep) { + redisAssert(j < argc); + keys[i++] = j; + } + *numkeys = i; + return keys; +} + +int *getKeysFromCommand(struct redisCommand *cmd,robj **argv, int argc, int *numkeys, int flags) { + if (cmd->getkeys_proc) { + return cmd->getkeys_proc(cmd,argv,argc,numkeys,flags); + } else { + return getKeysUsingCommandTable(cmd,argv,argc,numkeys); + } +} + +void getKeysFreeResult(int *result) { + zfree(result); +} + +int *noPreloadGetKeys(struct redisCommand *cmd,robj **argv, int argc, int *numkeys, int flags) { + if (flags & REDIS_GETKEYS_PRELOAD) { + *numkeys = 0; + return NULL; + } else { + return getKeysUsingCommandTable(cmd,argv,argc,numkeys); + } +} + +int *renameGetKeys(struct redisCommand *cmd,robj **argv, int argc, int *numkeys, int flags) { + if (flags & REDIS_GETKEYS_PRELOAD) { + int *keys = zmalloc(sizeof(int)); + *numkeys = 1; + keys[0] = 1; + return keys; + } else { + return getKeysUsingCommandTable(cmd,argv,argc,numkeys); + } +} + +int *zunionInterGetKeys(struct redisCommand *cmd,robj **argv, int argc, int *numkeys, int flags) { + int i, num, *keys; + REDIS_NOTUSED(cmd); + REDIS_NOTUSED(flags); + + num = atoi(argv[2]->ptr); + /* Sanity check. Don't return any key if the command is going to + * reply with syntax error. */ + if (num > (argc-3)) { + *numkeys = 0; + return NULL; + } + keys = zmalloc(sizeof(int)*num); + for (i = 0; i < num; i++) keys[i] = 3+i; + *numkeys = num; + return keys; +}