X-Git-Url: https://git.saurik.com/redis.git/blobdiff_plain/d7ed7fd2fcd29fbba605fdd06b9498ffd85c092f..09e2d9eeba3ff65fd60f905a5bcb0f684f7a883e:/src/redis.c?ds=sidebyside diff --git a/src/redis.c b/src/redis.c index 367c667d..0990d98a 100644 --- a/src/redis.c +++ b/src/redis.c @@ -28,6 +28,7 @@ */ #include "redis.h" +#include "slowlog.h" #ifdef HAVE_BACKTRACE #include @@ -50,7 +51,6 @@ #include #include #include -#include #include /* Our shared "common" objects */ @@ -70,12 +70,12 @@ struct redisServer server; /* server global state */ struct redisCommand *commandTable; struct redisCommand redisCommandTable[] = { {"get",getCommand,2,0,NULL,1,1,1,0,0}, - {"set",setCommand,3,REDIS_CMD_DENYOOM,NULL,0,0,0,0,0}, - {"setnx",setnxCommand,3,REDIS_CMD_DENYOOM,NULL,0,0,0,0,0}, - {"setex",setexCommand,4,REDIS_CMD_DENYOOM,NULL,0,0,0,0,0}, + {"set",setCommand,3,REDIS_CMD_DENYOOM,noPreloadGetKeys,1,1,1,0,0}, + {"setnx",setnxCommand,3,REDIS_CMD_DENYOOM,noPreloadGetKeys,1,1,1,0,0}, + {"setex",setexCommand,4,REDIS_CMD_DENYOOM,noPreloadGetKeys,2,2,1,0,0}, {"append",appendCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"strlen",strlenCommand,2,0,NULL,1,1,1,0,0}, - {"del",delCommand,-2,0,NULL,0,0,0,0,0}, + {"del",delCommand,-2,0,noPreloadGetKeys,1,-1,1,0,0}, {"exists",existsCommand,2,0,NULL,1,1,1,0,0}, {"setbit",setbitCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"getbit",getbitCommand,3,0,NULL,1,1,1,0,0}, @@ -85,8 +85,8 @@ struct redisCommand redisCommandTable[] = { {"incr",incrCommand,2,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"decr",decrCommand,2,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"mget",mgetCommand,-2,0,NULL,1,-1,1,0,0}, - {"rpush",rpushCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, - {"lpush",lpushCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, + {"rpush",rpushCommand,-3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, + {"lpush",lpushCommand,-3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"rpushx",rpushxCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"lpushx",lpushxCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"linsert",linsertCommand,5,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, @@ -94,7 +94,7 @@ struct redisCommand redisCommandTable[] = { {"lpop",lpopCommand,2,0,NULL,1,1,1,0,0}, {"brpop",brpopCommand,-3,0,NULL,1,1,1,0,0}, {"brpoplpush",brpoplpushCommand,4,REDIS_CMD_DENYOOM,NULL,1,2,1,0,0}, - {"blpop",blpopCommand,-3,0,NULL,1,1,1,0,0}, + {"blpop",blpopCommand,-3,0,NULL,1,-2,1,0,0}, {"llen",llenCommand,2,0,NULL,1,1,1,0,0}, {"lindex",lindexCommand,3,0,NULL,1,1,1,0,0}, {"lset",lsetCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, @@ -102,8 +102,8 @@ struct redisCommand redisCommandTable[] = { {"ltrim",ltrimCommand,4,0,NULL,1,1,1,0,0}, {"lrem",lremCommand,4,0,NULL,1,1,1,0,0}, {"rpoplpush",rpoplpushCommand,3,REDIS_CMD_DENYOOM,NULL,1,2,1,0,0}, - {"sadd",saddCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, - {"srem",sremCommand,3,0,NULL,1,1,1,0,0}, + {"sadd",saddCommand,-3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, + {"srem",sremCommand,-3,0,NULL,1,1,1,0,0}, {"smove",smoveCommand,4,0,NULL,1,2,1,0,0}, {"sismember",sismemberCommand,3,0,NULL,1,1,1,0,0}, {"scard",scardCommand,2,0,NULL,1,1,1,0,0}, @@ -116,13 +116,13 @@ struct redisCommand redisCommandTable[] = { {"sdiff",sdiffCommand,-2,REDIS_CMD_DENYOOM,NULL,1,-1,1,0,0}, {"sdiffstore",sdiffstoreCommand,-3,REDIS_CMD_DENYOOM,NULL,2,-1,1,0,0}, {"smembers",sinterCommand,2,0,NULL,1,1,1,0,0}, - {"zadd",zaddCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, + {"zadd",zaddCommand,-4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"zincrby",zincrbyCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, - {"zrem",zremCommand,3,0,NULL,1,1,1,0,0}, + {"zrem",zremCommand,-3,0,NULL,1,1,1,0,0}, {"zremrangebyscore",zremrangebyscoreCommand,4,0,NULL,1,1,1,0,0}, {"zremrangebyrank",zremrangebyrankCommand,4,0,NULL,1,1,1,0,0}, - {"zunionstore",zunionstoreCommand,-4,REDIS_CMD_DENYOOM,zunionInterBlockClientOnSwappedKeys,0,0,0,0,0}, - {"zinterstore",zinterstoreCommand,-4,REDIS_CMD_DENYOOM,zunionInterBlockClientOnSwappedKeys,0,0,0,0,0}, + {"zunionstore",zunionstoreCommand,-4,REDIS_CMD_DENYOOM,zunionInterGetKeys,0,0,0,0,0}, + {"zinterstore",zinterstoreCommand,-4,REDIS_CMD_DENYOOM,zunionInterGetKeys,0,0,0,0,0}, {"zrange",zrangeCommand,-4,0,NULL,1,1,1,0,0}, {"zrangebyscore",zrangebyscoreCommand,-4,0,NULL,1,1,1,0,0}, {"zrevrangebyscore",zrevrangebyscoreCommand,-4,0,NULL,1,1,1,0,0}, @@ -138,7 +138,7 @@ struct redisCommand redisCommandTable[] = { {"hmset",hmsetCommand,-4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, {"hmget",hmgetCommand,-3,0,NULL,1,1,1,0,0}, {"hincrby",hincrbyCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0}, - {"hdel",hdelCommand,3,0,NULL,1,1,1,0,0}, + {"hdel",hdelCommand,-3,0,NULL,1,1,1,0,0}, {"hlen",hlenCommand,2,0,NULL,1,1,1,0,0}, {"hkeys",hkeysCommand,2,0,NULL,1,1,1,0,0}, {"hvals",hvalsCommand,2,0,NULL,1,1,1,0,0}, @@ -152,10 +152,10 @@ struct redisCommand redisCommandTable[] = { {"randomkey",randomkeyCommand,1,0,NULL,0,0,0,0,0}, {"select",selectCommand,2,0,NULL,0,0,0,0,0}, {"move",moveCommand,3,0,NULL,1,1,1,0,0}, - {"rename",renameCommand,3,0,NULL,1,1,1,0,0}, - {"renamenx",renamenxCommand,3,0,NULL,1,1,1,0,0}, - {"expire",expireCommand,3,0,NULL,0,0,0,0,0}, - {"expireat",expireatCommand,3,0,NULL,0,0,0,0,0}, + {"rename",renameCommand,3,0,renameGetKeys,1,2,1,0,0}, + {"renamenx",renamenxCommand,3,0,renameGetKeys,1,2,1,0,0}, + {"expire",expireCommand,3,0,NULL,1,1,1,0,0}, + {"expireat",expireatCommand,3,0,NULL,1,1,1,0,0}, {"keys",keysCommand,2,0,NULL,0,0,0,0,0}, {"dbsize",dbsizeCommand,1,0,NULL,0,0,0,0,0}, {"auth",authCommand,2,0,NULL,0,0,0,0,0}, @@ -168,7 +168,7 @@ struct redisCommand redisCommandTable[] = { {"lastsave",lastsaveCommand,1,0,NULL,0,0,0,0,0}, {"type",typeCommand,2,0,NULL,1,1,1,0,0}, {"multi",multiCommand,1,0,NULL,0,0,0,0,0}, - {"exec",execCommand,1,REDIS_CMD_DENYOOM,execBlockClientOnSwappedKeys,0,0,0,0,0}, + {"exec",execCommand,1,REDIS_CMD_DENYOOM,NULL,0,0,0,0,0}, {"discard",discardCommand,1,0,NULL,0,0,0,0,0}, {"sync",syncCommand,1,0,NULL,0,0,0,0,0}, {"flushdb",flushdbCommand,1,0,NULL,0,0,0,0,0}, @@ -186,32 +186,41 @@ struct redisCommand redisCommandTable[] = { {"psubscribe",psubscribeCommand,-2,0,NULL,0,0,0,0,0}, {"punsubscribe",punsubscribeCommand,-1,0,NULL,0,0,0,0,0}, {"publish",publishCommand,3,REDIS_CMD_FORCE_REPLICATION,NULL,0,0,0,0,0}, - {"watch",watchCommand,-2,0,NULL,0,0,0,0,0}, - {"unwatch",unwatchCommand,1,0,NULL,0,0,0,0,0} + {"watch",watchCommand,-2,0,noPreloadGetKeys,1,-1,1,0,0}, + {"unwatch",unwatchCommand,1,0,NULL,0,0,0,0,0}, + {"cluster",clusterCommand,-2,0,NULL,0,0,0,0,0}, + {"restore",restoreCommand,4,0,NULL,0,0,0,0,0}, + {"migrate",migrateCommand,6,0,NULL,0,0,0,0,0}, + {"dump",dumpCommand,2,0,NULL,0,0,0,0,0}, + {"object",objectCommand,-2,0,NULL,0,0,0,0,0}, + {"client",clientCommand,-2,0,NULL,0,0,0,0,0}, + {"slowlog",slowlogCommand,-2,0,NULL,0,0,0,0,0} }; /*============================ Utility functions ============================ */ -void redisLog(int level, const char *fmt, ...) { +/* Low level logging. To use only for very big messages, otherwise + * redisLog() is to prefer. */ +void redisLogRaw(int level, const char *msg) { const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING }; const char *c = ".-*#"; time_t now = time(NULL); - va_list ap; FILE *fp; char buf[64]; - char msg[REDIS_MAX_LOGMSG_LEN]; + int rawmode = (level & REDIS_LOG_RAW); + level &= 0xff; /* clear flags */ if (level < server.verbosity) return; fp = (server.logfile == NULL) ? stdout : fopen(server.logfile,"a"); if (!fp) return; - va_start(ap, fmt); - vsnprintf(msg, sizeof(msg), fmt, ap); - va_end(ap); - - strftime(buf,sizeof(buf),"%d %b %H:%M:%S",localtime(&now)); - fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg); + if (rawmode) { + fprintf(fp,"%s",msg); + } else { + strftime(buf,sizeof(buf),"%d %b %H:%M:%S",localtime(&now)); + fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg); + } fflush(fp); if (server.logfile) fclose(fp); @@ -219,6 +228,22 @@ void redisLog(int level, const char *fmt, ...) { if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg); } +/* Like redisLogRaw() but with printf-alike support. This is the funciton that + * is used across the code. The raw version is only used in order to dump + * the INFO output on crash. */ +void redisLog(int level, const char *fmt, ...) { + va_list ap; + char msg[REDIS_MAX_LOGMSG_LEN]; + + if ((level&0xff) < server.verbosity) return; + + va_start(ap, fmt); + vsnprintf(msg, sizeof(msg), fmt, ap); + va_end(ap); + + redisLogRaw(level,msg); +} + /* Redis generally does not try to recover from out of memory conditions * when allocating objects or strings, it is not clear if it will be possible * to report this condition to the client since the networking layer itself @@ -428,6 +453,17 @@ dictType keylistDictType = { dictListDestructor /* val destructor */ }; +/* Cluster nodes hash table, mapping nodes addresses 1.2.3.4:6379 to + * clusterNode structures. */ +dictType clusterNodesDictType = { + dictSdsHash, /* hash function */ + NULL, /* key dup */ + NULL, /* val dup */ + dictSdsKeyCompare, /* key compare */ + dictSdsDestructor, /* key destructor */ + NULL /* val destructor */ +}; + int htNeedsResize(dict *dict) { long long size, used; @@ -551,6 +587,10 @@ int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) { */ updateLRUClock(); + /* Record the max memory used since the server was started. */ + if (zmalloc_used_memory() > server.stat_peak_memory) + server.stat_peak_memory = zmalloc_used_memory(); + /* We received a SIGTERM, shutting down here in a safe way, as it is * not ok doing so inside the signal handler. */ if (server.shutdown_asap) { @@ -594,6 +634,14 @@ int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) { if ((server.maxidletime && !(loops % 100)) || server.bpop_blocked_clients) closeTimedoutClients(); + /* Start a scheduled AOF rewrite if this was requested by the user while + * a BGSAVE was in progress. */ + if (server.bgsavechildpid == -1 && server.bgrewritechildpid == -1 && + server.aofrewrite_scheduled) + { + rewriteAppendOnlyFileBackground(); + } + /* Check if a background saving or AOF rewrite in progress terminated. */ if (server.bgsavechildpid != -1 || server.bgrewritechildpid != -1) { int statloc; @@ -612,25 +660,11 @@ int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) { } updateDictResizePolicy(); } - } else if (server.bgsavethread != (pthread_t) -1) { - if (server.bgsavethread != (pthread_t) -1) { - int state; - - pthread_mutex_lock(&server.bgsavethread_mutex); - state = server.bgsavethread_state; - pthread_mutex_unlock(&server.bgsavethread_mutex); - - if (state == REDIS_BGSAVE_THREAD_DONE_OK || - state == REDIS_BGSAVE_THREAD_DONE_ERR) - { - backgroundSaveDoneHandler( - (state == REDIS_BGSAVE_THREAD_DONE_OK) ? 0 : 1, 0); - } - } - } else if (!server.ds_enabled) { - /* If there is not a background saving in progress check if - * we have to save now */ + } else { time_t now = time(NULL); + + /* If there is not a background saving/rewrite in progress check if + * we have to save/rewrite now */ for (j = 0; j < server.saveparamslen; j++) { struct saveparam *sp = server.saveparams+j; @@ -642,6 +676,21 @@ int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) { break; } } + + /* Trigger an AOF rewrite if needed */ + if (server.bgsavechildpid == -1 && + server.bgrewritechildpid == -1 && + server.auto_aofrewrite_perc && + server.appendonly_current_size > server.auto_aofrewrite_min_size) + { + int base = server.auto_aofrewrite_base_size ? + server.auto_aofrewrite_base_size : 1; + long long growth = (server.appendonly_current_size*100/base) - 100; + if (growth >= server.auto_aofrewrite_perc) { + redisLog(REDIS_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth); + rewriteAppendOnlyFileBackground(); + } + } } /* Expire a few keys per cycle, only if this is a master. @@ -649,14 +698,13 @@ int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) { * in order to guarantee a strict consistency. */ if (server.masterhost == NULL) activeExpireCycle(); - /* Remove a few cached objects from memory if we are over the - * configured memory limit */ - if (server.ds_enabled) cacheCron(); - /* Replication cron function -- used to reconnect to master and * to detect transfer failures. */ if (!(loops % 10)) replicationCron(); + /* Run other sub-systems specific cron jobs */ + if (server.cluster_enabled && !(loops % 10)) clusterCron(); + server.cronloops++; return 100; } @@ -669,31 +717,6 @@ void beforeSleep(struct aeEventLoop *eventLoop) { listNode *ln; redisClient *c; - /* Awake clients that got all the on disk keys they requested */ - if (server.ds_enabled && listLength(server.io_ready_clients)) { - listIter li; - - listRewind(server.io_ready_clients,&li); - while((ln = listNext(&li))) { - c = ln->value; - struct redisCommand *cmd; - - /* Resume the client. */ - listDelNode(server.io_ready_clients,ln); - c->flags &= (~REDIS_IO_WAIT); - server.cache_blocked_clients--; - aeCreateFileEvent(server.el, c->fd, AE_READABLE, - readQueryFromClient, c); - cmd = lookupCommand(c->argv[0]->ptr); - redisAssert(cmd != NULL); - call(c,cmd); - resetClient(c); - /* There may be more data to process in the input buffer. */ - if (c->querybuf && sdslen(c->querybuf) > 0) - processInputBuffer(c); - } - } - /* Try to process pending commands for clients that were just unblocked. */ while (listLength(server.unblocked_clients)) { ln = listFirst(server.unblocked_clients); @@ -786,6 +809,10 @@ void initServerConfig() { server.appendonly = 0; server.appendfsync = APPENDFSYNC_EVERYSEC; server.no_appendfsync_on_rewrite = 0; + server.auto_aofrewrite_perc = REDIS_AUTO_AOFREWRITE_PERC; + server.auto_aofrewrite_min_size = REDIS_AUTO_AOFREWRITE_MIN_SIZE; + server.auto_aofrewrite_base_size = 0; + server.aofrewrite_scheduled = 0; server.lastfsync = time(NULL); server.appendfd = -1; server.appendseldb = -1; /* Make sure the first time will not match */ @@ -800,17 +827,16 @@ void initServerConfig() { server.maxmemory = 0; server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU; server.maxmemory_samples = 3; - server.ds_enabled = 0; - server.ds_path = sdsnew("/tmp/redis.ds"); - server.cache_max_memory = 64LL*1024*1024; /* 64 MB of RAM */ - server.cache_blocked_clients = 0; server.hash_max_zipmap_entries = REDIS_HASH_MAX_ZIPMAP_ENTRIES; server.hash_max_zipmap_value = REDIS_HASH_MAX_ZIPMAP_VALUE; server.list_max_ziplist_entries = REDIS_LIST_MAX_ZIPLIST_ENTRIES; server.list_max_ziplist_value = REDIS_LIST_MAX_ZIPLIST_VALUE; server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES; + server.zset_max_ziplist_entries = REDIS_ZSET_MAX_ZIPLIST_ENTRIES; + server.zset_max_ziplist_value = REDIS_ZSET_MAX_ZIPLIST_VALUE; server.shutdown_asap = 0; - server.cache_flush_delay = 0; + server.cluster_enabled = 0; + server.cluster.configfile = zstrdup("nodes.conf"); updateLRUClock(); resetServerSaveParams(); @@ -825,7 +851,9 @@ void initServerConfig() { server.masterport = 6379; server.master = NULL; server.replstate = REDIS_REPL_NONE; + server.repl_syncio_timeout = REDIS_REPL_SYNCIO_TIMEOUT; server.repl_serve_stale_data = 1; + server.repl_down_since = -1; /* Double constants initialization */ R_Zero = 0.0; @@ -840,6 +868,10 @@ void initServerConfig() { populateCommandTable(); server.delCommand = lookupCommandByCString("del"); server.multiCommand = lookupCommandByCString("multi"); + + /* Slow log */ + server.slowlog_log_slower_than = REDIS_SLOWLOG_LOG_SLOWER_THAN; + server.slowlog_max_len = REDIS_SLOWLOG_MAX_LEN; } void initServer() { @@ -847,27 +879,28 @@ void initServer() { signal(SIGHUP, SIG_IGN); signal(SIGPIPE, SIG_IGN); - setupSigSegvAction(); + setupSignalHandlers(); if (server.syslog_enabled) { openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT, server.syslog_facility); } - server.mainthread = pthread_self(); server.clients = listCreate(); server.slaves = listCreate(); server.monitors = listCreate(); server.unblocked_clients = listCreate(); - server.cache_io_queue = listCreate(); createSharedObjects(); server.el = aeCreateEventLoop(); server.db = zmalloc(sizeof(redisDb)*server.dbnum); - server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr); - if (server.ipfd == ANET_ERR) { - redisLog(REDIS_WARNING, "Opening port: %s", server.neterr); - exit(1); + + if (server.port != 0) { + server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr); + if (server.ipfd == ANET_ERR) { + redisLog(REDIS_WARNING, "Opening port: %s", server.neterr); + exit(1); + } } if (server.unixsocket != NULL) { unlink(server.unixsocket); /* don't care if this fails */ @@ -886,11 +919,6 @@ void initServer() { server.db[j].expires = dictCreate(&keyptrDictType,NULL); server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL); server.db[j].watched_keys = dictCreate(&keylistDictType,NULL); - if (server.ds_enabled) { - server.db[j].io_keys = dictCreate(&keylistDictType,NULL); - server.db[j].io_negcache = dictCreate(&setDictType,NULL); - server.db[j].io_queued = dictCreate(&setDictType,NULL); - } server.db[j].id = j; } server.pubsub_channels = dictCreate(&keylistDictType,NULL); @@ -900,8 +928,6 @@ void initServer() { server.cronloops = 0; server.bgsavechildpid = -1; server.bgrewritechildpid = -1; - server.bgsavethread_state = REDIS_BGSAVE_THREAD_UNACTIVE; - server.bgsavethread = (pthread_t) -1; server.bgrewritebuf = sdsempty(); server.aofbuf = sdsempty(); server.lastsave = time(NULL); @@ -913,6 +939,8 @@ void initServer() { server.stat_starttime = time(NULL); server.stat_keyspace_misses = 0; server.stat_keyspace_hits = 0; + server.stat_peak_memory = 0; + server.stat_fork_time = 0; server.unixtime = time(NULL); aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL); if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE, @@ -929,7 +957,9 @@ void initServer() { } } - if (server.ds_enabled) dsInit(); + if (server.cluster_enabled) clusterInit(); + slowlogInit(); + srand(time(NULL)^getpid()); } /* Populates the Redis Command Table starting from the hard coded list @@ -975,18 +1005,20 @@ struct redisCommand *lookupCommandByCString(char *s) { } /* Call() is the core of Redis execution of a command */ -void call(redisClient *c, struct redisCommand *cmd) { - long long dirty, start = ustime(); +void call(redisClient *c) { + long long dirty, start = ustime(), duration; dirty = server.dirty; - cmd->proc(c); + c->cmd->proc(c); dirty = server.dirty-dirty; - cmd->microseconds += ustime()-start; - cmd->calls++; + duration = ustime()-start; + c->cmd->microseconds += duration; + slowlogPushEntryIfNeeded(c->argv,c->argc,duration); + c->cmd->calls++; if (server.appendonly && dirty) - feedAppendOnlyFile(cmd,c->db->id,c->argv,c->argc); - if ((dirty || cmd->flags & REDIS_CMD_FORCE_REPLICATION) && + feedAppendOnlyFile(c->cmd,c->db->id,c->argv,c->argc); + if ((dirty || c->cmd->flags & REDIS_CMD_FORCE_REPLICATION) && listLength(server.slaves)) replicationFeedSlaves(server.slaves,c->db->id,c->argv,c->argc); if (listLength(server.monitors)) @@ -1003,8 +1035,6 @@ void call(redisClient *c, struct redisCommand *cmd) { * and other operations can be performed by the caller. Otherwise * if 0 is returned the client was destroied (i.e. after QUIT). */ int processCommand(redisClient *c) { - struct redisCommand *cmd; - /* The QUIT command is handled separately. Normal command procs will * go through checking for replication and QUIT will cause trouble * when FORCE_REPLICATION is enabled and would be implemented in @@ -1016,32 +1046,56 @@ int processCommand(redisClient *c) { } /* Now lookup the command and check ASAP about trivial error conditions - * such wrong arity, bad command name and so forth. */ - cmd = lookupCommand(c->argv[0]->ptr); - if (!cmd) { + * such as wrong arity, bad command name and so forth. */ + c->cmd = lookupCommand(c->argv[0]->ptr); + if (!c->cmd) { addReplyErrorFormat(c,"unknown command '%s'", (char*)c->argv[0]->ptr); return REDIS_OK; - } else if ((cmd->arity > 0 && cmd->arity != c->argc) || - (c->argc < -cmd->arity)) { + } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) || + (c->argc < -c->cmd->arity)) { addReplyErrorFormat(c,"wrong number of arguments for '%s' command", - cmd->name); + c->cmd->name); return REDIS_OK; } /* Check if the user is authenticated */ - if (server.requirepass && !c->authenticated && cmd->proc != authCommand) { + if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand) + { addReplyError(c,"operation not permitted"); return REDIS_OK; } + /* If cluster is enabled, redirect here */ + if (server.cluster_enabled && + !(c->cmd->getkeys_proc == NULL && c->cmd->firstkey == 0)) { + int hashslot; + + if (server.cluster.state != REDIS_CLUSTER_OK) { + addReplyError(c,"The cluster is down. Check with CLUSTER INFO for more information"); + return REDIS_OK; + } else { + int ask; + clusterNode *n = getNodeByQuery(c,c->cmd,c->argv,c->argc,&hashslot,&ask); + if (n == NULL) { + addReplyError(c,"Multi keys request invalid in cluster"); + return REDIS_OK; + } else if (n != server.cluster.myself) { + addReplySds(c,sdscatprintf(sdsempty(), + "-%s %d %s:%d\r\n", ask ? "ASK" : "MOVED", + hashslot,n->ip,n->port)); + return REDIS_OK; + } + } + } + /* Handle the maxmemory directive. * * First we try to free some memory if possible (if there are volatile * keys in the dataset). If there are not the only thing we can do * is returning an error. */ if (server.maxmemory) freeMemoryIfNeeded(); - if (server.maxmemory && (cmd->flags & REDIS_CMD_DENYOOM) && + if (server.maxmemory && (c->cmd->flags & REDIS_CMD_DENYOOM) && zmalloc_used_memory() > server.maxmemory) { addReplyError(c,"command not allowed when used memory > 'maxmemory'"); @@ -1051,8 +1105,10 @@ int processCommand(redisClient *c) { /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */ if ((dictSize(c->pubsub_channels) > 0 || listLength(c->pubsub_patterns) > 0) && - cmd->proc != subscribeCommand && cmd->proc != unsubscribeCommand && - cmd->proc != psubscribeCommand && cmd->proc != punsubscribeCommand) { + c->cmd->proc != subscribeCommand && + c->cmd->proc != unsubscribeCommand && + c->cmd->proc != psubscribeCommand && + c->cmd->proc != punsubscribeCommand) { addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context"); return REDIS_OK; } @@ -1061,7 +1117,7 @@ int processCommand(redisClient *c) { * we are a slave with a broken link with master. */ if (server.masterhost && server.replstate != REDIS_REPL_CONNECTED && server.repl_serve_stale_data == 0 && - cmd->proc != infoCommand && cmd->proc != slaveofCommand) + c->cmd->proc != infoCommand && c->cmd->proc != slaveofCommand) { addReplyError(c, "link with MASTER is down and slave-serve-stale-data is set to no"); @@ -1069,22 +1125,20 @@ int processCommand(redisClient *c) { } /* Loading DB? Return an error if the command is not INFO */ - if (server.loading && cmd->proc != infoCommand) { + if (server.loading && c->cmd->proc != infoCommand) { addReply(c, shared.loadingerr); return REDIS_OK; } /* Exec the command */ if (c->flags & REDIS_MULTI && - cmd->proc != execCommand && cmd->proc != discardCommand && - cmd->proc != multiCommand && cmd->proc != watchCommand) + c->cmd->proc != execCommand && c->cmd->proc != discardCommand && + c->cmd->proc != multiCommand && c->cmd->proc != watchCommand) { - queueMultiCommand(c,cmd); + queueMultiCommand(c); addReply(c,shared.queued); } else { - if (server.ds_enabled && blockClientOnSwappedKeys(c,cmd)) - return REDIS_ERR; - call(c,cmd); + call(c); } return REDIS_OK; } @@ -1101,9 +1155,7 @@ int prepareForShutdown() { kill(server.bgsavechildpid,SIGKILL); rdbRemoveTempFile(server.bgsavechildpid); } - if (server.ds_enabled) { - /* FIXME: flush all objects on disk */ - } else if (server.appendonly) { + if (server.appendonly) { /* Append only file: fsync() the AOF and exit */ aof_fsync(server.appendfd); } else if (server.saveparamslen > 0) { @@ -1173,7 +1225,6 @@ sds genRedisInfoString(char *section) { sds info = sdsempty(); time_t uptime = time(NULL)-server.stat_starttime; int j, numcommands; - char hmem[64]; struct rusage self_ru, c_ru; unsigned long lol, bib; int allsections = 0, defsections = 0; @@ -1187,7 +1238,6 @@ sds genRedisInfoString(char *section) { getrusage(RUSAGE_SELF, &self_ru); getrusage(RUSAGE_CHILDREN, &c_ru); getClientsMaxBuffers(&lol,&bib); - bytesToHuman(hmem,zmalloc_used_memory()); /* Server */ if (allsections || defsections || !strcasecmp(section,"server")) { @@ -1232,42 +1282,31 @@ sds genRedisInfoString(char *section) { /* Memory */ if (allsections || defsections || !strcasecmp(section,"memory")) { + char hmem[64]; + char peak_hmem[64]; + + bytesToHuman(hmem,zmalloc_used_memory()); + bytesToHuman(peak_hmem,server.stat_peak_memory); if (sections++) info = sdscat(info,"\r\n"); info = sdscatprintf(info, "# Memory\r\n" "used_memory:%zu\r\n" "used_memory_human:%s\r\n" "used_memory_rss:%zu\r\n" + "used_memory_peak:%zu\r\n" + "used_memory_peak_human:%s\r\n" "mem_fragmentation_ratio:%.2f\r\n" - "use_tcmalloc:%d\r\n", + "mem_allocator:%s\r\n", zmalloc_used_memory(), hmem, zmalloc_get_rss(), + server.stat_peak_memory, + peak_hmem, zmalloc_get_fragmentation_ratio(), - #ifdef USE_TCMALLOC - 1 - #else - 0 - #endif + ZMALLOC_LIB ); } - /* Allocation statistics */ - if (allsections || !strcasecmp(section,"allocstats")) { - if (sections++) info = sdscat(info,"\r\n"); - info = sdscat(info, "# Allocstats\r\nallocation_stats:"); - for (j = 0; j <= ZMALLOC_MAX_ALLOC_STAT; j++) { - size_t count = zmalloc_allocations_for_size(j); - if (count) { - if (info[sdslen(info)-1] != ':') info = sdscatlen(info,",",1); - info = sdscatprintf(info,"%s%d=%zu", - (j == ZMALLOC_MAX_ALLOC_STAT) ? ">=" : "", - j,count); - } - } - info = sdscat(info,"\r\n"); - } - /* Persistence */ if (allsections || defsections || !strcasecmp(section,"persistence")) { if (sections++) info = sdscat(info,"\r\n"); @@ -1282,11 +1321,20 @@ sds genRedisInfoString(char *section) { server.loading, server.appendonly, server.dirty, - server.bgsavechildpid != -1 || - server.bgsavethread != (pthread_t) -1, + server.bgsavechildpid != -1, server.lastsave, server.bgrewritechildpid != -1); + if (server.appendonly) { + info = sdscatprintf(info, + "aof_current_size:%lld\r\n" + "aof_base_size:%lld\r\n" + "aof_pending_rewrite:%d\r\n", + (long long) server.appendonly_current_size, + (long long) server.auto_aofrewrite_base_size, + server.aofrewrite_scheduled); + } + if (server.loading) { double perc; time_t eta, elapsed; @@ -1319,25 +1367,6 @@ sds genRedisInfoString(char *section) { } } - /* Diskstore */ - if (allsections || defsections || !strcasecmp(section,"diskstore")) { - if (sections++) info = sdscat(info,"\r\n"); - info = sdscatprintf(info, - "# Diskstore\r\n" - "ds_enabled:%d\r\n", - server.ds_enabled != 0); - if (server.ds_enabled) { - lockThreadedIO(); - info = sdscatprintf(info, - "cache_max_memory:%llu\r\n" - "cache_blocked_clients:%lu\r\n" - ,(unsigned long long) server.cache_max_memory, - (unsigned long) server.cache_blocked_clients - ); - unlockThreadedIO(); - } - } - /* Stats */ if (allsections || defsections || !strcasecmp(section,"stats")) { if (sections++) info = sdscat(info,"\r\n"); @@ -1350,7 +1379,8 @@ sds genRedisInfoString(char *section) { "keyspace_hits:%lld\r\n" "keyspace_misses:%lld\r\n" "pubsub_channels:%ld\r\n" - "pubsub_patterns:%u\r\n", + "pubsub_patterns:%u\r\n" + "latest_fork_usec:%lld\r\n", server.stat_numconnections, server.stat_numcommands, server.stat_expiredkeys, @@ -1358,7 +1388,8 @@ sds genRedisInfoString(char *section) { server.stat_keyspace_hits, server.stat_keyspace_misses, dictSize(server.pubsub_channels), - listLength(server.pubsub_patterns)); + listLength(server.pubsub_patterns), + server.stat_fork_time); } /* Replication */ @@ -1392,6 +1423,12 @@ sds genRedisInfoString(char *section) { (int)(time(NULL)-server.repl_transfer_lastio) ); } + + if (server.replstate != REDIS_REPL_CONNECTED) { + info = sdscatprintf(info, + "master_link_down_since_seconds:%ld\r\n", + (long)time(NULL)-server.repl_down_since); + } } info = sdscatprintf(info, "connected_slaves:%d\r\n", @@ -1405,8 +1442,8 @@ sds genRedisInfoString(char *section) { "# CPU\r\n" "used_cpu_sys:%.2f\r\n" "used_cpu_user:%.2f\r\n" - "used_cpu_sys_childrens:%.2f\r\n" - "used_cpu_user_childrens:%.2f\r\n", + "used_cpu_sys_children:%.2f\r\n" + "used_cpu_user_children:%.2f\r\n", (float)self_ru.ru_utime.tv_sec+(float)self_ru.ru_utime.tv_usec/1000000, (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000, (float)c_ru.ru_utime.tv_sec+(float)c_ru.ru_utime.tv_usec/1000000, @@ -1429,6 +1466,15 @@ sds genRedisInfoString(char *section) { } } + /* Clusetr */ + if (allsections || defsections || !strcasecmp(section,"cluster")) { + if (sections++) info = sdscat(info,"\r\n"); + info = sdscatprintf(info, + "# Cluster\r\n" + "cluster_enabled:%d\r\n", + server.cluster_enabled); + } + /* Key space */ if (allsections || defsections || !strcasecmp(section,"keyspace")) { if (sections++) info = sdscat(info,"\r\n"); @@ -1564,6 +1610,7 @@ void freeMemoryIfNeeded(void) { /* Finally remove the selected key. */ if (bestkey) { robj *keyobj = createStringObject(bestkey,sdslen(bestkey)); + propagateExpire(db,keyobj); dbDelete(db,keyobj); server.stat_evictedkeys++; decrRefCount(keyobj); @@ -1636,8 +1683,25 @@ void usage() { exit(1); } +void redisAsciiArt(void) { +#include "asciilogo.h" + char *buf = zmalloc(1024*16); + + snprintf(buf,1024*16,ascii_logo, + REDIS_VERSION, + redisGitSHA1(), + strtol(redisGitDirty(),NULL,10) > 0, + (sizeof(long) == 8) ? "64" : "32", + server.cluster_enabled ? "cluster" : "stand alone", + server.port, + (long) getpid() + ); + redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf); + zfree(buf); +} + int main(int argc, char **argv) { - time_t start; + long long start; initServerConfig(); if (argc == 2) { @@ -1654,19 +1718,18 @@ int main(int argc, char **argv) { if (server.daemonize) daemonize(); initServer(); if (server.daemonize) createPidFile(); + redisAsciiArt(); redisLog(REDIS_NOTICE,"Server started, Redis version " REDIS_VERSION); #ifdef __linux__ linuxOvercommitMemoryWarning(); #endif - start = time(NULL); - if (server.ds_enabled) { - redisLog(REDIS_NOTICE,"DB not loaded (running with disk back end)"); - } else if (server.appendonly) { + start = ustime(); + if (server.appendonly) { if (loadAppendOnlyFile(server.appendfilename) == REDIS_OK) - redisLog(REDIS_NOTICE,"DB loaded from append only file: %ld seconds",time(NULL)-start); + redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000); } else { if (rdbLoad(server.dbfilename) == REDIS_OK) - redisLog(REDIS_NOTICE,"DB loaded from disk: %ld seconds",time(NULL)-start); + redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",(float)(ustime()-start)/1000000); } if (server.ipfd > 0) redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port); @@ -1678,10 +1741,8 @@ int main(int argc, char **argv) { return 0; } -/* ============================= Backtrace support ========================= */ - #ifdef HAVE_BACKTRACE -void *getMcontextEip(ucontext_t *uc) { +static void *getMcontextEip(ucontext_t *uc) { #if defined(__FreeBSD__) return (void*) uc->uc_mcontext.mc_eip; #elif defined(__dietlibc__) @@ -1709,7 +1770,7 @@ void *getMcontextEip(ucontext_t *uc) { #endif } -void segvHandler(int sig, siginfo_t *info, void *secret) { +static void sigsegvHandler(int sig, siginfo_t *info, void *secret) { void *trace[100]; char **messages = NULL; int i, trace_size = 0; @@ -1721,7 +1782,7 @@ void segvHandler(int sig, siginfo_t *info, void *secret) { redisLog(REDIS_WARNING, "======= Ooops! Redis %s got signal: -%d- =======", REDIS_VERSION, sig); infostring = genRedisInfoString("all"); - redisLog(REDIS_WARNING, "%s",infostring); + redisLogRaw(REDIS_WARNING, infostring); /* It's not safe to sdsfree() the returned string under memory * corruption conditions. Let it leak as we are going to abort */ @@ -1748,37 +1809,35 @@ void segvHandler(int sig, siginfo_t *info, void *secret) { sigaction (sig, &act, NULL); kill(getpid(),sig); } +#endif /* HAVE_BACKTRACE */ -void sigtermHandler(int sig) { +static void sigtermHandler(int sig) { REDIS_NOTUSED(sig); - redisLog(REDIS_WARNING,"SIGTERM received, scheduling shutting down..."); + redisLog(REDIS_WARNING,"Received SIGTERM, scheduling shutdown..."); server.shutdown_asap = 1; } -void setupSigSegvAction(void) { +void setupSignalHandlers(void) { struct sigaction act; - sigemptyset (&act.sa_mask); - /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction - * is used. Otherwise, sa_handler is used */ - act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND | SA_SIGINFO; - act.sa_sigaction = segvHandler; - sigaction (SIGSEGV, &act, NULL); - sigaction (SIGBUS, &act, NULL); - sigaction (SIGFPE, &act, NULL); - sigaction (SIGILL, &act, NULL); - sigaction (SIGBUS, &act, NULL); - + /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used. + * Otherwise, sa_handler is used. */ + sigemptyset(&act.sa_mask); act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND; act.sa_handler = sigtermHandler; - sigaction (SIGTERM, &act, NULL); - return; -} + sigaction(SIGTERM, &act, NULL); -#else /* HAVE_BACKTRACE */ -void setupSigSegvAction(void) { +#ifdef HAVE_BACKTRACE + sigemptyset(&act.sa_mask); + act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND | SA_SIGINFO; + act.sa_sigaction = sigsegvHandler; + sigaction(SIGSEGV, &act, NULL); + sigaction(SIGBUS, &act, NULL); + sigaction(SIGFPE, &act, NULL); + sigaction(SIGILL, &act, NULL); +#endif + return; } -#endif /* HAVE_BACKTRACE */ /* The End */