redis.c 66.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
/*
 * Copyright (c) 2009-2010, Salvatore Sanfilippo <antirez at gmail dot com>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include "redis.h"

#ifdef HAVE_BACKTRACE
#include <execinfo.h>
#include <ucontext.h>
#endif /* HAVE_BACKTRACE */

#include <time.h>
#include <signal.h>
#include <sys/wait.h>
#include <errno.h>
#include <assert.h>
#include <ctype.h>
#include <stdarg.h>
#include <arpa/inet.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <sys/time.h>
#include <sys/resource.h>
#include <sys/uio.h>
#include <limits.h>
#include <float.h>
#include <math.h>
#include <pthread.h>
54
#include <sys/resource.h>
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

/* Our shared "common" objects */

struct sharedObjectsStruct shared;

/* Global vars that are actally used as constants. The following double
 * values are used for double on-disk serialization, and are initialized
 * at runtime to avoid strange compiler optimizations. */

double R_Zero, R_PosInf, R_NegInf, R_Nan;

/*================================= Globals ================================= */

/* Global vars */
struct redisServer server; /* server global state */
struct redisCommand *commandTable;
71
struct redisCommand redisCommandTable[] = {
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190
    {"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},
    {"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},
    {"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},
    {"setrange",setrangeCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"getrange",getrangeCommand,4,0,NULL,1,1,1,0,0},
    {"substr",getrangeCommand,4,0,NULL,1,1,1,0,0},
    {"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},
    {"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},
    {"rpop",rpopCommand,2,0,NULL,1,1,1,0,0},
    {"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},
    {"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},
    {"lrange",lrangeCommand,4,0,NULL,1,1,1,0,0},
    {"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},
    {"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},
    {"spop",spopCommand,2,0,NULL,1,1,1,0,0},
    {"srandmember",srandmemberCommand,2,0,NULL,1,1,1,0,0},
    {"sinter",sinterCommand,-2,REDIS_CMD_DENYOOM,NULL,1,-1,1,0,0},
    {"sinterstore",sinterstoreCommand,-3,REDIS_CMD_DENYOOM,NULL,2,-1,1,0,0},
    {"sunion",sunionCommand,-2,REDIS_CMD_DENYOOM,NULL,1,-1,1,0,0},
    {"sunionstore",sunionstoreCommand,-3,REDIS_CMD_DENYOOM,NULL,2,-1,1,0,0},
    {"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},
    {"zincrby",zincrbyCommand,4,REDIS_CMD_DENYOOM,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},
    {"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},
    {"zcount",zcountCommand,4,0,NULL,1,1,1,0,0},
    {"zrevrange",zrevrangeCommand,-4,0,NULL,1,1,1,0,0},
    {"zcard",zcardCommand,2,0,NULL,1,1,1,0,0},
    {"zscore",zscoreCommand,3,0,NULL,1,1,1,0,0},
    {"zrank",zrankCommand,3,0,NULL,1,1,1,0,0},
    {"zrevrank",zrevrankCommand,3,0,NULL,1,1,1,0,0},
    {"hset",hsetCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"hsetnx",hsetnxCommand,4,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"hget",hgetCommand,3,0,NULL,1,1,1,0,0},
    {"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},
    {"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},
    {"hgetall",hgetallCommand,2,0,NULL,1,1,1,0,0},
    {"hexists",hexistsCommand,3,0,NULL,1,1,1,0,0},
    {"incrby",incrbyCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"decrby",decrbyCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"getset",getsetCommand,3,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"mset",msetCommand,-3,REDIS_CMD_DENYOOM,NULL,1,-1,2,0,0},
    {"msetnx",msetnxCommand,-3,REDIS_CMD_DENYOOM,NULL,1,-1,2,0,0},
    {"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},
    {"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},
    {"ping",pingCommand,1,0,NULL,0,0,0,0,0},
    {"echo",echoCommand,2,0,NULL,0,0,0,0,0},
    {"save",saveCommand,1,0,NULL,0,0,0,0,0},
    {"bgsave",bgsaveCommand,1,0,NULL,0,0,0,0,0},
    {"bgrewriteaof",bgrewriteaofCommand,1,0,NULL,0,0,0,0,0},
    {"shutdown",shutdownCommand,1,0,NULL,0,0,0,0,0},
    {"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},
    {"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},
    {"flushall",flushallCommand,1,0,NULL,0,0,0,0,0},
    {"sort",sortCommand,-2,REDIS_CMD_DENYOOM,NULL,1,1,1,0,0},
    {"info",infoCommand,-1,0,NULL,0,0,0,0,0},
    {"monitor",monitorCommand,1,0,NULL,0,0,0,0,0},
    {"ttl",ttlCommand,2,0,NULL,1,1,1,0,0},
    {"persist",persistCommand,2,0,NULL,1,1,1,0,0},
    {"slaveof",slaveofCommand,3,0,NULL,0,0,0,0,0},
    {"debug",debugCommand,-2,0,NULL,0,0,0,0,0},
    {"config",configCommand,-2,0,NULL,0,0,0,0,0},
    {"subscribe",subscribeCommand,-2,0,NULL,0,0,0,0,0},
    {"unsubscribe",unsubscribeCommand,-1,0,NULL,0,0,0,0,0},
    {"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}
191 192 193 194
};

/*============================ Utility functions ============================ */

195 196 197
/* Low level logging. To use only for very big messages, otherwise
 * redisLog() is to prefer. */
void redisLogRaw(int level, const char *msg) {
J
Jonah H. Harris 已提交
198 199 200
    const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
    const char *c = ".-*#";
    time_t now = time(NULL);
201
    FILE *fp;
202 203 204
    char buf[64];

    if (level < server.verbosity) return;
205 206 207 208

    fp = (server.logfile == NULL) ? stdout : fopen(server.logfile,"a");
    if (!fp) return;

J
Jonah H. Harris 已提交
209 210 211 212
    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);

213
    if (server.logfile) fclose(fp);
J
Jonah H. Harris 已提交
214 215

    if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
216 217
}

218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233
/* 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 < server.verbosity) return;

    va_start(ap, fmt);
    vsnprintf(msg, sizeof(msg), fmt, ap);
    va_end(ap);

    redisLogRaw(level,msg);
}

234 235 236 237 238 239 240 241 242 243 244
/* 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
 * is based on heap allocation for send buffers, so we simply abort.
 * At least the code will be simpler to read... */
void oom(const char *msg) {
    redisLog(REDIS_WARNING, "%s: Out of memory\n",msg);
    sleep(1);
    abort();
}

A
antirez 已提交
245 246 247 248 249 250 251 252 253 254 255
/* Return the UNIX time in microseconds */
long long ustime(void) {
    struct timeval tv;
    long long ust;

    gettimeofday(&tv, NULL);
    ust = ((long long)tv.tv_sec)*1000000;
    ust += tv.tv_usec;
    return ust;
}

256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285
/*====================== Hash table type implementation  ==================== */

/* This is an hash table type that uses the SDS dynamic strings libary as
 * keys and radis objects as values (objects can hold SDS strings,
 * lists, sets). */

void dictVanillaFree(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);
    zfree(val);
}

void dictListDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);
    listRelease((list*)val);
}

int dictSdsKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    int l1,l2;
    DICT_NOTUSED(privdata);

    l1 = sdslen((sds)key1);
    l2 = sdslen((sds)key2);
    if (l1 != l2) return 0;
    return memcmp(key1, key2, l1) == 0;
}

286 287 288 289 290 291 292 293 294
/* A case insensitive version used for the command lookup table. */
int dictSdsKeyCaseCompare(void *privdata, const void *key1,
        const void *key2)
{
    DICT_NOTUSED(privdata);

    return strcasecmp(key1, key2) == 0;
}

295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
void dictRedisObjectDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);

    if (val == NULL) return; /* Values of swapped out keys as set to NULL */
    decrRefCount(val);
}

void dictSdsDestructor(void *privdata, void *val)
{
    DICT_NOTUSED(privdata);

    sdsfree(val);
}

int dictObjKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    const robj *o1 = key1, *o2 = key2;
    return dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
}

unsigned int dictObjHash(const void *key) {
    const robj *o = key;
    return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
}

unsigned int dictSdsHash(const void *key) {
    return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
}

326 327 328 329
unsigned int dictSdsCaseHash(const void *key) {
    return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
}

330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
int dictEncObjKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    robj *o1 = (robj*) key1, *o2 = (robj*) key2;
    int cmp;

    if (o1->encoding == REDIS_ENCODING_INT &&
        o2->encoding == REDIS_ENCODING_INT)
            return o1->ptr == o2->ptr;

    o1 = getDecodedObject(o1);
    o2 = getDecodedObject(o2);
    cmp = dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
    decrRefCount(o1);
    decrRefCount(o2);
    return cmp;
}

unsigned int dictEncObjHash(const void *key) {
    robj *o = (robj*) key;

    if (o->encoding == REDIS_ENCODING_RAW) {
        return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
    } else {
        if (o->encoding == REDIS_ENCODING_INT) {
            char buf[32];
            int len;

            len = ll2string(buf,32,(long)o->ptr);
            return dictGenHashFunction((unsigned char*)buf, len);
        } else {
            unsigned int hash;

            o = getDecodedObject(o);
            hash = dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
            decrRefCount(o);
            return hash;
        }
    }
}

A
antirez 已提交
371
/* Sets type and diskstore negative caching hash table */
372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
dictType setDictType = {
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
    dictRedisObjectDestructor, /* key destructor */
    NULL                       /* val destructor */
};

/* Sorted sets hash (note: a skiplist is used in addition to the hash table) */
dictType zsetDictType = {
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
    dictRedisObjectDestructor, /* key destructor */
388
    NULL                       /* val destructor */
389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410
};

/* Db->dict, keys are sds strings, vals are Redis objects. */
dictType dbDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
};

/* Db->expires */
dictType keyptrDictType = {
    dictSdsHash,               /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictSdsKeyCompare,         /* key compare */
    NULL,                      /* key destructor */
    NULL                       /* val destructor */
};

411 412 413 414 415 416 417 418 419 420
/* Command table. sds string -> command struct pointer. */
dictType commandTableDictType = {
    dictSdsCaseHash,           /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictSdsKeyCaseCompare,     /* key compare */
    dictSdsDestructor,         /* key destructor */
    NULL                       /* val destructor */
};

421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494
/* Hash type hash table (note that small hashes are represented with zimpaps) */
dictType hashDictType = {
    dictEncObjHash,             /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictEncObjKeyCompare,       /* key compare */
    dictRedisObjectDestructor,  /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
};

/* Keylist hash table type has unencoded redis objects as keys and
 * lists as values. It's used for blocking operations (BLPOP) and to
 * map swapped keys to a list of clients waiting for this keys to be loaded. */
dictType keylistDictType = {
    dictObjHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictObjKeyCompare,          /* key compare */
    dictRedisObjectDestructor,  /* key destructor */
    dictListDestructor          /* val destructor */
};

int htNeedsResize(dict *dict) {
    long long size, used;

    size = dictSlots(dict);
    used = dictSize(dict);
    return (size && used && size > DICT_HT_INITIAL_SIZE &&
            (used*100/size < REDIS_HT_MINFILL));
}

/* If the percentage of used slots in the HT reaches REDIS_HT_MINFILL
 * we resize the hash table to save memory */
void tryResizeHashTables(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        if (htNeedsResize(server.db[j].dict))
            dictResize(server.db[j].dict);
        if (htNeedsResize(server.db[j].expires))
            dictResize(server.db[j].expires);
    }
}

/* Our hash table implementation performs rehashing incrementally while
 * we write/read from the hash table. Still if the server is idle, the hash
 * table will use two tables for a long time. So we try to use 1 millisecond
 * of CPU time at every serverCron() loop in order to rehash some key. */
void incrementallyRehash(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        if (dictIsRehashing(server.db[j].dict)) {
            dictRehashMilliseconds(server.db[j].dict,1);
            break; /* already used our millisecond for this loop... */
        }
    }
}

/* This function is called once a background process of some kind terminates,
 * as we want to avoid resizing the hash tables when there is a child in order
 * to play well with copy-on-write (otherwise when a resize happens lots of
 * memory pages are copied). The goal of this function is to update the ability
 * for dict.c to resize the hash tables accordingly to the fact we have o not
 * running childs. */
void updateDictResizePolicy(void) {
    if (server.bgsavechildpid == -1 && server.bgrewritechildpid == -1)
        dictEnableResize();
    else
        dictDisableResize();
}

/* ======================= Cron: called every 100 ms ======================== */

495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535
/* Try to expire a few timed out keys. The algorithm used is adaptive and
 * will use few CPU cycles if there are few expiring keys, otherwise
 * it will get more aggressive to avoid that too much memory is used by
 * keys that can be removed from the keyspace. */
void activeExpireCycle(void) {
    int j;

    for (j = 0; j < server.dbnum; j++) {
        int expired;
        redisDb *db = server.db+j;

        /* Continue to expire if at the end of the cycle more than 25%
         * of the keys were expired. */
        do {
            long num = dictSize(db->expires);
            time_t now = time(NULL);

            expired = 0;
            if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
                num = REDIS_EXPIRELOOKUPS_PER_CRON;
            while (num--) {
                dictEntry *de;
                time_t t;

                if ((de = dictGetRandomKey(db->expires)) == NULL) break;
                t = (time_t) dictGetEntryVal(de);
                if (now > t) {
                    sds key = dictGetEntryKey(de);
                    robj *keyobj = createStringObject(key,sdslen(key));

                    propagateExpire(db,keyobj);
                    dbDelete(db,keyobj);
                    decrRefCount(keyobj);
                    expired++;
                    server.stat_expiredkeys++;
                }
            }
        } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
    }
}

536 537 538 539
void updateLRUClock(void) {
    server.lruclock = (time(NULL)/REDIS_LRU_CLOCK_RESOLUTION) &
                                                REDIS_LRU_CLOCK_MAX;
}
540

541
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
542
    int j, loops = server.cronloops;
543 544 545 546 547 548 549 550 551
    REDIS_NOTUSED(eventLoop);
    REDIS_NOTUSED(id);
    REDIS_NOTUSED(clientData);

    /* We take a cached value of the unix time in the global state because
     * with virtual memory and aging there is to store the current time
     * in objects at every object access, and accuracy is not needed.
     * To access a global var is faster than calling time(NULL) */
    server.unixtime = time(NULL);
552
    /* We have just 22 bits per object for LRU information.
553 554
     * So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
     * 2^22 bits with 10 seconds resoluton is more or less 1.5 years.
555
     *
556
     * Note that even if this will wrap after 1.5 years it's not a problem,
557
     * everything will still work but just some object will appear younger
558 559 560 561 562
     * to Redis. But for this to happen a given object should never be touched
     * for 1.5 years.
     *
     * Note that you can change the resolution altering the
     * REDIS_LRU_CLOCK_RESOLUTION define.
563
     */
564
    updateLRUClock();
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601

    /* 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) {
        if (prepareForShutdown() == REDIS_OK) exit(0);
        redisLog(REDIS_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
    }

    /* Show some info about non-empty databases */
    for (j = 0; j < server.dbnum; j++) {
        long long size, used, vkeys;

        size = dictSlots(server.db[j].dict);
        used = dictSize(server.db[j].dict);
        vkeys = dictSize(server.db[j].expires);
        if (!(loops % 50) && (used || vkeys)) {
            redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
            /* dictPrintStats(server.dict); */
        }
    }

    /* We don't want to resize the hash tables while a bacground saving
     * is in progress: the saving child is created using fork() that is
     * implemented with a copy-on-write semantic in most modern systems, so
     * if we resize the HT while there is the saving child at work actually
     * a lot of memory movements in the parent will cause a lot of pages
     * copied. */
    if (server.bgsavechildpid == -1 && server.bgrewritechildpid == -1) {
        if (!(loops % 10)) tryResizeHashTables();
        if (server.activerehashing) incrementallyRehash();
    }

    /* Show information about connected clients */
    if (!(loops % 50)) {
        redisLog(REDIS_VERBOSE,"%d clients connected (%d slaves), %zu bytes in use",
            listLength(server.clients)-listLength(server.slaves),
            listLength(server.slaves),
602
            zmalloc_used_memory());
603 604 605
    }

    /* Close connections of timedout clients */
606
    if ((server.maxidletime && !(loops % 100)) || server.bpop_blocked_clients)
607 608
        closeTimedoutClients();

A
antirez 已提交
609
    /* Check if a background saving or AOF rewrite in progress terminated. */
610 611 612 613 614
    if (server.bgsavechildpid != -1 || server.bgrewritechildpid != -1) {
        int statloc;
        pid_t pid;

        if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
615 616 617 618 619
            int exitcode = WEXITSTATUS(statloc);
            int bysignal = 0;
            
            if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);

620
            if (pid == server.bgsavechildpid) {
621
                backgroundSaveDoneHandler(exitcode,bysignal);
622
            } else {
623
                backgroundRewriteDoneHandler(exitcode,bysignal);
624 625 626
            }
            updateDictResizePolicy();
        }
A
antirez 已提交
627
    } else if (server.bgsavethread != (pthread_t) -1) {
628 629 630 631 632 633 634
        if (server.bgsavethread != (pthread_t) -1) {
            int state;

            pthread_mutex_lock(&server.bgsavethread_mutex);
            state = server.bgsavethread_state;
            pthread_mutex_unlock(&server.bgsavethread_mutex);

A
antirez 已提交
635 636
            if (state == REDIS_BGSAVE_THREAD_DONE_OK ||
                state == REDIS_BGSAVE_THREAD_DONE_ERR)
637 638
            {
                backgroundSaveDoneHandler(
A
antirez 已提交
639
                    (state == REDIS_BGSAVE_THREAD_DONE_OK) ? 0 : 1, 0);
640 641
            }
        }
642
    } else if (!server.ds_enabled) {
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658
        /* If there is not a background saving in progress check if
         * we have to save now */
         time_t now = time(NULL);
         for (j = 0; j < server.saveparamslen; j++) {
            struct saveparam *sp = server.saveparams+j;

            if (server.dirty >= sp->changes &&
                now-server.lastsave > sp->seconds) {
                redisLog(REDIS_NOTICE,"%d changes in %d seconds. Saving...",
                    sp->changes, sp->seconds);
                rdbSaveBackground(server.dbfilename);
                break;
            }
         }
    }

659 660 661 662
    /* Expire a few keys per cycle, only if this is a master.
     * On slaves we wait for DEL operations synthesized by the master
     * in order to guarantee a strict consistency. */
    if (server.masterhost == NULL) activeExpireCycle();
663

664 665
    /* Remove a few cached objects from memory if we are over the
     * configured memory limit */
666
    if (server.ds_enabled) cacheCron();
667

668 669
    /* Replication cron function -- used to reconnect to master and
     * to detect transfer failures. */
670
    if (!(loops % 10)) replicationCron();
671

672
    server.cronloops++;
673 674 675 676 677 678 679 680
    return 100;
}

/* This function gets called every time Redis is entering the
 * main loop of the event driven library, that is, before to sleep
 * for ready file descriptors. */
void beforeSleep(struct aeEventLoop *eventLoop) {
    REDIS_NOTUSED(eventLoop);
681 682
    listNode *ln;
    redisClient *c;
683

684 685
    /* Awake clients that got all the on disk keys they requested */
    if (server.ds_enabled && listLength(server.io_ready_clients)) {
686 687 688 689
        listIter li;

        listRewind(server.io_ready_clients,&li);
        while((ln = listNext(&li))) {
690
            c = ln->value;
691 692 693 694 695
            struct redisCommand *cmd;

            /* Resume the client. */
            listDelNode(server.io_ready_clients,ln);
            c->flags &= (~REDIS_IO_WAIT);
696
            server.cache_blocked_clients--;
697 698 699 700 701 702 703 704 705 706 707
            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);
        }
    }
708 709 710 711 712 713 714

    /* Try to process pending commands for clients that were just unblocked. */
    while (listLength(server.unblocked_clients)) {
        ln = listFirst(server.unblocked_clients);
        redisAssert(ln != NULL);
        c = ln->value;
        listDelNode(server.unblocked_clients,ln);
715
        c->flags &= ~REDIS_UNBLOCKED;
716 717 718 719 720 721

        /* Process remaining data in the input buffer. */
        if (c->querybuf && sdslen(c->querybuf) > 0)
            processInputBuffer(c);
    }

722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
    /* Write the AOF buffer on disk */
    flushAppendOnlyFile();
}

/* =========================== Server initialization ======================== */

void createSharedObjects(void) {
    int j;

    shared.crlf = createObject(REDIS_STRING,sdsnew("\r\n"));
    shared.ok = createObject(REDIS_STRING,sdsnew("+OK\r\n"));
    shared.err = createObject(REDIS_STRING,sdsnew("-ERR\r\n"));
    shared.emptybulk = createObject(REDIS_STRING,sdsnew("$0\r\n\r\n"));
    shared.czero = createObject(REDIS_STRING,sdsnew(":0\r\n"));
    shared.cone = createObject(REDIS_STRING,sdsnew(":1\r\n"));
    shared.cnegone = createObject(REDIS_STRING,sdsnew(":-1\r\n"));
    shared.nullbulk = createObject(REDIS_STRING,sdsnew("$-1\r\n"));
    shared.nullmultibulk = createObject(REDIS_STRING,sdsnew("*-1\r\n"));
    shared.emptymultibulk = createObject(REDIS_STRING,sdsnew("*0\r\n"));
    shared.pong = createObject(REDIS_STRING,sdsnew("+PONG\r\n"));
    shared.queued = createObject(REDIS_STRING,sdsnew("+QUEUED\r\n"));
    shared.wrongtypeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR Operation against a key holding the wrong kind of value\r\n"));
    shared.nokeyerr = createObject(REDIS_STRING,sdsnew(
        "-ERR no such key\r\n"));
    shared.syntaxerr = createObject(REDIS_STRING,sdsnew(
        "-ERR syntax error\r\n"));
    shared.sameobjecterr = createObject(REDIS_STRING,sdsnew(
        "-ERR source and destination objects are the same\r\n"));
    shared.outofrangeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR index out of range\r\n"));
753 754
    shared.loadingerr = createObject(REDIS_STRING,sdsnew(
        "-LOADING Redis is loading the dataset in memory\r\n"));
755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783
    shared.space = createObject(REDIS_STRING,sdsnew(" "));
    shared.colon = createObject(REDIS_STRING,sdsnew(":"));
    shared.plus = createObject(REDIS_STRING,sdsnew("+"));
    shared.select0 = createStringObject("select 0\r\n",10);
    shared.select1 = createStringObject("select 1\r\n",10);
    shared.select2 = createStringObject("select 2\r\n",10);
    shared.select3 = createStringObject("select 3\r\n",10);
    shared.select4 = createStringObject("select 4\r\n",10);
    shared.select5 = createStringObject("select 5\r\n",10);
    shared.select6 = createStringObject("select 6\r\n",10);
    shared.select7 = createStringObject("select 7\r\n",10);
    shared.select8 = createStringObject("select 8\r\n",10);
    shared.select9 = createStringObject("select 9\r\n",10);
    shared.messagebulk = createStringObject("$7\r\nmessage\r\n",13);
    shared.pmessagebulk = createStringObject("$8\r\npmessage\r\n",14);
    shared.subscribebulk = createStringObject("$9\r\nsubscribe\r\n",15);
    shared.unsubscribebulk = createStringObject("$11\r\nunsubscribe\r\n",18);
    shared.psubscribebulk = createStringObject("$10\r\npsubscribe\r\n",17);
    shared.punsubscribebulk = createStringObject("$12\r\npunsubscribe\r\n",19);
    shared.mbulk3 = createStringObject("*3\r\n",4);
    shared.mbulk4 = createStringObject("*4\r\n",4);
    for (j = 0; j < REDIS_SHARED_INTEGERS; j++) {
        shared.integers[j] = createObject(REDIS_STRING,(void*)(long)j);
        shared.integers[j]->encoding = REDIS_ENCODING_INT;
    }
}

void initServerConfig() {
    server.port = REDIS_SERVERPORT;
784
    server.bindaddr = NULL;
785
    server.unixsocket = NULL;
786 787 788
    server.ipfd = -1;
    server.sofd = -1;
    server.dbnum = REDIS_DEFAULT_DBNUM;
789 790 791
    server.verbosity = REDIS_VERBOSE;
    server.maxidletime = REDIS_MAXIDLETIME;
    server.saveparams = NULL;
792
    server.loading = 0;
793
    server.logfile = NULL; /* NULL = log on standard output */
J
Jonah H. Harris 已提交
794 795 796
    server.syslog_enabled = 0;
    server.syslog_ident = zstrdup("redis");
    server.syslog_facility = LOG_LOCAL0;
797 798 799 800 801 802 803 804 805 806 807 808 809 810
    server.daemonize = 0;
    server.appendonly = 0;
    server.appendfsync = APPENDFSYNC_EVERYSEC;
    server.no_appendfsync_on_rewrite = 0;
    server.lastfsync = time(NULL);
    server.appendfd = -1;
    server.appendseldb = -1; /* Make sure the first time will not match */
    server.pidfile = zstrdup("/var/run/redis.pid");
    server.dbfilename = zstrdup("dump.rdb");
    server.appendfilename = zstrdup("appendonly.aof");
    server.requirepass = NULL;
    server.rdbcompression = 1;
    server.activerehashing = 1;
    server.maxclients = 0;
811
    server.bpop_blocked_clients = 0;
812
    server.maxmemory = 0;
813 814
    server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU;
    server.maxmemory_samples = 3;
815
    server.ds_enabled = 0;
816
    server.ds_path = sdsnew("/tmp/redis.ds");
817 818
    server.cache_max_memory = 64LL*1024*1024; /* 64 MB of RAM */
    server.cache_blocked_clients = 0;
819 820 821 822
    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;
823
    server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES;
824
    server.shutdown_asap = 0;
825
    server.cache_flush_delay = 0;
826

A
antirez 已提交
827
    updateLRUClock();
828 829 830 831 832 833 834 835 836 837 838 839
    resetServerSaveParams();

    appendServerSaveParams(60*60,1);  /* save after 1 hour and 1 change */
    appendServerSaveParams(300,100);  /* save after 5 minutes and 100 changes */
    appendServerSaveParams(60,10000); /* save after 1 minute and 10000 changes */
    /* Replication related */
    server.isslave = 0;
    server.masterauth = NULL;
    server.masterhost = NULL;
    server.masterport = 6379;
    server.master = NULL;
    server.replstate = REDIS_REPL_NONE;
840
    server.repl_serve_stale_data = 1;
841 842 843 844 845 846

    /* Double constants initialization */
    R_Zero = 0.0;
    R_PosInf = 1.0/R_Zero;
    R_NegInf = -1.0/R_Zero;
    R_Nan = R_Zero/R_Zero;
847 848 849 850 851 852 853 854

    /* Command table -- we intiialize it here as it is part of the
     * initial configuration, since command names may be changed via
     * redis.conf using the rename-command directive. */
    server.commands = dictCreate(&commandTableDictType,NULL);
    populateCommandTable();
    server.delCommand = lookupCommandByCString("del");
    server.multiCommand = lookupCommandByCString("multi");
855 856 857 858 859 860 861 862 863
}

void initServer() {
    int j;

    signal(SIGHUP, SIG_IGN);
    signal(SIGPIPE, SIG_IGN);
    setupSigSegvAction();

J
Jonah H. Harris 已提交
864 865 866 867 868
    if (server.syslog_enabled) {
        openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
            server.syslog_facility);
    }

869
    server.mainthread = pthread_self();
870 871 872
    server.clients = listCreate();
    server.slaves = listCreate();
    server.monitors = listCreate();
873
    server.unblocked_clients = listCreate();
874
    server.cache_io_queue = listCreate();
875

876 877 878
    createSharedObjects();
    server.el = aeCreateEventLoop();
    server.db = zmalloc(sizeof(redisDb)*server.dbnum);
879

880
    if (server.port != 0) {
881
        server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr);
882 883 884 885
        if (server.ipfd == ANET_ERR) {
            redisLog(REDIS_WARNING, "Opening port: %s", server.neterr);
            exit(1);
        }
886
    }
887 888 889
    if (server.unixsocket != NULL) {
        unlink(server.unixsocket); /* don't care if this fails */
        server.sofd = anetUnixServer(server.neterr,server.unixsocket);
890 891 892 893
        if (server.sofd == ANET_ERR) {
            redisLog(REDIS_WARNING, "Opening socket: %s", server.neterr);
            exit(1);
        }
894
    }
895 896
    if (server.ipfd < 0 && server.sofd < 0) {
        redisLog(REDIS_WARNING, "Configured to not listen anywhere, exiting.");
897 898 899 900 901 902 903
        exit(1);
    }
    for (j = 0; j < server.dbnum; j++) {
        server.db[j].dict = dictCreate(&dbDictType,NULL);
        server.db[j].expires = dictCreate(&keyptrDictType,NULL);
        server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL);
        server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
A
antirez 已提交
904
        if (server.ds_enabled) {
905
            server.db[j].io_keys = dictCreate(&keylistDictType,NULL);
A
antirez 已提交
906
            server.db[j].io_negcache = dictCreate(&setDictType,NULL);
907
            server.db[j].io_queued = dictCreate(&setDictType,NULL);
A
antirez 已提交
908
        }
909 910 911 912 913 914 915 916 917
        server.db[j].id = j;
    }
    server.pubsub_channels = dictCreate(&keylistDictType,NULL);
    server.pubsub_patterns = listCreate();
    listSetFreeMethod(server.pubsub_patterns,freePubsubPattern);
    listSetMatchMethod(server.pubsub_patterns,listMatchPubsubPattern);
    server.cronloops = 0;
    server.bgsavechildpid = -1;
    server.bgrewritechildpid = -1;
918 919
    server.bgsavethread_state = REDIS_BGSAVE_THREAD_UNACTIVE;
    server.bgsavethread = (pthread_t) -1;
920 921 922 923 924 925 926
    server.bgrewritebuf = sdsempty();
    server.aofbuf = sdsempty();
    server.lastsave = time(NULL);
    server.dirty = 0;
    server.stat_numcommands = 0;
    server.stat_numconnections = 0;
    server.stat_expiredkeys = 0;
927
    server.stat_evictedkeys = 0;
928
    server.stat_starttime = time(NULL);
929 930
    server.stat_keyspace_misses = 0;
    server.stat_keyspace_hits = 0;
931 932
    server.unixtime = time(NULL);
    aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL);
933
    if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE,
934
        acceptTcpHandler,NULL) == AE_ERR) oom("creating file event");
935
    if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
936
        acceptUnixHandler,NULL) == AE_ERR) oom("creating file event");
937 938 939 940 941 942 943 944 945 946

    if (server.appendonly) {
        server.appendfd = open(server.appendfilename,O_WRONLY|O_APPEND|O_CREAT,0644);
        if (server.appendfd == -1) {
            redisLog(REDIS_WARNING, "Can't open the append-only file: %s",
                strerror(errno));
            exit(1);
        }
    }

947
    if (server.ds_enabled) dsInit();
A
antirez 已提交
948
    srand(time(NULL)^getpid());
949 950
}

951 952 953 954
/* Populates the Redis Command Table starting from the hard coded list
 * we have on top of redis.c file. */
void populateCommandTable(void) {
    int j;
955
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
956 957

    for (j = 0; j < numcommands; j++) {
958
        struct redisCommand *c = redisCommandTable+j;
959
        int retval;
960

961 962 963
        retval = dictAdd(server.commands, sdsnew(c->name), c);
        assert(retval == DICT_OK);
    }
964 965
}

966 967 968 969 970 971 972 973 974 975 976 977
void resetCommandTableStats(void) {
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
    int j;

    for (j = 0; j < numcommands; j++) {
        struct redisCommand *c = redisCommandTable+j;

        c->microseconds = 0;
        c->calls = 0;
    }
}

978 979
/* ====================== Commands lookup and execution ===================== */

980 981 982 983 984 985 986 987 988 989 990
struct redisCommand *lookupCommand(sds name) {
    return dictFetchValue(server.commands, name);
}

struct redisCommand *lookupCommandByCString(char *s) {
    struct redisCommand *cmd;
    sds name = sdsnew(s);

    cmd = dictFetchValue(server.commands, name);
    sdsfree(name);
    return cmd;
991 992 993 994
}

/* Call() is the core of Redis execution of a command */
void call(redisClient *c, struct redisCommand *cmd) {
A
antirez 已提交
995
    long long dirty, start = ustime();
996 997 998 999

    dirty = server.dirty;
    cmd->proc(c);
    dirty = server.dirty-dirty;
A
antirez 已提交
1000
    cmd->microseconds += ustime()-start;
1001
    cmd->calls++;
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023

    if (server.appendonly && dirty)
        feedAppendOnlyFile(cmd,c->db->id,c->argv,c->argc);
    if ((dirty || cmd->flags & REDIS_CMD_FORCE_REPLICATION) &&
        listLength(server.slaves))
        replicationFeedSlaves(server.slaves,c->db->id,c->argv,c->argc);
    if (listLength(server.monitors))
        replicationFeedMonitors(server.monitors,c->db->id,c->argv,c->argc);
    server.stat_numcommands++;
}

/* If this function gets called we already read a whole
 * command, argments are in the client argv/argc fields.
 * processCommand() execute the command or prepare the
 * server for a bulk read from the client.
 *
 * If 1 is returned the client is still alive and valid and
 * 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;

P
Pieter Noordhuis 已提交
1024 1025 1026 1027
    /* 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
     * a regular command proc. */
1028
    if (!strcasecmp(c->argv[0]->ptr,"quit")) {
P
Pieter Noordhuis 已提交
1029
        addReply(c,shared.ok);
1030
        c->flags |= REDIS_CLOSE_AFTER_REPLY;
1031
        return REDIS_ERR;
1032 1033 1034 1035 1036 1037
    }

    /* 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) {
1038 1039
        addReplyErrorFormat(c,"unknown command '%s'",
            (char*)c->argv[0]->ptr);
1040
        return REDIS_OK;
1041 1042
    } else if ((cmd->arity > 0 && cmd->arity != c->argc) ||
               (c->argc < -cmd->arity)) {
1043 1044
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
            cmd->name);
1045
        return REDIS_OK;
1046 1047 1048 1049
    }

    /* Check if the user is authenticated */
    if (server.requirepass && !c->authenticated && cmd->proc != authCommand) {
1050
        addReplyError(c,"operation not permitted");
1051
        return REDIS_OK;
1052 1053
    }

1054 1055 1056 1057 1058 1059
    /* 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();
1060
    if (server.maxmemory && (cmd->flags & REDIS_CMD_DENYOOM) &&
1061
        zmalloc_used_memory() > server.maxmemory)
1062
    {
1063
        addReplyError(c,"command not allowed when used memory > 'maxmemory'");
1064
        return REDIS_OK;
1065 1066 1067 1068 1069 1070 1071
    }

    /* 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) {
1072
        addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context");
1073
        return REDIS_OK;
1074 1075
    }

1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086
    /* Only allow INFO and SLAVEOF when slave-serve-stale-data is no and
     * 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)
    {
        addReplyError(c,
            "link with MASTER is down and slave-serve-stale-data is set to no");
        return REDIS_OK;
    }

1087 1088 1089 1090 1091 1092
    /* Loading DB? Return an error if the command is not INFO */
    if (server.loading && cmd->proc != infoCommand) {
        addReply(c, shared.loadingerr);
        return REDIS_OK;
    }

1093 1094 1095 1096 1097 1098 1099 1100
    /* Exec the command */
    if (c->flags & REDIS_MULTI &&
        cmd->proc != execCommand && cmd->proc != discardCommand &&
        cmd->proc != multiCommand && cmd->proc != watchCommand)
    {
        queueMultiCommand(c,cmd);
        addReply(c,shared.queued);
    } else {
1101 1102
        if (server.ds_enabled && blockClientOnSwappedKeys(c,cmd))
            return REDIS_ERR;
1103 1104
        call(c,cmd);
    }
1105
    return REDIS_OK;
1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
}

/*================================== Shutdown =============================== */

int prepareForShutdown() {
    redisLog(REDIS_WARNING,"User requested shutdown, saving DB...");
    /* Kill the saving child if there is a background saving in progress.
       We want to avoid race conditions, for instance our saving child may
       overwrite the synchronous saving did by SHUTDOWN. */
    if (server.bgsavechildpid != -1) {
        redisLog(REDIS_WARNING,"There is a live saving child. Killing it!");
        kill(server.bgsavechildpid,SIGKILL);
        rdbRemoveTempFile(server.bgsavechildpid);
    }
1120 1121 1122
    if (server.ds_enabled) {
        /* FIXME: flush all objects on disk */
    } else if (server.appendonly) {
1123 1124
        /* Append only file: fsync() the AOF and exit */
        aof_fsync(server.appendfd);
1125
    } else if (server.saveparamslen > 0) {
1126
        /* Snapshotting. Perform a SYNC SAVE and exit */
1127
        if (rdbSave(server.dbfilename) != REDIS_OK) {
1128 1129 1130 1131 1132 1133 1134 1135
            /* Ooops.. error saving! The best we can do is to continue
             * operating. Note that if there was a background saving process,
             * in the next cron() Redis will be notified that the background
             * saving aborted, handling special stuff like slaves pending for
             * synchronization... */
            redisLog(REDIS_WARNING,"Error trying to save the DB, can't exit");
            return REDIS_ERR;
        }
1136 1137
    } else {
        redisLog(REDIS_WARNING,"Not saving DB.");
1138
    }
1139
    if (server.daemonize) unlink(server.pidfile);
1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
    redisLog(REDIS_WARNING,"Server exit now, bye bye...");
    return REDIS_OK;
}

/*================================== Commands =============================== */

void authCommand(redisClient *c) {
    if (!server.requirepass || !strcmp(c->argv[1]->ptr, server.requirepass)) {
      c->authenticated = 1;
      addReply(c,shared.ok);
    } else {
      c->authenticated = 0;
1152
      addReplyError(c,"invalid password");
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
    }
}

void pingCommand(redisClient *c) {
    addReply(c,shared.pong);
}

void echoCommand(redisClient *c) {
    addReplyBulk(c,c->argv[1]);
}

/* Convert an amount of bytes into a human readable string in the form
 * of 100B, 2G, 100M, 4K, and so forth. */
void bytesToHuman(char *s, unsigned long long n) {
    double d;

    if (n < 1024) {
        /* Bytes */
        sprintf(s,"%lluB",n);
        return;
    } else if (n < (1024*1024)) {
        d = (double)n/(1024);
        sprintf(s,"%.2fK",d);
    } else if (n < (1024LL*1024*1024)) {
        d = (double)n/(1024*1024);
        sprintf(s,"%.2fM",d);
    } else if (n < (1024LL*1024*1024*1024)) {
        d = (double)n/(1024LL*1024*1024);
        sprintf(s,"%.2fG",d);
    }
}

/* Create the string returned by the INFO command. This is decoupled
 * by the INFO command itself as we need to report the same information
 * on memory corruption problems. */
1188 1189
sds genRedisInfoString(char *section) {
    sds info = sdsempty();
1190
    time_t uptime = time(NULL)-server.stat_starttime;
A
antirez 已提交
1191
    int j, numcommands;
1192
    char hmem[64];
1193
    struct rusage self_ru, c_ru;
1194
    unsigned long lol, bib;
1195 1196 1197 1198 1199
    int allsections = 0, defsections = 0;
    int sections = 0;
    
    if (section) {
        allsections = strcasecmp(section,"all") == 0;
1200
        defsections = strcasecmp(section,"default") == 0;
1201
    }
1202 1203 1204

    getrusage(RUSAGE_SELF, &self_ru);
    getrusage(RUSAGE_CHILDREN, &c_ru);
1205
    getClientsMaxBuffers(&lol,&bib);
1206
    bytesToHuman(hmem,zmalloc_used_memory());
1207 1208 1209 1210

    /* Server */
    if (allsections || defsections || !strcasecmp(section,"server")) {
        if (sections++) info = sdscat(info,"\r\n");
1211
        info = sdscatprintf(info,
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
            "# Server\r\n"
            "redis_version:%s\r\n"
            "redis_git_sha1:%s\r\n"
            "redis_git_dirty:%d\r\n"
            "arch_bits:%s\r\n"
            "multiplexing_api:%s\r\n"
            "process_id:%ld\r\n"
            "tcp_port:%d\r\n"
            "uptime_in_seconds:%ld\r\n"
            "uptime_in_days:%ld\r\n"
            "lru_clock:%ld\r\n",
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
            (sizeof(long) == 8) ? "64" : "32",
            aeGetApiName(),
            (long) getpid(),
            server.port,
            uptime,
            uptime/(3600*24),
            (unsigned long) server.lruclock);
    }

    /* Clients */
    if (allsections || defsections || !strcasecmp(section,"clients")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Clients\r\n"
            "connected_clients:%d\r\n"
            "client_longest_output_list:%lu\r\n"
            "client_biggest_input_buf:%lu\r\n"
            "blocked_clients:%d\r\n",
            listLength(server.clients)-listLength(server.slaves),
            lol, bib,
            server.bpop_blocked_clients);
    }

    /* Memory */
    if (allsections || defsections || !strcasecmp(section,"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"
            "mem_fragmentation_ratio:%.2f\r\n"
            "use_tcmalloc:%d\r\n",
            zmalloc_used_memory(),
            hmem,
            zmalloc_get_rss(),
            zmalloc_get_fragmentation_ratio(),
    #ifdef USE_TCMALLOC
            1
    #else
            0
    #endif
1268
            );
1269 1270 1271 1272 1273 1274
    }

    /* Allocation statistics */
    if (allsections || !strcasecmp(section,"allocstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscat(info, "# Allocstats\r\nallocation_stats:");
1275 1276 1277 1278 1279 1280 1281 1282
        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);
            }
1283
        }
1284
        info = sdscat(info,"\r\n");
1285
    }
1286 1287 1288 1289

    /* Persistence */
    if (allsections || defsections || !strcasecmp(section,"persistence")) {
        if (sections++) info = sdscat(info,"\r\n");
1290
        info = sdscatprintf(info,
1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335
            "# Persistence\r\n"
            "loading:%d\r\n"
            "aof_enabled:%d\r\n"
            "changes_since_last_save:%lld\r\n"
            "bgsave_in_progress:%d\r\n"
            "last_save_time:%ld\r\n"
            "bgrewriteaof_in_progress:%d\r\n",
            server.loading,
            server.appendonly,
            server.dirty,
                server.bgsavechildpid != -1 ||
                server.bgsavethread != (pthread_t) -1,
            server.lastsave,
            server.bgrewritechildpid != -1);

        if (server.loading) {
            double perc;
            time_t eta, elapsed;
            off_t remaining_bytes = server.loading_total_bytes-
                                    server.loading_loaded_bytes;

            perc = ((double)server.loading_loaded_bytes /
                   server.loading_total_bytes) * 100;

            elapsed = time(NULL)-server.loading_start_time;
            if (elapsed == 0) {
                eta = 1; /* A fake 1 second figure if we don't have
                            enough info */
            } else {
                eta = (elapsed*remaining_bytes)/server.loading_loaded_bytes;
            }

            info = sdscatprintf(info,
                "loading_start_time:%ld\r\n"
                "loading_total_bytes:%llu\r\n"
                "loading_loaded_bytes:%llu\r\n"
                "loading_loaded_perc:%.2f\r\n"
                "loading_eta_seconds:%ld\r\n"
                ,(unsigned long) server.loading_start_time,
                (unsigned long long) server.loading_total_bytes,
                (unsigned long long) server.loading_loaded_bytes,
                perc,
                eta
            );
        }
1336
    }
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349

    /* 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"
A
antirez 已提交
1350 1351 1352 1353 1354
                "cache_io_queue_len:%lu\r\n"
                "cache_io_jobs_new:%lu\r\n"
                "cache_io_jobs_processing:%lu\r\n"
                "cache_io_jobs_processed:%lu\r\n"
                "cache_io_ready_clients:%lu\r\n"
1355
                ,(unsigned long long) server.cache_max_memory,
A
antirez 已提交
1356 1357 1358 1359 1360 1361
                (unsigned long) server.cache_blocked_clients,
                (unsigned long) listLength(server.cache_io_queue),
                (unsigned long) listLength(server.io_newjobs),
                (unsigned long) listLength(server.io_processing),
                (unsigned long) listLength(server.io_processed),
                (unsigned long) listLength(server.io_ready_clients)
1362 1363
            );
            unlockThreadedIO();
1364
        }
1365
    }
1366

1367 1368 1369
    /* Stats */
    if (allsections || defsections || !strcasecmp(section,"stats")) {
        if (sections++) info = sdscat(info,"\r\n");
1370
        info = sdscatprintf(info,
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387
            "# Stats\r\n"
            "total_connections_received:%lld\r\n"
            "total_commands_processed:%lld\r\n"
            "expired_keys:%lld\r\n"
            "evicted_keys:%lld\r\n"
            "keyspace_hits:%lld\r\n"
            "keyspace_misses:%lld\r\n"
            "pubsub_channels:%ld\r\n"
            "pubsub_patterns:%u\r\n",
            server.stat_numconnections,
            server.stat_numcommands,
            server.stat_expiredkeys,
            server.stat_evictedkeys,
            server.stat_keyspace_hits,
            server.stat_keyspace_misses,
            dictSize(server.pubsub_channels),
            listLength(server.pubsub_patterns));
1388
    }
A
antirez 已提交
1389

1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
    /* Replication */
    if (allsections || defsections || !strcasecmp(section,"replication")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Replication\r\n"
            "role:%s\r\n",
            server.masterhost == NULL ? "master" : "slave");
        if (server.masterhost) {
            info = sdscatprintf(info,
                "master_host:%s\r\n"
                "master_port:%d\r\n"
                "master_link_status:%s\r\n"
                "master_last_io_seconds_ago:%d\r\n"
                "master_sync_in_progress:%d\r\n"
                ,server.masterhost,
                server.masterport,
                (server.replstate == REDIS_REPL_CONNECTED) ?
                    "up" : "down",
                server.master ?
                ((int)(time(NULL)-server.master->lastinteraction)) : -1,
                server.replstate == REDIS_REPL_TRANSFER
            );

            if (server.replstate == REDIS_REPL_TRANSFER) {
                info = sdscatprintf(info,
                    "master_sync_left_bytes:%ld\r\n"
                    "master_sync_last_io_seconds_ago:%d\r\n"
                    ,(long)server.repl_transfer_left,
                    (int)(time(NULL)-server.repl_transfer_lastio)
                );
            }
A
antirez 已提交
1421
        }
1422 1423 1424
        info = sdscatprintf(info,
            "connected_slaves:%d\r\n",
            listLength(server.slaves));
A
antirez 已提交
1425 1426
    }

1427 1428
    /* CPU */
    if (allsections || defsections || !strcasecmp(section,"cpu")) {
1429 1430
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
1431
        "# CPU\r\n"
1432 1433 1434 1435 1436 1437 1438 1439
        "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",
        (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,
        (float)c_ru.ru_stime.tv_sec+(float)c_ru.ru_stime.tv_usec/1000000);
1440
    }
1441

1442 1443 1444 1445
    /* cmdtime */
    if (allsections || !strcasecmp(section,"commandstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Commandstats\r\n");
1446
        numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1447
        for (j = 0; j < numcommands; j++) {
1448
            struct redisCommand *c = redisCommandTable+j;
1449

1450 1451 1452 1453 1454
            if (!c->calls) continue;
            info = sdscatprintf(info,
                "cmdstat_%s:calls=%lld,usec=%lld,usec_per_call=%.2f\r\n",
                c->name, c->calls, c->microseconds,
                (c->calls == 0) ? 0 : ((float)c->microseconds/c->calls));
1455
        }
A
antirez 已提交
1456 1457
    }

1458 1459 1460 1461 1462 1463
    /* Key space */
    if (allsections || defsections || !strcasecmp(section,"keyspace")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Keyspace\r\n");
        for (j = 0; j < server.dbnum; j++) {
            long long keys, vkeys;
1464

1465 1466 1467 1468 1469 1470
            keys = dictSize(server.db[j].dict);
            vkeys = dictSize(server.db[j].expires);
            if (keys || vkeys) {
                info = sdscatprintf(info, "db%d:keys=%lld,expires=%lld\r\n",
                    j, keys, vkeys);
            }
1471 1472 1473 1474 1475 1476
        }
    }
    return info;
}

void infoCommand(redisClient *c) {
1477 1478 1479 1480 1481 1482 1483
    char *section = c->argc == 2 ? c->argv[1]->ptr : "default";

    if (c->argc > 2) {
        addReply(c,shared.syntaxerr);
        return;
    }
    sds info = genRedisInfoString(section);
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513
    addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
        (unsigned long)sdslen(info)));
    addReplySds(c,info);
    addReply(c,shared.crlf);
}

void monitorCommand(redisClient *c) {
    /* ignore MONITOR if aleady slave or in monitor mode */
    if (c->flags & REDIS_SLAVE) return;

    c->flags |= (REDIS_SLAVE|REDIS_MONITOR);
    c->slaveseldb = 0;
    listAddNodeTail(server.monitors,c);
    addReply(c,shared.ok);
}

/* ============================ Maxmemory directive  ======================== */

/* This function gets called when 'maxmemory' is set on the config file to limit
 * the max memory used by the server, and we are out of memory.
 * This function will try to, in order:
 *
 * - Free objects from the free list
 * - Try to remove keys with an EXPIRE set
 *
 * It is not possible to free enough memory to reach used-memory < maxmemory
 * the server will start refusing commands that will enlarge even more the
 * memory usage.
 */
void freeMemoryIfNeeded(void) {
1514 1515
    /* Remove keys accordingly to the active policy as long as we are
     * over the memory limit. */
1516 1517
    if (server.maxmemory_policy == REDIS_MAXMEMORY_NO_EVICTION) return;

1518
    while (server.maxmemory && zmalloc_used_memory() > server.maxmemory) {
1519 1520
        int j, k, freed = 0;

1521
        for (j = 0; j < server.dbnum; j++) {
1522
            long bestval = 0; /* just to prevent warning */
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555
            sds bestkey = NULL;
            struct dictEntry *de;
            redisDb *db = server.db+j;
            dict *dict;

            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM)
            {
                dict = server.db[j].dict;
            } else {
                dict = server.db[j].expires;
            }
            if (dictSize(dict) == 0) continue;

            /* volatile-random and allkeys-random policy */
            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM ||
                server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_RANDOM)
            {
                de = dictGetRandomKey(dict);
                bestkey = dictGetEntryKey(de);
            }

            /* volatile-lru and allkeys-lru policy */
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
            {
                for (k = 0; k < server.maxmemory_samples; k++) {
                    sds thiskey;
                    long thisval;
                    robj *o;

                    de = dictGetRandomKey(dict);
                    thiskey = dictGetEntryKey(de);
1556 1557 1558 1559
                    /* When policy is volatile-lru we need an additonal lookup
                     * to locate the real key, as dict is set to db->expires. */
                    if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
                        de = dictFind(db->dict, thiskey);
1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592
                    o = dictGetEntryVal(de);
                    thisval = estimateObjectIdleTime(o);

                    /* Higher idle time is better candidate for deletion */
                    if (bestkey == NULL || thisval > bestval) {
                        bestkey = thiskey;
                        bestval = thisval;
                    }
                }
            }

            /* volatile-ttl */
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_TTL) {
                for (k = 0; k < server.maxmemory_samples; k++) {
                    sds thiskey;
                    long thisval;

                    de = dictGetRandomKey(dict);
                    thiskey = dictGetEntryKey(de);
                    thisval = (long) dictGetEntryVal(de);

                    /* Expire sooner (minor expire unix timestamp) is better
                     * candidate for deletion */
                    if (bestkey == NULL || thisval < bestval) {
                        bestkey = thiskey;
                        bestval = thisval;
                    }
                }
            }

            /* Finally remove the selected key. */
            if (bestkey) {
                robj *keyobj = createStringObject(bestkey,sdslen(bestkey));
1593
                propagateExpire(db,keyobj);
1594
                dbDelete(db,keyobj);
1595
                server.stat_evictedkeys++;
1596 1597 1598 1599 1600 1601
                decrRefCount(keyobj);
                freed++;
            }
        }
        if (!freed) return; /* nothing to free... */
    }
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627
}

/* =================================== Main! ================================ */

#ifdef __linux__
int linuxOvercommitMemoryValue(void) {
    FILE *fp = fopen("/proc/sys/vm/overcommit_memory","r");
    char buf[64];

    if (!fp) return -1;
    if (fgets(buf,64,fp) == NULL) {
        fclose(fp);
        return -1;
    }
    fclose(fp);

    return atoi(buf);
}

void linuxOvercommitMemoryWarning(void) {
    if (linuxOvercommitMemoryValue() == 0) {
        redisLog(REDIS_WARNING,"WARNING overcommit_memory is set to 0! Background save may fail under low memory condition. To fix this issue add 'vm.overcommit_memory = 1' to /etc/sysctl.conf and then reboot or run the command 'sysctl vm.overcommit_memory=1' for this to take effect.");
    }
}
#endif /* __linux__ */

1628 1629 1630 1631
void createPidFile(void) {
    /* Try to write the pid file in a best-effort way. */
    FILE *fp = fopen(server.pidfile,"w");
    if (fp) {
1632
        fprintf(fp,"%d\n",(int)getpid());
1633 1634 1635 1636
        fclose(fp);
    }
}

1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666
void daemonize(void) {
    int fd;

    if (fork() != 0) exit(0); /* parent exits */
    setsid(); /* create a new session */

    /* Every output goes to /dev/null. If Redis is daemonized but
     * the 'logfile' is set to 'stdout' in the configuration file
     * it will not log at all. */
    if ((fd = open("/dev/null", O_RDWR, 0)) != -1) {
        dup2(fd, STDIN_FILENO);
        dup2(fd, STDOUT_FILENO);
        dup2(fd, STDERR_FILENO);
        if (fd > STDERR_FILENO) close(fd);
    }
}

void version() {
    printf("Redis server version %s (%s:%d)\n", REDIS_VERSION,
        redisGitSHA1(), atoi(redisGitDirty()) > 0);
    exit(0);
}

void usage() {
    fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf]\n");
    fprintf(stderr,"       ./redis-server - (read config from stdin)\n");
    exit(1);
}

int main(int argc, char **argv) {
1667
    long long start;
1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682

    initServerConfig();
    if (argc == 2) {
        if (strcmp(argv[1], "-v") == 0 ||
            strcmp(argv[1], "--version") == 0) version();
        if (strcmp(argv[1], "--help") == 0) usage();
        resetServerSaveParams();
        loadServerConfig(argv[1]);
    } else if ((argc > 2)) {
        usage();
    } else {
        redisLog(REDIS_WARNING,"Warning: no config file specified, using the default config. In order to specify a config file use 'redis-server /path/to/redis.conf'");
    }
    if (server.daemonize) daemonize();
    initServer();
1683
    if (server.daemonize) createPidFile();
1684 1685 1686 1687
    redisLog(REDIS_NOTICE,"Server started, Redis version " REDIS_VERSION);
#ifdef __linux__
    linuxOvercommitMemoryWarning();
#endif
1688
    start = ustime();
A
antirez 已提交
1689
    if (server.ds_enabled) {
A
antirez 已提交
1690
        redisLog(REDIS_NOTICE,"DB not loaded (running with disk back end)");
A
antirez 已提交
1691
    } else if (server.appendonly) {
1692
        if (loadAppendOnlyFile(server.appendfilename) == REDIS_OK)
1693
            redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
1694 1695
    } else {
        if (rdbLoad(server.dbfilename) == REDIS_OK)
1696
            redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",(float)(ustime()-start)/1000000);
1697
    }
1698 1699 1700
    if (server.ipfd > 0)
        redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
    if (server.sofd > 0)
1701
        redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727
    aeSetBeforeSleepProc(server.el,beforeSleep);
    aeMain(server.el);
    aeDeleteEventLoop(server.el);
    return 0;
}

/* ============================= Backtrace support ========================= */

#ifdef HAVE_BACKTRACE
void *getMcontextEip(ucontext_t *uc) {
#if defined(__FreeBSD__)
    return (void*) uc->uc_mcontext.mc_eip;
#elif defined(__dietlibc__)
    return (void*) uc->uc_mcontext.eip;
#elif defined(__APPLE__) && !defined(MAC_OS_X_VERSION_10_6)
  #if __x86_64__
    return (void*) uc->uc_mcontext->__ss.__rip;
  #else
    return (void*) uc->uc_mcontext->__ss.__eip;
  #endif
#elif defined(__APPLE__) && defined(MAC_OS_X_VERSION_10_6)
  #if defined(_STRUCT_X86_THREAD_STATE64) && !defined(__i386__)
    return (void*) uc->uc_mcontext->__ss.__rip;
  #else
    return (void*) uc->uc_mcontext->__ss.__eip;
  #endif
1728 1729 1730 1731
#elif defined(__i386__)
    return (void*) uc->uc_mcontext.gregs[14]; /* Linux 32 */
#elif defined(__X86_64__) || defined(__x86_64__)
    return (void*) uc->uc_mcontext.gregs[16]; /* Linux 64 */
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744
#elif defined(__ia64__) /* Linux IA64 */
    return (void*) uc->uc_mcontext.sc_ip;
#else
    return NULL;
#endif
}

void segvHandler(int sig, siginfo_t *info, void *secret) {
    void *trace[100];
    char **messages = NULL;
    int i, trace_size = 0;
    ucontext_t *uc = (ucontext_t*) secret;
    sds infostring;
1745
    struct sigaction act;
1746 1747 1748 1749
    REDIS_NOTUSED(info);

    redisLog(REDIS_WARNING,
        "======= Ooops! Redis %s got signal: -%d- =======", REDIS_VERSION, sig);
1750
    infostring = genRedisInfoString("all");
1751
    redisLogRaw(REDIS_WARNING, infostring);
1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765
    /* It's not safe to sdsfree() the returned string under memory
     * corruption conditions. Let it leak as we are going to abort */

    trace_size = backtrace(trace, 100);
    /* overwrite sigaction with caller's address */
    if (getMcontextEip(uc) != NULL) {
        trace[1] = getMcontextEip(uc);
    }
    messages = backtrace_symbols(trace, trace_size);

    for (i=1; i<trace_size; ++i)
        redisLog(REDIS_WARNING,"%s", messages[i]);

    /* free(messages); Don't call free() with possibly corrupted memory. */
1766
    if (server.daemonize) unlink(server.pidfile);
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776

    /* Make sure we exit with the right signal at the end. So for instance
     * the core will be dumped if enabled. */
    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;
    act.sa_handler = SIG_DFL;
    sigaction (sig, &act, NULL);
    kill(getpid(),sig);
1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811
}

void sigtermHandler(int sig) {
    REDIS_NOTUSED(sig);

    redisLog(REDIS_WARNING,"SIGTERM received, scheduling shutting down...");
    server.shutdown_asap = 1;
}

void setupSigSegvAction(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);

    act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND;
    act.sa_handler = sigtermHandler;
    sigaction (SIGTERM, &act, NULL);
    return;
}

#else /* HAVE_BACKTRACE */
void setupSigSegvAction(void) {
}
#endif /* HAVE_BACKTRACE */

/* The End */