redis.c 101.8 KB
Newer Older
1
/*
2
 * Copyright (c) 2009-2012, Salvatore Sanfilippo <antirez at gmail dot com>
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
 * 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"
31
#include "slowlog.h"
32
#include "bio.h"
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

#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>
50
#include <sys/resource.h>
51
#include <sys/utsname.h>
52 53 54 55 56

/* Our shared "common" objects */

struct sharedObjectsStruct shared;

57
/* Global vars that are actually used as constants. The following double
58 59 60 61 62 63 64 65 66 67
 * 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;
68

69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
/* Our command table.
 *
 * Every entry is composed of the following fields:
 *
 * name: a string representing the command name.
 * function: pointer to the C function implementing the command.
 * arity: number of arguments, it is possible to use -N to say >= N
 * sflags: command flags as string. See below for a table of flags.
 * flags: flags as bitmask. Computed by Redis using the 'sflags' field.
 * get_keys_proc: an optional function to get key arguments from a command.
 *                This is only used when the following three fields are not
 *                enough to specify what arguments are keys.
 * first_key_index: first argument that is a key
 * last_key_index: last argument that is a key
 * key_step: step to get all the keys from first to last argument. For instance
 *           in MSET the step is two since arguments are key,val,key,val,...
 * microseconds: microseconds of total execution time for this command.
 * calls: total number of calls of this command.
 *
 * The flags, microseconds and calls fields are computed by Redis and should
 * always be set to zero.
 *
 * Command flags are expressed using strings where every character represents
 * a flag. Later the populateCommandTable() function will take care of
 * populating the real 'flags' field using this characters.
94 95 96 97 98 99 100 101
 *
 * This is the meaning of the flags:
 *
 * w: write command (may modify the key space).
 * r: read command  (will never modify the key space).
 * m: may increase memory usage once called. Don't allow if out of memory.
 * a: admin command, like SAVE or SHUTDOWN.
 * p: Pub/Sub related command.
G
guiquanz 已提交
102
 * f: force replication of this command, regardless of server.dirty.
103
 * s: command not allowed in scripts.
104
 * R: random command. Command is not deterministic, that is, the same command
105
 *    with the same arguments, with the same key space, may have different
106 107 108
 *    results. For instance SPOP and RANDOMKEY are two random commands.
 * S: Sort command output array if called from script, so that the output
 *    is deterministic.
109 110 111 112
 * l: Allow command while loading the database.
 * t: Allow command while a slave has stale data but is not allowed to
 *    server this data. Normally no command is accepted in this condition
 *    but just a few.
113
 * M: Do not automatically propagate the command on MONITOR.
114
 */
115
struct redisCommand redisCommandTable[] = {
116 117 118
    {"get",getCommand,2,"r",0,NULL,1,1,1,0,0},
    {"set",setCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
    {"setnx",setnxCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
119 120
    {"setex",setexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
    {"psetex",psetexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
    {"append",appendCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"strlen",strlenCommand,2,"r",0,NULL,1,1,1,0,0},
    {"del",delCommand,-2,"w",0,noPreloadGetKeys,1,-1,1,0,0},
    {"exists",existsCommand,2,"r",0,NULL,1,1,1,0,0},
    {"setbit",setbitCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"getbit",getbitCommand,3,"r",0,NULL,1,1,1,0,0},
    {"setrange",setrangeCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"getrange",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"substr",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"incr",incrCommand,2,"wm",0,NULL,1,1,1,0,0},
    {"decr",decrCommand,2,"wm",0,NULL,1,1,1,0,0},
    {"mget",mgetCommand,-2,"r",0,NULL,1,-1,1,0,0},
    {"rpush",rpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"lpush",lpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"rpushx",rpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"lpushx",lpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"linsert",linsertCommand,5,"wm",0,NULL,1,1,1,0,0},
    {"rpop",rpopCommand,2,"w",0,NULL,1,1,1,0,0},
    {"lpop",lpopCommand,2,"w",0,NULL,1,1,1,0,0},
A
antirez 已提交
140 141 142
    {"brpop",brpopCommand,-3,"ws",0,NULL,1,1,1,0,0},
    {"brpoplpush",brpoplpushCommand,4,"wms",0,NULL,1,2,1,0,0},
    {"blpop",blpopCommand,-3,"ws",0,NULL,1,-2,1,0,0},
143 144 145 146 147 148 149 150 151 152 153 154
    {"llen",llenCommand,2,"r",0,NULL,1,1,1,0,0},
    {"lindex",lindexCommand,3,"r",0,NULL,1,1,1,0,0},
    {"lset",lsetCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"lrange",lrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"ltrim",ltrimCommand,4,"w",0,NULL,1,1,1,0,0},
    {"lrem",lremCommand,4,"w",0,NULL,1,1,1,0,0},
    {"rpoplpush",rpoplpushCommand,3,"wm",0,NULL,1,2,1,0,0},
    {"sadd",saddCommand,-3,"wm",0,NULL,1,1,1,0,0},
    {"srem",sremCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"smove",smoveCommand,4,"w",0,NULL,1,2,1,0,0},
    {"sismember",sismemberCommand,3,"r",0,NULL,1,1,1,0,0},
    {"scard",scardCommand,2,"r",0,NULL,1,1,1,0,0},
155
    {"spop",spopCommand,2,"wRs",0,NULL,1,1,1,0,0},
156
    {"srandmember",srandmemberCommand,-2,"rR",0,NULL,1,1,1,0,0},
157
    {"sinter",sinterCommand,-2,"rS",0,NULL,1,-1,1,0,0},
158
    {"sinterstore",sinterstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
159
    {"sunion",sunionCommand,-2,"rS",0,NULL,1,-1,1,0,0},
160
    {"sunionstore",sunionstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
161
    {"sdiff",sdiffCommand,-2,"rS",0,NULL,1,-1,1,0,0},
162
    {"sdiffstore",sdiffstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
163
    {"smembers",sinterCommand,2,"rS",0,NULL,1,1,1,0,0},
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
    {"zadd",zaddCommand,-4,"wm",0,NULL,1,1,1,0,0},
    {"zincrby",zincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"zrem",zremCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"zremrangebyscore",zremrangebyscoreCommand,4,"w",0,NULL,1,1,1,0,0},
    {"zremrangebyrank",zremrangebyrankCommand,4,"w",0,NULL,1,1,1,0,0},
    {"zunionstore",zunionstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zinterstore",zinterstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zrange",zrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrangebyscore",zrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrevrangebyscore",zrevrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zcount",zcountCommand,4,"r",0,NULL,1,1,1,0,0},
    {"zrevrange",zrevrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zcard",zcardCommand,2,"r",0,NULL,1,1,1,0,0},
    {"zscore",zscoreCommand,3,"r",0,NULL,1,1,1,0,0},
    {"zrank",zrankCommand,3,"r",0,NULL,1,1,1,0,0},
    {"zrevrank",zrevrankCommand,3,"r",0,NULL,1,1,1,0,0},
    {"hset",hsetCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"hsetnx",hsetnxCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"hget",hgetCommand,3,"r",0,NULL,1,1,1,0,0},
    {"hmset",hmsetCommand,-4,"wm",0,NULL,1,1,1,0,0},
    {"hmget",hmgetCommand,-3,"r",0,NULL,1,1,1,0,0},
    {"hincrby",hincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
A
antirez 已提交
186
    {"hincrbyfloat",hincrbyfloatCommand,4,"wm",0,NULL,1,1,1,0,0},
187 188
    {"hdel",hdelCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"hlen",hlenCommand,2,"r",0,NULL,1,1,1,0,0},
189 190
    {"hkeys",hkeysCommand,2,"rS",0,NULL,1,1,1,0,0},
    {"hvals",hvalsCommand,2,"rS",0,NULL,1,1,1,0,0},
191 192 193 194
    {"hgetall",hgetallCommand,2,"r",0,NULL,1,1,1,0,0},
    {"hexists",hexistsCommand,3,"r",0,NULL,1,1,1,0,0},
    {"incrby",incrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"decrby",decrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
A
antirez 已提交
195
    {"incrbyfloat",incrbyfloatCommand,3,"wm",0,NULL,1,1,1,0,0},
196 197 198
    {"getset",getsetCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"mset",msetCommand,-3,"wm",0,NULL,1,-1,2,0,0},
    {"msetnx",msetnxCommand,-3,"wm",0,NULL,1,-1,2,0,0},
199
    {"randomkey",randomkeyCommand,1,"rR",0,NULL,0,0,0,0,0},
200 201 202 203
    {"select",selectCommand,2,"r",0,NULL,0,0,0,0,0},
    {"move",moveCommand,3,"w",0,NULL,1,1,1,0,0},
    {"rename",renameCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
    {"renamenx",renamenxCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
204 205 206 207
    {"expire",expireCommand,3,"w",0,NULL,1,1,1,0,0},
    {"expireat",expireatCommand,3,"w",0,NULL,1,1,1,0,0},
    {"pexpire",pexpireCommand,3,"w",0,NULL,1,1,1,0,0},
    {"pexpireat",pexpireatCommand,3,"w",0,NULL,1,1,1,0,0},
208
    {"keys",keysCommand,2,"rS",0,NULL,0,0,0,0,0},
209
    {"dbsize",dbsizeCommand,1,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
210
    {"auth",authCommand,2,"rs",0,NULL,0,0,0,0,0},
211 212
    {"ping",pingCommand,1,"r",0,NULL,0,0,0,0,0},
    {"echo",echoCommand,2,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
213
    {"save",saveCommand,1,"ars",0,NULL,0,0,0,0,0},
214 215
    {"bgsave",bgsaveCommand,1,"ar",0,NULL,0,0,0,0,0},
    {"bgrewriteaof",bgrewriteaofCommand,1,"ar",0,NULL,0,0,0,0,0},
216
    {"shutdown",shutdownCommand,-1,"ar",0,NULL,0,0,0,0,0},
A
antirez 已提交
217
    {"lastsave",lastsaveCommand,1,"rR",0,NULL,0,0,0,0,0},
218
    {"type",typeCommand,2,"r",0,NULL,1,1,1,0,0},
219
    {"multi",multiCommand,1,"rs",0,NULL,0,0,0,0,0},
220
    {"exec",execCommand,1,"sM",0,NULL,0,0,0,0,0},
221 222
    {"discard",discardCommand,1,"rs",0,NULL,0,0,0,0,0},
    {"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
223
    {"psync",syncCommand,3,"ars",0,NULL,0,0,0,0,0},
A
antirez 已提交
224
    {"replconf",replconfCommand,-1,"ars",0,NULL,0,0,0,0,0},
225 226
    {"flushdb",flushdbCommand,1,"w",0,NULL,0,0,0,0,0},
    {"flushall",flushallCommand,1,"w",0,NULL,0,0,0,0,0},
227
    {"sort",sortCommand,-2,"wm",0,NULL,1,1,1,0,0},
228
    {"info",infoCommand,-1,"rlt",0,NULL,0,0,0,0,0},
229
    {"monitor",monitorCommand,1,"ars",0,NULL,0,0,0,0,0},
230 231
    {"ttl",ttlCommand,2,"r",0,NULL,1,1,1,0,0},
    {"pttl",pttlCommand,2,"r",0,NULL,1,1,1,0,0},
232
    {"persist",persistCommand,2,"w",0,NULL,1,1,1,0,0},
233
    {"slaveof",slaveofCommand,3,"ast",0,NULL,0,0,0,0,0},
234
    {"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
235
    {"config",configCommand,-2,"ar",0,NULL,0,0,0,0,0},
236 237 238 239 240
    {"subscribe",subscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
    {"unsubscribe",unsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
    {"psubscribe",psubscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
    {"punsubscribe",punsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
    {"publish",publishCommand,3,"pflt",0,NULL,0,0,0,0,0},
241 242
    {"watch",watchCommand,-2,"rs",0,noPreloadGetKeys,1,-1,1,0,0},
    {"unwatch",unwatchCommand,1,"rs",0,NULL,0,0,0,0,0},
243
    {"restore",restoreCommand,4,"awm",0,NULL,1,1,1,0,0},
244
    {"migrate",migrateCommand,6,"aw",0,NULL,0,0,0,0,0},
245 246
    {"dump",dumpCommand,2,"ar",0,NULL,1,1,1,0,0},
    {"object",objectCommand,-2,"r",0,NULL,2,2,2,0,0},
247
    {"client",clientCommand,-2,"ar",0,NULL,0,0,0,0,0},
248 249
    {"eval",evalCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
    {"evalsha",evalShaCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
250
    {"slowlog",slowlogCommand,-2,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
251
    {"script",scriptCommand,-2,"ras",0,NULL,0,0,0,0,0},
A
antirez 已提交
252 253 254
    {"time",timeCommand,1,"rR",0,NULL,0,0,0,0,0},
    {"bitop",bitopCommand,-4,"wm",0,NULL,2,-1,1,0,0},
    {"bitcount",bitcountCommand,-2,"r",0,NULL,1,1,1,0,0}
255 256 257 258
};

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

259 260 261
/* 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
262 263
    const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
    const char *c = ".-*#";
264
    FILE *fp;
265
    char buf[64];
A
antirez 已提交
266
    int rawmode = (level & REDIS_LOG_RAW);
267

A
antirez 已提交
268
    level &= 0xff; /* clear flags */
269
    if (level < server.verbosity) return;
270 271 272 273

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

A
antirez 已提交
274 275 276
    if (rawmode) {
        fprintf(fp,"%s",msg);
    } else {
277 278 279 280 281 282
        int off;
        struct timeval tv;

        gettimeofday(&tv,NULL);
        off = strftime(buf,sizeof(buf),"%d %b %H:%M:%S.",localtime(&tv.tv_sec));
        snprintf(buf+off,sizeof(buf)-off,"%03d",(int)tv.tv_usec/1000);
A
antirez 已提交
283 284
        fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg);
    }
J
Jonah H. Harris 已提交
285 286
    fflush(fp);

287
    if (server.logfile) fclose(fp);
J
Jonah H. Harris 已提交
288 289

    if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
290 291
}

G
guiquanz 已提交
292
/* Like redisLogRaw() but with printf-alike support. This is the function that
293 294 295 296 297 298
 * 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];

A
antirez 已提交
299
    if ((level&0xff) < server.verbosity) return;
300 301 302 303 304 305 306 307

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

    redisLogRaw(level,msg);
}

A
antirez 已提交
308 309 310 311 312 313 314 315 316 317 318 319 320 321
/* Log a fixed message without printf-alike capabilities, in a way that is
 * safe to call from a signal handler.
 *
 * We actually use this only for signals that are not fatal from the point
 * of view of Redis. Signals that are going to kill the server anyway and
 * where we need printf-alike features are served by redisLog(). */
void redisLogFromHandler(int level, const char *msg) {
    int fd;
    char buf[64];

    if ((level&0xff) < server.verbosity ||
        (server.logfile == NULL && server.daemonize)) return;
    fd = server.logfile ?
        open(server.logfile, O_APPEND|O_CREAT|O_WRONLY, 0644) :
A
antirez 已提交
322
        STDOUT_FILENO;
A
antirez 已提交
323 324
    if (fd == -1) return;
    ll2string(buf,sizeof(buf),getpid());
325 326 327
    if (write(fd,"[",1) == -1) goto err;
    if (write(fd,buf,strlen(buf)) == -1) goto err;
    if (write(fd," | signal handler] (",20) == -1) goto err;
A
antirez 已提交
328
    ll2string(buf,sizeof(buf),time(NULL));
329 330 331 332 333
    if (write(fd,buf,strlen(buf)) == -1) goto err;
    if (write(fd,") ",2) == -1) goto err;
    if (write(fd,msg,strlen(msg)) == -1) goto err;
    if (write(fd,"\n",1) == -1) goto err;
err:
A
antirez 已提交
334
    if (server.logfile) close(fd);
A
antirez 已提交
335 336
}

A
antirez 已提交
337 338 339 340 341 342 343 344 345 346 347
/* 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;
}

348 349 350 351 352
/* Return the UNIX time in milliseconds */
long long mstime(void) {
    return ustime()/1000;
}

353 354 355 356 357 358 359 360 361 362 363 364
/* After an RDB dump or AOF rewrite we exit from children using _exit() instead of
 * exit(), because the latter may interact with the same file objects used by
 * the parent process. However if we are testing the coverage normal exit() is
 * used in order to obtain the right coverage information. */
void exitFromChild(int retcode) {
#ifdef COVERAGE_TEST
    exit(retcode);
#else
    _exit(retcode);
#endif
}

365 366
/*====================== Hash table type implementation  ==================== */

G
guiquanz 已提交
367
/* This is an hash table type that uses the SDS dynamic strings library as
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
 * 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;
}

A
antirez 已提交
395 396
/* A case insensitive version used for the command lookup table and other
 * places where case insensitive non binary-safe comparison is needed. */
397 398 399 400 401 402 403 404
int dictSdsKeyCaseCompare(void *privdata, const void *key1,
        const void *key2)
{
    DICT_NOTUSED(privdata);

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

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
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));
}

436 437 438 439
unsigned int dictSdsCaseHash(const void *key) {
    return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
}

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
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;
        }
    }
}

481
/* Sets type hash table */
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497
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 */
498
    NULL                       /* val destructor */
499 500 501 502 503 504 505 506 507 508
};

/* 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 */
A
antirez 已提交
509 510 511 512 513 514 515 516 517 518
};

/* server.lua_scripts sha (as sds string) -> scripts (as robj) cache. */
dictType shaScriptObjectDictType = {
    dictSdsCaseHash,            /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCaseCompare,      /* key compare */
    dictSdsDestructor,          /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
519 520 521 522 523 524 525 526 527 528 529 530
};

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

531 532 533 534 535 536 537 538 539 540
/* 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 */
};

541
/* Hash type hash table (note that small hashes are represented with ziplists) */
542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592
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++) {
593
        /* Keys dictionary */
594 595 596 597
        if (dictIsRehashing(server.db[j].dict)) {
            dictRehashMilliseconds(server.db[j].dict,1);
            break; /* already used our millisecond for this loop... */
        }
598 599 600 601 602
        /* Expires */
        if (dictIsRehashing(server.db[j].expires)) {
            dictRehashMilliseconds(server.db[j].expires,1);
            break; /* already used our millisecond for this loop... */
        }
603 604 605 606 607 608 609 610 611 612
    }
}

/* 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) {
A
antirez 已提交
613
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
614 615 616 617 618 619 620
        dictEnableResize();
    else
        dictDisableResize();
}

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

621 622 623 624 625
/* 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) {
626
    int j, iteration = 0;
627
    long long start = ustime(), timelimit;
628 629 630

    /* We can use at max REDIS_EXPIRELOOKUPS_TIME_PERC percentage of CPU time
     * per iteration. Since this function gets called with a frequency of
631
     * server.hz times per second, the following is the max amount of
632
     * microseconds we can spend in this function. */
633
    timelimit = 1000000*REDIS_EXPIRELOOKUPS_TIME_PERC/server.hz/100;
634
    if (timelimit <= 0) timelimit = 1;
635 636

    for (j = 0; j < server.dbnum; j++) {
637
        int expired;
638 639 640 641 642
        redisDb *db = server.db+j;

        /* Continue to expire if at the end of the cycle more than 25%
         * of the keys were expired. */
        do {
643 644
            unsigned long num = dictSize(db->expires);
            unsigned long slots = dictSlots(db->expires);
645
            long long now = mstime();
646

647 648 649 650 651 652
            /* When there are less than 1% filled slots getting random
             * keys is expensive, so stop here waiting for better times...
             * The dictionary will be resized asap. */
            if (num && slots > DICT_HT_INITIAL_SIZE &&
                (num*100/slots < 1)) break;

653 654
            /* The main collection cycle. Sample random keys among keys
             * with an expire set, checking for expired ones. */
655 656 657 658 659
            expired = 0;
            if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
                num = REDIS_EXPIRELOOKUPS_PER_CRON;
            while (num--) {
                dictEntry *de;
660
                long long t;
661 662

                if ((de = dictGetRandomKey(db->expires)) == NULL) break;
663
                t = dictGetSignedIntegerVal(de);
664
                if (now > t) {
665
                    sds key = dictGetKey(de);
666 667 668 669
                    robj *keyobj = createStringObject(key,sdslen(key));

                    propagateExpire(db,keyobj);
                    dbDelete(db,keyobj);
670 671
                    notifyKeyspaceEvent(REDIS_NOTIFY_EXPIRED,
                        "expired",keyobj,db->id);
672 673 674 675 676
                    decrRefCount(keyobj);
                    expired++;
                    server.stat_expiredkeys++;
                }
            }
677 678 679 680
            /* We can't block forever here even if there are many keys to
             * expire. So after a given amount of milliseconds return to the
             * caller waiting for the other active expire cycle. */
            iteration++;
681
            if ((iteration & 0xf) == 0 && /* check once every 16 cycles. */
682
                (ustime()-start) > timelimit) return;
683 684 685 686
        } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
    }
}

687
void updateLRUClock(void) {
688
    server.lruclock = (server.unixtime/REDIS_LRU_CLOCK_RESOLUTION) &
689 690
                                                REDIS_LRU_CLOCK_MAX;
}
691

692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716

/* Add a sample to the operations per second array of samples. */
void trackOperationsPerSecond(void) {
    long long t = mstime() - server.ops_sec_last_sample_time;
    long long ops = server.stat_numcommands - server.ops_sec_last_sample_ops;
    long long ops_sec;

    ops_sec = t > 0 ? (ops*1000/t) : 0;

    server.ops_sec_samples[server.ops_sec_idx] = ops_sec;
    server.ops_sec_idx = (server.ops_sec_idx+1) % REDIS_OPS_SEC_SAMPLES;
    server.ops_sec_last_sample_time = mstime();
    server.ops_sec_last_sample_ops = server.stat_numcommands;
}

/* Return the mean of all the samples. */
long long getOperationsPerSecond(void) {
    int j;
    long long sum = 0;

    for (j = 0; j < REDIS_OPS_SEC_SAMPLES; j++)
        sum += server.ops_sec_samples[j];
    return sum / REDIS_OPS_SEC_SAMPLES;
}

A
antirez 已提交
717 718
/* Check for timeouts. Returns non-zero if the client was terminated */
int clientsCronHandleTimeout(redisClient *c) {
719
    time_t now = server.unixtime;
720 721 722 723 724 725 726 727 728 729 730

    if (server.maxidletime &&
        !(c->flags & REDIS_SLAVE) &&    /* no timeout for slaves */
        !(c->flags & REDIS_MASTER) &&   /* no timeout for masters */
        !(c->flags & REDIS_BLOCKED) &&  /* no timeout for BLPOP */
        dictSize(c->pubsub_channels) == 0 && /* no timeout for pubsub */
        listLength(c->pubsub_patterns) == 0 &&
        (now - c->lastinteraction > server.maxidletime))
    {
        redisLog(REDIS_VERBOSE,"Closing idle client");
        freeClient(c);
A
antirez 已提交
731
        return 1;
732 733 734 735 736 737
    } else if (c->flags & REDIS_BLOCKED) {
        if (c->bpop.timeout != 0 && c->bpop.timeout < now) {
            addReply(c,shared.nullmultibulk);
            unblockClientWaitingData(c);
        }
    }
A
antirez 已提交
738
    return 0;
739 740
}

741
/* The client query buffer is an sds.c string that can end with a lot of
A
antirez 已提交
742 743
 * free space not used, this function reclaims space if needed.
 *
G
guiquanz 已提交
744
 * The function always returns 0 as it never terminates the client. */
A
antirez 已提交
745
int clientsCronResizeQueryBuffer(redisClient *c) {
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763
    size_t querybuf_size = sdsAllocSize(c->querybuf);
    time_t idletime = server.unixtime - c->lastinteraction;

    /* There are two conditions to resize the query buffer:
     * 1) Query buffer is > BIG_ARG and too big for latest peak.
     * 2) Client is inactive and the buffer is bigger than 1k. */
    if (((querybuf_size > REDIS_MBULK_BIG_ARG) &&
         (querybuf_size/(c->querybuf_peak+1)) > 2) ||
         (querybuf_size > 1024 && idletime > 2))
    {
        /* Only resize the query buffer if it is actually wasting space. */
        if (sdsavail(c->querybuf) > 1024) {
            c->querybuf = sdsRemoveFreeSpace(c->querybuf);
        }
    }
    /* Reset the peak again to capture the peak memory usage in the next
     * cycle. */
    c->querybuf_peak = 0;
A
antirez 已提交
764
    return 0;
765 766
}

767
void clientsCron(void) {
768 769
    /* Make sure to process at least 1/(server.hz*10) of clients per call.
     * Since this function is called server.hz times per second we are sure that
770 771 772
     * in the worst case we process all the clients in 10 seconds.
     * In normal conditions (a reasonable number of clients) we process
     * all the clients in a shorter time. */
773
    int numclients = listLength(server.clients);
774
    int iterations = numclients/(server.hz*10);
775

776 777
    if (iterations < 50)
        iterations = (numclients < 50) ? numclients : 50;
778 779 780 781 782 783 784 785 786 787
    while(listLength(server.clients) && iterations--) {
        redisClient *c;
        listNode *head;

        /* Rotate the list, take the current head, process.
         * This way if the client must be removed from the list it's the
         * first element and we don't incur into O(N) computation. */
        listRotate(server.clients);
        head = listFirst(server.clients);
        c = listNodeValue(head);
A
antirez 已提交
788 789 790 791 792
        /* The following functions do different service checks on the client.
         * The protocol is that they return non-zero if the client was
         * terminated. */
        if (clientsCronHandleTimeout(c)) continue;
        if (clientsCronResizeQueryBuffer(c)) continue;
793 794 795
    }
}

796
/* This is our timer interrupt, called server.hz times per second.
797 798 799 800 801
 * Here is where we do a number of things that need to be done asynchronously.
 * For instance:
 *
 * - Active expired keys collection (it is also performed in a lazy way on
 *   lookup).
G
guiquanz 已提交
802
 * - Software watchdog.
803 804 805
 * - Update some statistic.
 * - Incremental rehashing of the DBs hash tables.
 * - Triggering BGSAVE / AOF rewrite, and handling of terminated children.
G
guiquanz 已提交
806
 * - Clients timeout of different kinds.
807 808 809
 * - Replication reconnection.
 * - Many more...
 *
810
 * Everything directly called here will be called server.hz times per second,
811 812 813 814
 * so in order to throttle execution of things we want to do less frequently
 * a macro is used: run_with_period(milliseconds) { .... }
 */

815
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
816
    int j;
817 818 819 820
    REDIS_NOTUSED(eventLoop);
    REDIS_NOTUSED(id);
    REDIS_NOTUSED(clientData);

A
antirez 已提交
821 822 823 824
    /* Software watchdog: deliver the SIGALRM that will reach the signal
     * handler if we don't return here fast enough. */
    if (server.watchdog_period) watchdogScheduleSignal(server.watchdog_period);

825 826 827 828 829
    /* 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);
A
antirez 已提交
830

831
    run_with_period(100) trackOperationsPerSecond();
832

833
    /* We have just 22 bits per object for LRU information.
834
     * So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
G
guiquanz 已提交
835
     * 2^22 bits with 10 seconds resolution is more or less 1.5 years.
836
     *
837
     * Note that even if this will wrap after 1.5 years it's not a problem,
838
     * everything will still work but just some object will appear younger
839 840 841 842 843
     * 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.
844
     */
845
    updateLRUClock();
846

847 848 849 850
    /* 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();

851 852 853
    /* 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) {
854
        if (prepareForShutdown(0) == REDIS_OK) exit(0);
855
        redisLog(REDIS_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
856
        server.shutdown_asap = 0;
857 858 859
    }

    /* Show some info about non-empty databases */
860 861 862 863 864 865 866 867 868 869 870
    run_with_period(5000) {
        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 (used || vkeys) {
                redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
                /* dictPrintStats(server.dict); */
            }
871 872 873
        }
    }

G
guiquanz 已提交
874
    /* We don't want to resize the hash tables while a background saving
875 876 877 878 879
     * 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. */
A
antirez 已提交
880
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
881
        tryResizeHashTables();
882 883 884 885
        if (server.activerehashing) incrementallyRehash();
    }

    /* Show information about connected clients */
886 887 888
    if (!server.sentinel_mode) {
        run_with_period(5000) {
            redisLog(REDIS_VERBOSE,
889
                "%lu clients connected (%lu slaves), %zu bytes in use",
890 891 892 893
                listLength(server.clients)-listLength(server.slaves),
                listLength(server.slaves),
                zmalloc_used_memory());
        }
894 895
    }

896 897
    /* We need to do a few operations on clients asynchronously. */
    clientsCron();
898

899 900
    /* Start a scheduled AOF rewrite if this was requested by the user while
     * a BGSAVE was in progress. */
A
antirez 已提交
901
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
902
        server.aof_rewrite_scheduled)
903 904 905 906
    {
        rewriteAppendOnlyFileBackground();
    }

A
antirez 已提交
907
    /* Check if a background saving or AOF rewrite in progress terminated. */
A
antirez 已提交
908
    if (server.rdb_child_pid != -1 || server.aof_child_pid != -1) {
909 910 911 912
        int statloc;
        pid_t pid;

        if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
913 914 915 916 917
            int exitcode = WEXITSTATUS(statloc);
            int bysignal = 0;
            
            if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);

A
antirez 已提交
918
            if (pid == server.rdb_child_pid) {
919
                backgroundSaveDoneHandler(exitcode,bysignal);
920
            } else if (pid == server.aof_child_pid) {
921
                backgroundRewriteDoneHandler(exitcode,bysignal);
922 923 924 925
            } else {
                redisLog(REDIS_WARNING,
                    "Warning, detected child with unmatched pid: %ld",
                    (long)pid);
926 927 928
            }
            updateDictResizePolicy();
        }
A
antirez 已提交
929
    } else {
930 931
        /* If there is not a background saving/rewrite in progress check if
         * we have to save/rewrite now */
932 933 934 935
         for (j = 0; j < server.saveparamslen; j++) {
            struct saveparam *sp = server.saveparams+j;

            if (server.dirty >= sp->changes &&
936
                server.unixtime-server.lastsave > sp->seconds) {
937
                redisLog(REDIS_NOTICE,"%d changes in %d seconds. Saving...",
938
                    sp->changes, (int)sp->seconds);
A
antirez 已提交
939
                rdbSaveBackground(server.rdb_filename);
940 941 942
                break;
            }
         }
943 944

         /* Trigger an AOF rewrite if needed */
A
antirez 已提交
945
         if (server.rdb_child_pid == -1 &&
A
antirez 已提交
946
             server.aof_child_pid == -1 &&
947 948
             server.aof_rewrite_perc &&
             server.aof_current_size > server.aof_rewrite_min_size)
949
         {
950 951 952 953
            long long base = server.aof_rewrite_base_size ?
                            server.aof_rewrite_base_size : 1;
            long long growth = (server.aof_current_size*100/base) - 100;
            if (growth >= server.aof_rewrite_perc) {
954
                redisLog(REDIS_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth);
955 956 957
                rewriteAppendOnlyFileBackground();
            }
         }
958 959
    }

960 961 962 963 964

    /* If we postponed an AOF buffer flush, let's try to do it every time the
     * cron function is called. */
    if (server.aof_flush_postponed_start) flushAppendOnlyFile(0);

965 966 967 968
    /* 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();
969

970 971 972
    /* Close clients that need to be closed asynchronous */
    freeClientsInAsyncFreeQueue();

973 974
    /* Replication cron function -- used to reconnect to master and
     * to detect transfer failures. */
975
    run_with_period(1000) replicationCron();
976

977 978 979 980 981
    /* Run the sentinel timer if we are in sentinel mode. */
    run_with_period(100) {
        if (server.sentinel_mode) sentinelTimer();
    }

982
    server.cronloops++;
983
    return 1000/server.hz;
984 985 986 987 988 989 990
}

/* 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);
991 992
    listNode *ln;
    redisClient *c;
993

994 995 996 997 998 999
    /* 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);
1000
        c->flags &= ~REDIS_UNBLOCKED;
1001 1002

        /* Process remaining data in the input buffer. */
1003 1004
        if (c->querybuf && sdslen(c->querybuf) > 0) {
            server.current_client = c;
1005
            processInputBuffer(c);
1006 1007
            server.current_client = NULL;
        }
1008 1009
    }

1010
    /* Write the AOF buffer on disk */
1011
    flushAppendOnlyFile(0);
1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
}

/* =========================== 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(
1032
        "-WRONGTYPE Operation against a key holding the wrong kind of value\r\n"));
1033 1034 1035 1036 1037 1038 1039 1040
    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"));
A
antirez 已提交
1041 1042
    shared.noscripterr = createObject(REDIS_STRING,sdsnew(
        "-NOSCRIPT No matching script. Please use EVAL.\r\n"));
1043 1044
    shared.loadingerr = createObject(REDIS_STRING,sdsnew(
        "-LOADING Redis is loading the dataset in memory\r\n"));
1045
    shared.slowscripterr = createObject(REDIS_STRING,sdsnew(
1046
        "-BUSY Redis is busy running a script. You can only call SCRIPT KILL or SHUTDOWN NOSAVE.\r\n"));
1047 1048
    shared.masterdownerr = createObject(REDIS_STRING,sdsnew(
        "-MASTERDOWN Link with MASTER is down and slave-serve-stale-data is set to 'no'.\r\n"));
1049
    shared.bgsaveerr = createObject(REDIS_STRING,sdsnew(
1050 1051 1052
        "-MISCONF Redis is configured to save RDB snapshots, but is currently not able to persist on disk. Commands that may modify the data set are disabled. Please check Redis logs for details about the error.\r\n"));
    shared.roslaveerr = createObject(REDIS_STRING,sdsnew(
        "-READONLY You can't write against a read only slave.\r\n"));
1053 1054
    shared.noautherr = createObject(REDIS_STRING,sdsnew(
        "-NOAUTH Authentication required.\r\n"));
1055 1056
    shared.oomerr = createObject(REDIS_STRING,sdsnew(
        "-OOM command not allowed when used memory > 'maxmemory'.\r\n"));
A
antirez 已提交
1057 1058
    shared.execaborterr = createObject(REDIS_STRING,sdsnew(
        "-EXECABORT Transaction discarded because of previous errors.\r\n"));
1059 1060 1061
    shared.space = createObject(REDIS_STRING,sdsnew(" "));
    shared.colon = createObject(REDIS_STRING,sdsnew(":"));
    shared.plus = createObject(REDIS_STRING,sdsnew("+"));
1062 1063

    for (j = 0; j < REDIS_SHARED_SELECT_CMDS; j++) {
1064 1065 1066 1067
        char dictid_str[64];
        int dictid_len;

        dictid_len = ll2string(dictid_str,sizeof(dictid_str),j);
1068
        shared.select[j] = createObject(REDIS_STRING,
1069 1070 1071
            sdscatprintf(sdsempty(),
                "*2\r\n$6\r\nSELECT\r\n$%d\r\n%s\r\n",
                dictid_len, dictid_str));
1072
    }
1073 1074 1075 1076 1077 1078
    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("$12\r\npunsubscribe\r\n",19);
1079
    shared.del = createStringObject("DEL",3);
1080 1081
    shared.rpop = createStringObject("RPOP",4);
    shared.lpop = createStringObject("LPOP",4);
1082
    shared.lpush = createStringObject("LPUSH",5);
1083 1084 1085 1086
    for (j = 0; j < REDIS_SHARED_INTEGERS; j++) {
        shared.integers[j] = createObject(REDIS_STRING,(void*)(long)j);
        shared.integers[j]->encoding = REDIS_ENCODING_INT;
    }
1087 1088 1089 1090 1091 1092
    for (j = 0; j < REDIS_SHARED_BULKHDR_LEN; j++) {
        shared.mbulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"*%d\r\n",j));
        shared.bulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"$%d\r\n",j));
    }
1093 1094 1095
}

void initServerConfig() {
A
antirez 已提交
1096
    getRandomHexChars(server.runid,REDIS_RUN_ID_SIZE);
1097
    server.hz = REDIS_DEFAULT_HZ;
A
antirez 已提交
1098
    server.runid[REDIS_RUN_ID_SIZE] = '\0';
1099
    server.arch_bits = (sizeof(long) == 8) ? 64 : 32;
1100
    server.port = REDIS_SERVERPORT;
1101
    server.bindaddr = NULL;
1102
    server.unixsocket = NULL;
1103
    server.unixsocketperm = 0;
1104 1105 1106
    server.ipfd = -1;
    server.sofd = -1;
    server.dbnum = REDIS_DEFAULT_DBNUM;
A
antirez 已提交
1107
    server.verbosity = REDIS_NOTICE;
1108
    server.maxidletime = REDIS_MAXIDLETIME;
1109
    server.tcpkeepalive = 0;
1110
    server.client_max_querybuf_len = REDIS_MAX_QUERYBUF_LEN;
1111
    server.saveparams = NULL;
1112
    server.loading = 0;
1113
    server.logfile = NULL; /* NULL = log on standard output */
J
Jonah H. Harris 已提交
1114 1115 1116
    server.syslog_enabled = 0;
    server.syslog_ident = zstrdup("redis");
    server.syslog_facility = LOG_LOCAL0;
1117
    server.daemonize = 0;
1118
    server.aof_state = REDIS_AOF_OFF;
1119 1120 1121 1122 1123 1124
    server.aof_fsync = AOF_FSYNC_EVERYSEC;
    server.aof_no_fsync_on_rewrite = 0;
    server.aof_rewrite_perc = REDIS_AOF_REWRITE_PERC;
    server.aof_rewrite_min_size = REDIS_AOF_REWRITE_MIN_SIZE;
    server.aof_rewrite_base_size = 0;
    server.aof_rewrite_scheduled = 0;
A
antirez 已提交
1125
    server.aof_last_fsync = time(NULL);
1126 1127
    server.aof_rewrite_time_last = -1;
    server.aof_rewrite_time_start = -1;
1128
    server.aof_lastbgrewrite_status = REDIS_OK;
1129
    server.aof_delayed_fsync = 0;
A
antirez 已提交
1130 1131
    server.aof_fd = -1;
    server.aof_selected_db = -1; /* Make sure the first time will not match */
1132
    server.aof_flush_postponed_start = 0;
1133
    server.pidfile = zstrdup("/var/run/redis.pid");
A
antirez 已提交
1134
    server.rdb_filename = zstrdup("dump.rdb");
1135
    server.aof_filename = zstrdup("appendonly.aof");
1136
    server.requirepass = NULL;
A
antirez 已提交
1137
    server.rdb_compression = 1;
1138
    server.rdb_checksum = 1;
1139
    server.activerehashing = 1;
A
antirez 已提交
1140
    server.notify_keyspace_events = 0;
1141
    server.maxclients = REDIS_MAX_CLIENTS;
1142
    server.bpop_blocked_clients = 0;
1143
    server.maxmemory = 0;
1144 1145
    server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU;
    server.maxmemory_samples = 3;
1146 1147
    server.hash_max_ziplist_entries = REDIS_HASH_MAX_ZIPLIST_ENTRIES;
    server.hash_max_ziplist_value = REDIS_HASH_MAX_ZIPLIST_VALUE;
1148 1149
    server.list_max_ziplist_entries = REDIS_LIST_MAX_ZIPLIST_ENTRIES;
    server.list_max_ziplist_value = REDIS_LIST_MAX_ZIPLIST_VALUE;
1150
    server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES;
1151 1152
    server.zset_max_ziplist_entries = REDIS_ZSET_MAX_ZIPLIST_ENTRIES;
    server.zset_max_ziplist_value = REDIS_ZSET_MAX_ZIPLIST_VALUE;
1153
    server.shutdown_asap = 0;
A
1fe4cd5  
antirez 已提交
1154 1155
    server.repl_ping_slave_period = REDIS_REPL_PING_SLAVE_PERIOD;
    server.repl_timeout = REDIS_REPL_TIMEOUT;
1156
    server.lua_caller = NULL;
A
antirez 已提交
1157
    server.lua_time_limit = REDIS_LUA_TIME_LIMIT;
1158
    server.lua_client = NULL;
1159
    server.lua_timedout = 0;
1160

A
antirez 已提交
1161
    updateLRUClock();
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
    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.masterauth = NULL;
    server.masterhost = NULL;
    server.masterport = 6379;
    server.master = NULL;
1172 1173
    server.cached_master = NULL;
    server.repl_master_initial_offset = -1;
A
antirez 已提交
1174
    server.repl_state = REDIS_REPL_NONE;
1175
    server.repl_syncio_timeout = REDIS_REPL_SYNCIO_TIMEOUT;
1176
    server.repl_serve_stale_data = 1;
1177
    server.repl_slave_ro = 1;
1178
    server.repl_down_since = time(NULL);
1179
    server.repl_disable_tcp_nodelay = 0;
1180
    server.slave_priority = REDIS_DEFAULT_SLAVE_PRIORITY;
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190
    server.master_repl_offset = 0;

    /* Replication partial resync backlog */
    server.repl_backlog = NULL;
    server.repl_backlog_size = REDIS_DEFAULT_REPL_BACKLOG_SIZE;
    server.repl_backlog_histlen = 0;
    server.repl_backlog_idx = 0;
    server.repl_backlog_off = 0;
    server.repl_backlog_time_limit = REDIS_DEFAULT_REPL_BACKLOG_TIME_LIMIT;
    server.repl_no_slaves_since = time(NULL);
1191

1192 1193 1194 1195
    /* Client output buffer limits */
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].hard_limit_bytes = 0;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_bytes = 0;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_seconds = 0;
1196 1197
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].hard_limit_bytes = 1024*1024*256;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_bytes = 1024*1024*64;
1198
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_seconds = 60;
1199 1200
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].hard_limit_bytes = 1024*1024*32;
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_bytes = 1024*1024*8;
1201 1202
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_seconds = 60;

1203 1204 1205 1206 1207
    /* 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;
1208

G
guiquanz 已提交
1209
    /* Command table -- we initiialize it here as it is part of the
1210 1211 1212 1213 1214 1215
     * 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");
1216
    server.lpushCommand = lookupCommandByCString("lpush");
1217 1218
    server.lpopCommand = lookupCommandByCString("lpop");
    server.rpopCommand = lookupCommandByCString("rpop");
1219 1220 1221 1222
    
    /* Slow log */
    server.slowlog_log_slower_than = REDIS_SLOWLOG_LOG_SLOWER_THAN;
    server.slowlog_max_len = REDIS_SLOWLOG_MAX_LEN;
A
antirez 已提交
1223

A
antirez 已提交
1224
    /* Debugging */
A
antirez 已提交
1225 1226 1227 1228
    server.assert_failed = "<no assertion failed>";
    server.assert_file = "<no file>";
    server.assert_line = 0;
    server.bug_report_start = 0;
A
antirez 已提交
1229
    server.watchdog_period = 0;
1230 1231
}

1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253
/* This function will try to raise the max number of open files accordingly to
 * the configured max number of clients. It will also account for 32 additional
 * file descriptors as we need a few more for persistence, listening
 * sockets, log files and so forth.
 *
 * If it will not be possible to set the limit accordingly to the configured
 * max number of clients, the function will do the reverse setting
 * server.maxclients to the value that we can actually handle. */
void adjustOpenFilesLimit(void) {
    rlim_t maxfiles = server.maxclients+32;
    struct rlimit limit;

    if (getrlimit(RLIMIT_NOFILE,&limit) == -1) {
        redisLog(REDIS_WARNING,"Unable to obtain the current NOFILE limit (%s), assuming 1024 and setting the max clients configuration accordingly.",
            strerror(errno));
        server.maxclients = 1024-32;
    } else {
        rlim_t oldlimit = limit.rlim_cur;

        /* Set the max number of files if the current limit is not enough
         * for our needs. */
        if (oldlimit < maxfiles) {
1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
            rlim_t f;
            
            f = maxfiles;
            while(f > oldlimit) {
                limit.rlim_cur = f;
                limit.rlim_max = f;
                if (setrlimit(RLIMIT_NOFILE,&limit) != -1) break;
                f -= 128;
            }
            if (f < oldlimit) f = oldlimit;
            if (f != maxfiles) {
                server.maxclients = f-32;
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275
                redisLog(REDIS_WARNING,"Unable to set the max number of files limit to %d (%s), setting the max clients configuration to %d.",
                    (int) maxfiles, strerror(errno), (int) server.maxclients);
            } else {
                redisLog(REDIS_NOTICE,"Max number of open files set to %d",
                    (int) maxfiles);
            }
        }
    }
}

1276 1277 1278 1279 1280
void initServer() {
    int j;

    signal(SIGHUP, SIG_IGN);
    signal(SIGPIPE, SIG_IGN);
1281
    setupSignalHandlers();
1282

J
Jonah H. Harris 已提交
1283 1284 1285 1286 1287
    if (server.syslog_enabled) {
        openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
            server.syslog_facility);
    }

1288
    server.current_client = NULL;
1289
    server.clients = listCreate();
1290
    server.clients_to_close = listCreate();
1291 1292
    server.slaves = listCreate();
    server.monitors = listCreate();
1293
    server.slaveseldb = -1; /* Force to emit the first SELECT command. */
1294
    server.unblocked_clients = listCreate();
1295
    server.ready_keys = listCreate();
1296

1297
    createSharedObjects();
1298 1299
    adjustOpenFilesLimit();
    server.el = aeCreateEventLoop(server.maxclients+1024);
1300
    server.db = zmalloc(sizeof(redisDb)*server.dbnum);
1301

1302
    if (server.port != 0) {
1303
        server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr);
1304
        if (server.ipfd == ANET_ERR) {
1305 1306
            redisLog(REDIS_WARNING, "Opening port %d: %s",
                server.port, server.neterr);
1307 1308
            exit(1);
        }
1309
    }
1310 1311
    if (server.unixsocket != NULL) {
        unlink(server.unixsocket); /* don't care if this fails */
1312
        server.sofd = anetUnixServer(server.neterr,server.unixsocket,server.unixsocketperm);
1313 1314 1315 1316
        if (server.sofd == ANET_ERR) {
            redisLog(REDIS_WARNING, "Opening socket: %s", server.neterr);
            exit(1);
        }
1317
    }
1318 1319
    if (server.ipfd < 0 && server.sofd < 0) {
        redisLog(REDIS_WARNING, "Configured to not listen anywhere, exiting.");
1320 1321 1322 1323 1324 1325
        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);
1326
        server.db[j].ready_keys = dictCreate(&setDictType,NULL);
1327 1328 1329 1330 1331 1332 1333 1334
        server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
        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;
A
antirez 已提交
1335
    server.rdb_child_pid = -1;
A
antirez 已提交
1336
    server.aof_child_pid = -1;
1337
    aofRewriteBufferReset();
A
antirez 已提交
1338
    server.aof_buf = sdsempty();
1339
    server.lastsave = time(NULL);
1340 1341
    server.rdb_save_time_last = -1;
    server.rdb_save_time_start = -1;
1342 1343 1344 1345
    server.dirty = 0;
    server.stat_numcommands = 0;
    server.stat_numconnections = 0;
    server.stat_expiredkeys = 0;
1346
    server.stat_evictedkeys = 0;
1347
    server.stat_starttime = time(NULL);
1348 1349
    server.stat_keyspace_misses = 0;
    server.stat_keyspace_hits = 0;
1350
    server.stat_peak_memory = 0;
1351
    server.stat_fork_time = 0;
1352
    server.stat_rejected_conn = 0;
1353 1354 1355
    server.stat_sync_full = 0;
    server.stat_sync_partial_ok = 0;
    server.stat_sync_partial_err = 0;
1356 1357 1358 1359
    memset(server.ops_sec_samples,0,sizeof(server.ops_sec_samples));
    server.ops_sec_idx = 0;
    server.ops_sec_last_sample_time = mstime();
    server.ops_sec_last_sample_ops = 0;
1360
    server.unixtime = time(NULL);
1361
    server.lastbgsave_status = REDIS_OK;
1362
    server.stop_writes_on_bgsave_err = 1;
stamhe's avatar
stamhe 已提交
1363
    if(aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL) == AE_ERR) {
1364
        redisPanic("Can't create the serverCron time event.");
stamhe's avatar
stamhe 已提交
1365 1366
        exit(1);
    }
1367
    if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE,
A
antirez 已提交
1368
        acceptTcpHandler,NULL) == AE_ERR) redisPanic("Unrecoverable error creating server.ipfd file event.");
1369
    if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
A
antirez 已提交
1370
        acceptUnixHandler,NULL) == AE_ERR) redisPanic("Unrecoverable error creating server.sofd file event.");
1371

1372
    if (server.aof_state == REDIS_AOF_ON) {
A
antirez 已提交
1373
        server.aof_fd = open(server.aof_filename,
1374
                               O_WRONLY|O_APPEND|O_CREAT,0644);
A
antirez 已提交
1375
        if (server.aof_fd == -1) {
1376 1377 1378 1379 1380 1381
            redisLog(REDIS_WARNING, "Can't open the append-only file: %s",
                strerror(errno));
            exit(1);
        }
    }

1382 1383
    /* 32 bit instances are limited to 4GB of address space, so if there is
     * no explicit limit in the user provided configuration we set a limit
1384 1385
     * at 3 GB using maxmemory with 'noeviction' policy'. This avoids
     * useless crashes of the Redis instance for out of memory. */
1386
    if (server.arch_bits == 32 && server.maxmemory == 0) {
1387 1388
        redisLog(REDIS_WARNING,"Warning: 32 bit instance detected but no memory limit set. Setting 3 GB maxmemory limit with 'noeviction' policy now.");
        server.maxmemory = 3072LL*(1024*1024); /* 3 GB */
1389 1390 1391
        server.maxmemory_policy = REDIS_MAXMEMORY_NO_EVICTION;
    }

1392
    scriptingInit();
1393
    slowlogInit();
1394
    bioInit();
1395 1396
}

1397 1398 1399 1400
/* Populates the Redis Command Table starting from the hard coded list
 * we have on top of redis.c file. */
void populateCommandTable(void) {
    int j;
1401
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1402 1403

    for (j = 0; j < numcommands; j++) {
1404
        struct redisCommand *c = redisCommandTable+j;
1405
        char *f = c->sflags;
1406
        int retval;
1407

1408 1409 1410 1411 1412 1413 1414 1415
        while(*f != '\0') {
            switch(*f) {
            case 'w': c->flags |= REDIS_CMD_WRITE; break;
            case 'r': c->flags |= REDIS_CMD_READONLY; break;
            case 'm': c->flags |= REDIS_CMD_DENYOOM; break;
            case 'a': c->flags |= REDIS_CMD_ADMIN; break;
            case 'p': c->flags |= REDIS_CMD_PUBSUB; break;
            case 'f': c->flags |= REDIS_CMD_FORCE_REPLICATION; break;
1416 1417
            case 's': c->flags |= REDIS_CMD_NOSCRIPT; break;
            case 'R': c->flags |= REDIS_CMD_RANDOM; break;
1418
            case 'S': c->flags |= REDIS_CMD_SORT_FOR_SCRIPT; break;
1419 1420
            case 'l': c->flags |= REDIS_CMD_LOADING; break;
            case 't': c->flags |= REDIS_CMD_STALE; break;
1421
            case 'M': c->flags |= REDIS_CMD_SKIP_MONITOR; break;
1422 1423 1424 1425 1426
            default: redisPanic("Unsupported command flag"); break;
            }
            f++;
        }

1427 1428 1429
        retval = dictAdd(server.commands, sdsnew(c->name), c);
        assert(retval == DICT_OK);
    }
1430 1431
}

1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
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;
    }
}

1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480
/* ========================== Redis OP Array API ============================ */

void redisOpArrayInit(redisOpArray *oa) {
    oa->ops = NULL;
    oa->numops = 0;
}

int redisOpArrayAppend(redisOpArray *oa, struct redisCommand *cmd, int dbid,
                       robj **argv, int argc, int target)
{
    redisOp *op;

    oa->ops = zrealloc(oa->ops,sizeof(redisOp)*(oa->numops+1));
    op = oa->ops+oa->numops;
    op->cmd = cmd;
    op->dbid = dbid;
    op->argv = argv;
    op->argc = argc;
    op->target = target;
    oa->numops++;
    return oa->numops;
}

void redisOpArrayFree(redisOpArray *oa) {
    while(oa->numops) {
        int j;
        redisOp *op;

        oa->numops--;
        op = oa->ops+oa->numops;
        for (j = 0; j < op->argc; j++)
            decrRefCount(op->argv[j]);
        zfree(op->argv);
    }
    zfree(oa->ops);
}

1481 1482
/* ====================== Commands lookup and execution ===================== */

1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493
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;
1494 1495
}

A
antirez 已提交
1496
/* Propagate the specified command (in the context of the specified database id)
1497
 * to AOF and Slaves.
A
antirez 已提交
1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508
 *
 * flags are an xor between:
 * + REDIS_PROPAGATE_NONE (no propagation of command at all)
 * + REDIS_PROPAGATE_AOF (propagate into the AOF file if is enabled)
 * + REDIS_PROPAGATE_REPL (propagate into the replication link)
 */
void propagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
               int flags)
{
    if (server.aof_state != REDIS_AOF_OFF && flags & REDIS_PROPAGATE_AOF)
        feedAppendOnlyFile(cmd,dbid,argv,argc);
1509
    if (flags & REDIS_PROPAGATE_REPL)
A
antirez 已提交
1510 1511 1512
        replicationFeedSlaves(server.slaves,dbid,argv,argc);
}

1513 1514
/* Used inside commands to schedule the propagation of additional commands
 * after the current command is propagated to AOF / Replication. */
1515 1516 1517
void alsoPropagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
                   int target)
{
1518
    redisOpArrayAppend(&server.also_propagate,cmd,dbid,argv,argc,target);
1519 1520
}

1521
/* Call() is the core of Redis execution of a command */
1522
void call(redisClient *c, int flags) {
1523
    long long dirty, start = ustime(), duration;
1524

A
antirez 已提交
1525
    /* Sent the command to clients in MONITOR mode, only if the commands are
G
guiquanz 已提交
1526
     * not generated from reading an AOF. */
1527 1528 1529 1530
    if (listLength(server.monitors) &&
        !server.loading &&
        !(c->cmd->flags & REDIS_CMD_SKIP_MONITOR))
    {
1531
        replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
1532
    }
A
antirez 已提交
1533 1534

    /* Call the command. */
1535
    redisOpArrayInit(&server.also_propagate);
1536
    dirty = server.dirty;
1537
    c->cmd->proc(c);
1538
    dirty = server.dirty-dirty;
1539
    duration = ustime()-start;
1540 1541 1542 1543 1544 1545

    /* When EVAL is called loading the AOF we don't want commands called
     * from Lua to go into the slowlog or to populate statistics. */
    if (server.loading && c->flags & REDIS_LUA_CLIENT)
        flags &= ~(REDIS_CALL_SLOWLOG | REDIS_CALL_STATS);

A
antirez 已提交
1546 1547
    /* Log the command into the Slow log if needed, and populate the
     * per-command statistics that we show in INFO commandstats. */
1548
    if (flags & REDIS_CALL_SLOWLOG && c->cmd->proc != execCommand)
1549 1550 1551 1552 1553
        slowlogPushEntryIfNeeded(c->argv,c->argc,duration);
    if (flags & REDIS_CALL_STATS) {
        c->cmd->microseconds += duration;
        c->cmd->calls++;
    }
A
antirez 已提交
1554 1555

    /* Propagate the command into the AOF and replication link */
1556
    if (flags & REDIS_CALL_PROPAGATE) {
A
antirez 已提交
1557 1558 1559 1560 1561 1562 1563 1564
        int flags = REDIS_PROPAGATE_NONE;

        if (c->cmd->flags & REDIS_CMD_FORCE_REPLICATION)
            flags |= REDIS_PROPAGATE_REPL;
        if (dirty)
            flags |= (REDIS_PROPAGATE_REPL | REDIS_PROPAGATE_AOF);
        if (flags != REDIS_PROPAGATE_NONE)
            propagate(c->cmd,c->db->id,c->argv,c->argc,flags);
1565
    }
1566 1567 1568

    /* Handle the alsoPropagate() API to handle commands that want to propagate
     * multiple separated commands. */
1569
    if (server.also_propagate.numops) {
1570
        int j;
1571
        redisOp *rop;
1572

1573 1574 1575 1576 1577
        for (j = 0; j < server.also_propagate.numops; j++) {
            rop = &server.also_propagate.ops[j];
            propagate(rop->cmd, rop->dbid, rop->argv, rop->argc, rop->target);
        }
        redisOpArrayFree(&server.also_propagate);
1578
    }
1579 1580 1581 1582
    server.stat_numcommands++;
}

/* If this function gets called we already read a whole
1583
 * command, arguments are in the client argv/argc fields.
1584 1585 1586 1587
 * 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
G
guiquanz 已提交
1588 1589
 * other operations can be performed by the caller. Otherwise
 * if 0 is returned the client was destroyed (i.e. after QUIT). */
1590
int processCommand(redisClient *c) {
P
Pieter Noordhuis 已提交
1591 1592 1593 1594
    /* 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. */
1595
    if (!strcasecmp(c->argv[0]->ptr,"quit")) {
P
Pieter Noordhuis 已提交
1596
        addReply(c,shared.ok);
1597
        c->flags |= REDIS_CLOSE_AFTER_REPLY;
1598
        return REDIS_ERR;
1599 1600 1601
    }

    /* Now lookup the command and check ASAP about trivial error conditions
1602
     * such as wrong arity, bad command name and so forth. */
1603
    c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
1604
    if (!c->cmd) {
A
antirez 已提交
1605
        flagTransaction(c);
1606 1607
        addReplyErrorFormat(c,"unknown command '%s'",
            (char*)c->argv[0]->ptr);
1608
        return REDIS_OK;
1609 1610
    } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) ||
               (c->argc < -c->cmd->arity)) {
A
antirez 已提交
1611
        flagTransaction(c);
1612
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
1613
            c->cmd->name);
1614
        return REDIS_OK;
1615 1616 1617
    }

    /* Check if the user is authenticated */
1618 1619
    if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand)
    {
A
antirez 已提交
1620
        flagTransaction(c);
1621
        addReply(c,shared.noautherr);
1622
        return REDIS_OK;
1623 1624
    }

1625 1626 1627 1628 1629
    /* 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. */
1630 1631 1632
    if (server.maxmemory) {
        int retval = freeMemoryIfNeeded();
        if ((c->cmd->flags & REDIS_CMD_DENYOOM) && retval == REDIS_ERR) {
A
antirez 已提交
1633
            flagTransaction(c);
1634
            addReply(c, shared.oomerr);
1635 1636
            return REDIS_OK;
        }
1637 1638
    }

1639
    /* Don't accept write commands if there are problems persisting on disk. */
1640 1641 1642
    if (server.stop_writes_on_bgsave_err &&
        server.saveparamslen > 0
        && server.lastbgsave_status == REDIS_ERR &&
1643 1644
        c->cmd->flags & REDIS_CMD_WRITE)
    {
A
antirez 已提交
1645
        flagTransaction(c);
1646 1647
        addReply(c, shared.bgsaveerr);
        return REDIS_OK;
1648 1649
    }

1650
    /* Don't accept write commands if this is a read only slave. But
1651 1652 1653 1654 1655 1656 1657
     * accept write commands if this is our master. */
    if (server.masterhost && server.repl_slave_ro &&
        !(c->flags & REDIS_MASTER) &&
        c->cmd->flags & REDIS_CMD_WRITE)
    {
        addReply(c, shared.roslaveerr);
        return REDIS_OK;
1658 1659
    }

1660 1661 1662
    /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */
    if ((dictSize(c->pubsub_channels) > 0 || listLength(c->pubsub_patterns) > 0)
        &&
1663 1664 1665 1666
        c->cmd->proc != subscribeCommand &&
        c->cmd->proc != unsubscribeCommand &&
        c->cmd->proc != psubscribeCommand &&
        c->cmd->proc != punsubscribeCommand) {
1667
        addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context");
1668
        return REDIS_OK;
1669 1670
    }

1671 1672
    /* Only allow INFO and SLAVEOF when slave-serve-stale-data is no and
     * we are a slave with a broken link with master. */
A
antirez 已提交
1673
    if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED &&
1674
        server.repl_serve_stale_data == 0 &&
1675
        !(c->cmd->flags & REDIS_CMD_STALE))
1676
    {
A
antirez 已提交
1677
        flagTransaction(c);
1678
        addReply(c, shared.masterdownerr);
1679 1680 1681
        return REDIS_OK;
    }

1682 1683 1684
    /* Loading DB? Return an error if the command has not the
     * REDIS_CMD_LOADING flag. */
    if (server.loading && !(c->cmd->flags & REDIS_CMD_LOADING)) {
1685 1686 1687 1688
        addReply(c, shared.loadingerr);
        return REDIS_OK;
    }

1689
    /* Lua script too slow? Only allow commands with REDIS_CMD_STALE flag. */
1690
    if (server.lua_timedout &&
1691
          c->cmd->proc != authCommand &&
1692
        !(c->cmd->proc == shutdownCommand &&
1693 1694 1695 1696 1697 1698
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'n') &&
        !(c->cmd->proc == scriptCommand &&
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'k'))
    {
A
antirez 已提交
1699
        flagTransaction(c);
1700 1701 1702 1703
        addReply(c, shared.slowscripterr);
        return REDIS_OK;
    }

1704 1705
    /* Exec the command */
    if (c->flags & REDIS_MULTI &&
1706 1707
        c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
        c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
1708
    {
1709
        queueMultiCommand(c);
1710 1711
        addReply(c,shared.queued);
    } else {
1712
        call(c,REDIS_CALL_FULL);
1713 1714
        if (listLength(server.ready_keys))
            handleClientsBlockedOnLists();
1715
    }
1716
    return REDIS_OK;
1717 1718 1719 1720
}

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

1721 1722 1723 1724
int prepareForShutdown(int flags) {
    int save = flags & REDIS_SHUTDOWN_SAVE;
    int nosave = flags & REDIS_SHUTDOWN_NOSAVE;

1725
    redisLog(REDIS_WARNING,"User requested shutdown...");
1726 1727 1728
    /* 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. */
A
antirez 已提交
1729
    if (server.rdb_child_pid != -1) {
1730
        redisLog(REDIS_WARNING,"There is a child saving an .rdb. Killing it!");
1731
        kill(server.rdb_child_pid,SIGUSR1);
A
antirez 已提交
1732
        rdbRemoveTempFile(server.rdb_child_pid);
1733
    }
1734
    if (server.aof_state != REDIS_AOF_OFF) {
1735 1736
        /* Kill the AOF saving child as the AOF we already have may be longer
         * but contains the full dataset anyway. */
A
antirez 已提交
1737
        if (server.aof_child_pid != -1) {
1738 1739
            redisLog(REDIS_WARNING,
                "There is a child rewriting the AOF. Killing it!");
1740
            kill(server.aof_child_pid,SIGUSR1);
1741
        }
1742
        /* Append only file: fsync() the AOF and exit */
1743
        redisLog(REDIS_NOTICE,"Calling fsync() on the AOF file.");
A
antirez 已提交
1744
        aof_fsync(server.aof_fd);
1745
    }
1746
    if ((server.saveparamslen > 0 && !nosave) || save) {
1747
        redisLog(REDIS_NOTICE,"Saving the final RDB snapshot before exiting.");
1748
        /* Snapshotting. Perform a SYNC SAVE and exit */
A
antirez 已提交
1749
        if (rdbSave(server.rdb_filename) != REDIS_OK) {
1750 1751 1752 1753 1754
            /* 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... */
1755
            redisLog(REDIS_WARNING,"Error trying to save the DB, can't exit.");
1756 1757 1758
            return REDIS_ERR;
        }
    }
1759 1760 1761 1762
    if (server.daemonize) {
        redisLog(REDIS_NOTICE,"Removing the pid file.");
        unlink(server.pidfile);
    }
1763 1764 1765
    /* Close the listening sockets. Apparently this allows faster restarts. */
    if (server.ipfd != -1) close(server.ipfd);
    if (server.sofd != -1) close(server.sofd);
1766 1767 1768 1769
    if (server.unixsocket) {
        redisLog(REDIS_NOTICE,"Removing the unix socket file.");
        unlink(server.unixsocket); /* don't care if this fails */
    }
1770

1771
    redisLog(REDIS_WARNING,"Redis is now ready to exit, bye bye...");
1772 1773 1774 1775 1776
    return REDIS_OK;
}

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

1777
/* Return zero if strings are the same, non-zero if they are not.
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 1812 1813 1814 1815 1816 1817 1818
 * The comparison is performed in a way that prevents an attacker to obtain
 * information about the nature of the strings just monitoring the execution
 * time of the function.
 *
 * Note that limiting the comparison length to strings up to 512 bytes we
 * can avoid leaking any information about the password length and any
 * possible branch misprediction related leak.
 */
int time_independent_strcmp(char *a, char *b) {
    char bufa[REDIS_AUTHPASS_MAX_LEN], bufb[REDIS_AUTHPASS_MAX_LEN];
    /* The above two strlen perform len(a) + len(b) operations where either
     * a or b are fixed (our password) length, and the difference is only
     * relative to the length of the user provided string, so no information
     * leak is possible in the following two lines of code. */
    int alen = strlen(a);
    int blen = strlen(b);
    int j;
    int diff = 0;

    /* We can't compare strings longer than our static buffers.
     * Note that this will never pass the first test in practical circumstances
     * so there is no info leak. */
    if (alen > sizeof(bufa) || blen > sizeof(bufb)) return 1;

    memset(bufa,0,sizeof(bufa));        /* Constant time. */
    memset(bufb,0,sizeof(bufb));        /* Constant time. */
    /* Again the time of the following two copies is proportional to
     * len(a) + len(b) so no info is leaked. */
    memcpy(bufa,a,alen);
    memcpy(bufb,b,blen);

    /* Always compare all the chars in the two buffers without
     * conditional expressions. */
    for (j = 0; j < sizeof(bufa); j++) {
        diff |= (bufa[j] ^ bufb[j]);
    }
    /* Length must be equal as well. */
    diff |= alen ^ blen;
    return diff; /* If zero strings are the same. */
}

1819
void authCommand(redisClient *c) {
1820 1821
    if (!server.requirepass) {
        addReplyError(c,"Client sent AUTH, but no password is set");
1822
    } else if (!time_independent_strcmp(c->argv[1]->ptr, server.requirepass)) {
1823 1824 1825 1826
      c->authenticated = 1;
      addReply(c,shared.ok);
    } else {
      c->authenticated = 0;
1827
      addReplyError(c,"invalid password");
1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838
    }
}

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

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

A
antirez 已提交
1839 1840 1841
void timeCommand(redisClient *c) {
    struct timeval tv;

G
guiquanz 已提交
1842
    /* gettimeofday() can only fail if &tv is a bad address so we
A
antirez 已提交
1843 1844 1845 1846 1847 1848 1849
     * don't check for errors. */
    gettimeofday(&tv,NULL);
    addReplyMultiBulkLen(c,2);
    addReplyBulkLongLong(c,tv.tv_sec);
    addReplyBulkLongLong(c,tv.tv_usec);
}

1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873
/* 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. */
1874 1875
sds genRedisInfoString(char *section) {
    sds info = sdsempty();
1876
    time_t uptime = server.unixtime-server.stat_starttime;
A
antirez 已提交
1877
    int j, numcommands;
1878
    struct rusage self_ru, c_ru;
1879
    unsigned long lol, bib;
1880 1881
    int allsections = 0, defsections = 0;
    int sections = 0;
1882

1883 1884
    if (section) {
        allsections = strcasecmp(section,"all") == 0;
1885
        defsections = strcasecmp(section,"default") == 0;
1886
    }
1887 1888 1889

    getrusage(RUSAGE_SELF, &self_ru);
    getrusage(RUSAGE_CHILDREN, &c_ru);
1890
    getClientsMaxBuffers(&lol,&bib);
1891 1892 1893

    /* Server */
    if (allsections || defsections || !strcasecmp(section,"server")) {
1894
        struct utsname name;
1895
        char *mode;
1896

A
antirez 已提交
1897
        if (server.sentinel_mode) mode = "sentinel";
1898 1899
        else mode = "standalone";
    
1900
        if (sections++) info = sdscat(info,"\r\n");
1901
        uname(&name);
1902
        info = sdscatprintf(info,
1903 1904 1905 1906
            "# Server\r\n"
            "redis_version:%s\r\n"
            "redis_git_sha1:%s\r\n"
            "redis_git_dirty:%d\r\n"
1907
            "redis_build_id:%llx\r\n"
1908
            "redis_mode:%s\r\n"
1909
            "os:%s %s %s\r\n"
1910
            "arch_bits:%d\r\n"
1911
            "multiplexing_api:%s\r\n"
A
antirez 已提交
1912
            "gcc_version:%d.%d.%d\r\n"
1913
            "process_id:%ld\r\n"
A
antirez 已提交
1914
            "run_id:%s\r\n"
1915 1916 1917
            "tcp_port:%d\r\n"
            "uptime_in_seconds:%ld\r\n"
            "uptime_in_days:%ld\r\n"
1918
            "hz:%d\r\n"
1919 1920 1921 1922
            "lru_clock:%ld\r\n",
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
1923
            (unsigned long long) redisBuildId(),
1924
            mode,
1925
            name.sysname, name.release, name.machine,
1926
            server.arch_bits,
1927
            aeGetApiName(),
A
antirez 已提交
1928 1929 1930 1931 1932
#ifdef __GNUC__
            __GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__,
#else
            0,0,0,
#endif
1933
            (long) getpid(),
A
antirez 已提交
1934
            server.runid,
1935 1936 1937
            server.port,
            uptime,
            uptime/(3600*24),
1938
            server.hz,
1939 1940 1941 1942 1943 1944 1945 1946
            (unsigned long) server.lruclock);
    }

    /* Clients */
    if (allsections || defsections || !strcasecmp(section,"clients")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Clients\r\n"
1947
            "connected_clients:%lu\r\n"
1948 1949 1950 1951 1952 1953 1954 1955 1956 1957
            "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")) {
1958 1959 1960 1961 1962
        char hmem[64];
        char peak_hmem[64];

        bytesToHuman(hmem,zmalloc_used_memory());
        bytesToHuman(peak_hmem,server.stat_peak_memory);
1963 1964 1965 1966 1967 1968
        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"
1969 1970
            "used_memory_peak:%zu\r\n"
            "used_memory_peak_human:%s\r\n"
A
antirez 已提交
1971
            "used_memory_lua:%lld\r\n"
1972
            "mem_fragmentation_ratio:%.2f\r\n"
1973
            "mem_allocator:%s\r\n",
1974 1975 1976
            zmalloc_used_memory(),
            hmem,
            zmalloc_get_rss(),
1977 1978
            server.stat_peak_memory,
            peak_hmem,
A
antirez 已提交
1979
            ((long long)lua_gc(server.lua,LUA_GCCOUNT,0))*1024LL,
1980
            zmalloc_get_fragmentation_ratio(),
1981
            ZMALLOC_LIB
1982
            );
1983 1984
    }

1985 1986 1987
    /* Persistence */
    if (allsections || defsections || !strcasecmp(section,"persistence")) {
        if (sections++) info = sdscat(info,"\r\n");
1988
        info = sdscatprintf(info,
1989 1990
            "# Persistence\r\n"
            "loading:%d\r\n"
1991 1992 1993 1994 1995 1996
            "rdb_changes_since_last_save:%lld\r\n"
            "rdb_bgsave_in_progress:%d\r\n"
            "rdb_last_save_time:%ld\r\n"
            "rdb_last_bgsave_status:%s\r\n"
            "rdb_last_bgsave_time_sec:%ld\r\n"
            "rdb_current_bgsave_time_sec:%ld\r\n"
1997
            "aof_enabled:%d\r\n"
1998 1999 2000
            "aof_rewrite_in_progress:%d\r\n"
            "aof_rewrite_scheduled:%d\r\n"
            "aof_last_rewrite_time_sec:%ld\r\n"
2001 2002
            "aof_current_rewrite_time_sec:%ld\r\n"
            "aof_last_bgrewrite_status:%s\r\n",
2003 2004
            server.loading,
            server.dirty,
A
antirez 已提交
2005
            server.rdb_child_pid != -1,
2006
            server.lastsave,
2007
            (server.lastbgsave_status == REDIS_OK) ? "ok" : "err",
2008 2009 2010 2011
            server.rdb_save_time_last,
            (server.rdb_child_pid == -1) ?
                -1 : time(NULL)-server.rdb_save_time_start,
            server.aof_state != REDIS_AOF_OFF,
2012
            server.aof_child_pid != -1,
2013 2014 2015
            server.aof_rewrite_scheduled,
            server.aof_rewrite_time_last,
            (server.aof_child_pid == -1) ?
2016 2017
                -1 : time(NULL)-server.aof_rewrite_time_start,
            (server.aof_lastbgrewrite_status == REDIS_OK) ? "ok" : "err");
2018

2019
        if (server.aof_state != REDIS_AOF_OFF) {
2020 2021 2022
            info = sdscatprintf(info,
                "aof_current_size:%lld\r\n"
                "aof_base_size:%lld\r\n"
2023 2024
                "aof_pending_rewrite:%d\r\n"
                "aof_buffer_length:%zu\r\n"
2025
                "aof_rewrite_buffer_length:%lu\r\n"
2026 2027
                "aof_pending_bio_fsync:%llu\r\n"
                "aof_delayed_fsync:%lu\r\n",
2028 2029 2030
                (long long) server.aof_current_size,
                (long long) server.aof_rewrite_base_size,
                server.aof_rewrite_scheduled,
A
antirez 已提交
2031
                sdslen(server.aof_buf),
2032
                aofRewriteBufferSize(),
2033 2034
                bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC),
                server.aof_delayed_fsync);
2035 2036
        }

2037 2038 2039 2040 2041 2042 2043 2044 2045
        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;

2046
            elapsed = server.unixtime-server.loading_start_time;
2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066
            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
            );
        }
2067
    }
2068 2069 2070 2071

    /* Stats */
    if (allsections || defsections || !strcasecmp(section,"stats")) {
        if (sections++) info = sdscat(info,"\r\n");
2072
        info = sdscatprintf(info,
2073 2074 2075
            "# Stats\r\n"
            "total_connections_received:%lld\r\n"
            "total_commands_processed:%lld\r\n"
2076
            "instantaneous_ops_per_sec:%lld\r\n"
2077
            "rejected_connections:%lld\r\n"
2078 2079 2080
            "sync_full:%lld\r\n"
            "sync_partial_ok:%lld\r\n"
            "sync_partial_err:%lld\r\n"
2081 2082 2083 2084 2085
            "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"
2086
            "pubsub_patterns:%lu\r\n"
2087
            "latest_fork_usec:%lld\r\n",
2088 2089
            server.stat_numconnections,
            server.stat_numcommands,
2090
            getOperationsPerSecond(),
2091
            server.stat_rejected_conn,
2092 2093 2094
            server.stat_sync_full,
            server.stat_sync_partial_ok,
            server.stat_sync_partial_err,
2095 2096 2097 2098 2099
            server.stat_expiredkeys,
            server.stat_evictedkeys,
            server.stat_keyspace_hits,
            server.stat_keyspace_misses,
            dictSize(server.pubsub_channels),
2100 2101
            listLength(server.pubsub_patterns),
            server.stat_fork_time);
2102
    }
A
antirez 已提交
2103

2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
    /* 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"
2118
                "slave_repl_offset:%lld\r\n"
2119 2120
                ,server.masterhost,
                server.masterport,
A
antirez 已提交
2121
                (server.repl_state == REDIS_REPL_CONNECTED) ?
2122 2123
                    "up" : "down",
                server.master ?
2124
                ((int)(server.unixtime-server.master->lastinteraction)) : -1,
2125 2126
                server.repl_state == REDIS_REPL_TRANSFER,
                server.master ? server.master->reploff : -1
2127 2128
            );

A
antirez 已提交
2129
            if (server.repl_state == REDIS_REPL_TRANSFER) {
2130
                info = sdscatprintf(info,
2131
                    "master_sync_left_bytes:%lld\r\n"
2132
                    "master_sync_last_io_seconds_ago:%d\r\n"
2133 2134
                    , (long long)
                        (server.repl_transfer_size - server.repl_transfer_read),
2135
                    (int)(server.unixtime-server.repl_transfer_lastio)
2136 2137
                );
            }
2138

A
antirez 已提交
2139
            if (server.repl_state != REDIS_REPL_CONNECTED) {
2140 2141
                info = sdscatprintf(info,
                    "master_link_down_since_seconds:%ld\r\n",
2142
                    (long)server.unixtime-server.repl_down_since);
2143
            }
2144
            info = sdscatprintf(info,
2145 2146 2147 2148
                "slave_priority:%d\r\n"
                "slave_read_only:%d\r\n",
                server.slave_priority,
                server.repl_slave_ro);
A
antirez 已提交
2149
        }
2150
        info = sdscatprintf(info,
2151
            "connected_slaves:%lu\r\n",
2152
            listLength(server.slaves));
2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179
        if (listLength(server.slaves)) {
            int slaveid = 0;
            listNode *ln;
            listIter li;

            listRewind(server.slaves,&li);
            while((ln = listNext(&li))) {
                redisClient *slave = listNodeValue(ln);
                char *state = NULL;
                char ip[32];
                int port;

                if (anetPeerToString(slave->fd,ip,&port) == -1) continue;
                switch(slave->replstate) {
                case REDIS_REPL_WAIT_BGSAVE_START:
                case REDIS_REPL_WAIT_BGSAVE_END:
                    state = "wait_bgsave";
                    break;
                case REDIS_REPL_SEND_BULK:
                    state = "send_bulk";
                    break;
                case REDIS_REPL_ONLINE:
                    state = "online";
                    break;
                }
                if (state == NULL) continue;
                info = sdscatprintf(info,"slave%d:%s,%d,%s\r\n",
A
antirez 已提交
2180
                    slaveid,ip,slave->slave_listening_port,state);
2181 2182 2183
                slaveid++;
            }
        }
2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
        info = sdscatprintf(info,
            "master_repl_offset:%lld\r\n"
            "repl_backlog_active:%d\r\n"
            "repl_backlog_size:%lld\r\n"
            "repl_backlog_first_byte_offset:%lld\r\n"
            "repl_backlog_histlen:%lld\r\n",
            server.master_repl_offset,
            server.repl_backlog != NULL,
            server.repl_backlog_size,
            server.repl_backlog_off,
            server.repl_backlog_histlen);
A
antirez 已提交
2195 2196
    }

2197 2198
    /* CPU */
    if (allsections || defsections || !strcasecmp(section,"cpu")) {
2199 2200
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
2201
        "# CPU\r\n"
2202 2203
        "used_cpu_sys:%.2f\r\n"
        "used_cpu_user:%.2f\r\n"
2204 2205
        "used_cpu_sys_children:%.2f\r\n"
        "used_cpu_user_children:%.2f\r\n",
2206
        (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000,
2207 2208 2209
        (float)self_ru.ru_utime.tv_sec+(float)self_ru.ru_utime.tv_usec/1000000,
        (float)c_ru.ru_stime.tv_sec+(float)c_ru.ru_stime.tv_usec/1000000,
        (float)c_ru.ru_utime.tv_sec+(float)c_ru.ru_utime.tv_usec/1000000);
2210
    }
2211

2212 2213 2214 2215
    /* cmdtime */
    if (allsections || !strcasecmp(section,"commandstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Commandstats\r\n");
2216
        numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
2217
        for (j = 0; j < numcommands; j++) {
2218
            struct redisCommand *c = redisCommandTable+j;
2219

2220 2221 2222 2223 2224
            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));
2225
        }
A
antirez 已提交
2226 2227
    }

2228 2229 2230 2231 2232 2233
    /* 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;
2234

2235 2236 2237 2238 2239 2240
            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);
            }
2241 2242 2243 2244 2245 2246
        }
    }
    return info;
}

void infoCommand(redisClient *c) {
2247 2248 2249 2250 2251 2252 2253
    char *section = c->argc == 2 ? c->argv[1]->ptr : "default";

    if (c->argc > 2) {
        addReply(c,shared.syntaxerr);
        return;
    }
    sds info = genRedisInfoString(section);
2254 2255 2256 2257 2258 2259 2260
    addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
        (unsigned long)sdslen(info)));
    addReplySds(c,info);
    addReply(c,shared.crlf);
}

void monitorCommand(redisClient *c) {
2261
    /* ignore MONITOR if already slave or in monitor mode */
2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
    if (c->flags & REDIS_SLAVE) return;

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

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

/* This function gets called when 'maxmemory' is set on the config file to limit
2272
 * the max memory used by the server, before processing a command.
2273
 *
2274 2275
 * The goal of the function is to free enough memory to keep Redis under the
 * configured memory limit.
2276
 *
2277 2278 2279 2280 2281 2282 2283 2284
 * The function starts calculating how many bytes should be freed to keep
 * Redis under the limit, and enters a loop selecting the best keys to
 * evict accordingly to the configured policy.
 *
 * If all the bytes needed to return back under the limit were freed the
 * function returns REDIS_OK, otherwise REDIS_ERR is returned, and the caller
 * should block the execution of commands that will result in more memory
 * used by the server.
2285
 */
2286 2287 2288 2289
int freeMemoryIfNeeded(void) {
    size_t mem_used, mem_tofree, mem_freed;
    int slaves = listLength(server.slaves);

2290 2291
    /* Remove the size of slaves output buffers and AOF buffer from the
     * count of used memory. */
2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306
    mem_used = zmalloc_used_memory();
    if (slaves) {
        listIter li;
        listNode *ln;

        listRewind(server.slaves,&li);
        while((ln = listNext(&li))) {
            redisClient *slave = listNodeValue(ln);
            unsigned long obuf_bytes = getClientOutputBufferMemoryUsage(slave);
            if (obuf_bytes > mem_used)
                mem_used = 0;
            else
                mem_used -= obuf_bytes;
        }
    }
2307 2308
    if (server.aof_state != REDIS_AOF_OFF) {
        mem_used -= sdslen(server.aof_buf);
2309
        mem_used -= aofRewriteBufferSize();
2310
    }
2311

2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322
    /* Check if we are over the memory limit. */
    if (mem_used <= server.maxmemory) return REDIS_OK;

    if (server.maxmemory_policy == REDIS_MAXMEMORY_NO_EVICTION)
        return REDIS_ERR; /* We need to free memory, but policy forbids. */

    /* Compute how much memory we need to free. */
    mem_tofree = mem_used - server.maxmemory;
    mem_freed = 0;
    while (mem_freed < mem_tofree) {
        int j, k, keys_freed = 0;
2323

2324
        for (j = 0; j < server.dbnum; j++) {
2325
            long bestval = 0; /* just to prevent warning */
2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
            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);
2345
                bestkey = dictGetKey(de);
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357
            }

            /* 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);
2358
                    thiskey = dictGetKey(de);
2359
                    /* When policy is volatile-lru we need an additional lookup
2360 2361 2362
                     * 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);
2363
                    o = dictGetVal(de);
2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380
                    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);
2381 2382
                    thiskey = dictGetKey(de);
                    thisval = (long) dictGetVal(de);
2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394

                    /* 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) {
2395 2396
                long long delta;

2397
                robj *keyobj = createStringObject(bestkey,sdslen(bestkey));
2398
                propagateExpire(db,keyobj);
2399 2400 2401 2402 2403 2404 2405 2406 2407
                /* We compute the amount of memory freed by dbDelete() alone.
                 * It is possible that actually the memory needed to propagate
                 * the DEL in AOF and replication link is greater than the one
                 * we are freeing removing the key, but we can't account for
                 * that otherwise we would never exit the loop.
                 *
                 * AOF and Output buffer memory will be freed eventually so
                 * we only care about memory used by the key space. */
                delta = (long long) zmalloc_used_memory();
2408
                dbDelete(db,keyobj);
2409 2410
                delta -= (long long) zmalloc_used_memory();
                mem_freed += delta;
2411
                server.stat_evictedkeys++;
2412 2413
                notifyKeyspaceEvent(REDIS_NOTIFY_EVICTED, "evicted",
                    keyobj, db->id);
2414
                decrRefCount(keyobj);
2415 2416 2417 2418 2419 2420
                keys_freed++;

                /* When the memory to free starts to be big enough, we may
                 * start spending so much time here that is impossible to
                 * deliver data to the slaves fast enough, so we force the
                 * transmission here inside the loop. */
A
antirez 已提交
2421
                if (slaves) flushSlavesOutputBuffers();
2422 2423
            }
        }
2424
        if (!keys_freed) return REDIS_ERR; /* nothing to free... */
2425
    }
2426
    return REDIS_OK;
2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452
}

/* =================================== 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__ */

2453 2454 2455 2456
void createPidFile(void) {
    /* Try to write the pid file in a best-effort way. */
    FILE *fp = fopen(server.pidfile,"w");
    if (fp) {
2457
        fprintf(fp,"%d\n",(int)getpid());
2458 2459 2460 2461
        fclose(fp);
    }
}

2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479
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() {
2480
    printf("Redis server v=%s sha=%s:%d malloc=%s bits=%d build=%llx\n",
A
antirez 已提交
2481 2482 2483 2484
        REDIS_VERSION,
        redisGitSHA1(),
        atoi(redisGitDirty()) > 0,
        ZMALLOC_LIB,
2485
        sizeof(long) == 4 ? 32 : 64,
2486
        (unsigned long long) redisBuildId());
2487 2488 2489 2490
    exit(0);
}

void usage() {
2491
    fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf] [options]\n");
2492
    fprintf(stderr,"       ./redis-server - (read config from stdin)\n");
2493
    fprintf(stderr,"       ./redis-server -v or --version\n");
2494 2495
    fprintf(stderr,"       ./redis-server -h or --help\n");
    fprintf(stderr,"       ./redis-server --test-memory <megabytes>\n\n");
2496 2497 2498 2499 2500
    fprintf(stderr,"Examples:\n");
    fprintf(stderr,"       ./redis-server (run the server with default conf)\n");
    fprintf(stderr,"       ./redis-server /etc/redis/6379.conf\n");
    fprintf(stderr,"       ./redis-server --port 7777\n");
    fprintf(stderr,"       ./redis-server --port 7777 --slaveof 127.0.0.1 8888\n");
2501 2502 2503
    fprintf(stderr,"       ./redis-server /etc/myredis.conf --loglevel verbose\n\n");
    fprintf(stderr,"Sentinel mode:\n");
    fprintf(stderr,"       ./redis-server /etc/sentinel.conf --sentinel\n");
2504 2505 2506
    exit(1);
}

A
antirez 已提交
2507 2508 2509
void redisAsciiArt(void) {
#include "asciilogo.h"
    char *buf = zmalloc(1024*16);
2510 2511
    char *mode = "stand alone";

A
antirez 已提交
2512
    if (server.sentinel_mode) mode = "sentinel";
A
antirez 已提交
2513 2514 2515 2516 2517 2518

    snprintf(buf,1024*16,ascii_logo,
        REDIS_VERSION,
        redisGitSHA1(),
        strtol(redisGitDirty(),NULL,10) > 0,
        (sizeof(long) == 8) ? "64" : "32",
2519
        mode, server.port,
A
antirez 已提交
2520 2521 2522 2523 2524 2525
        (long) getpid()
    );
    redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf);
    zfree(buf);
}

2526
static void sigtermHandler(int sig) {
2527 2528
    REDIS_NOTUSED(sig);

A
antirez 已提交
2529
    redisLogFromHandler(REDIS_WARNING,"Received SIGTERM, scheduling shutdown...");
2530 2531 2532
    server.shutdown_asap = 1;
}

2533
void setupSignalHandlers(void) {
2534 2535
    struct sigaction act;

2536 2537 2538
    /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used.
     * Otherwise, sa_handler is used. */
    sigemptyset(&act.sa_mask);
2539
    act.sa_flags = 0;
2540
    act.sa_handler = sigtermHandler;
2541
    sigaction(SIGTERM, &act, NULL);
2542

2543 2544
#ifdef HAVE_BACKTRACE
    sigemptyset(&act.sa_mask);
2545
    act.sa_flags = SA_NODEFER | SA_RESETHAND | SA_SIGINFO;
2546 2547 2548 2549 2550 2551 2552
    act.sa_sigaction = sigsegvHandler;
    sigaction(SIGSEGV, &act, NULL);
    sigaction(SIGBUS, &act, NULL);
    sigaction(SIGFPE, &act, NULL);
    sigaction(SIGILL, &act, NULL);
#endif
    return;
2553 2554
}

2555 2556
void memtest(size_t megabytes, int passes);

2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582
/* Returns 1 if there is --sentinel among the arguments or if
 * argv[0] is exactly "redis-sentinel". */
int checkForSentinelMode(int argc, char **argv) {
    int j;

    if (strstr(argv[0],"redis-sentinel") != NULL) return 1;
    for (j = 1; j < argc; j++)
        if (!strcmp(argv[j],"--sentinel")) return 1;
    return 0;
}

/* Function called at startup to load RDB or AOF file in memory. */
void loadDataFromDisk(void) {
    long long start = ustime();
    if (server.aof_state == REDIS_AOF_ON) {
        if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
            redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
    } else {
        if (rdbLoad(server.rdb_filename) == REDIS_OK) {
            redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",
                (float)(ustime()-start)/1000000);
        } else if (errno != ENOENT) {
            redisLog(REDIS_WARNING,"Fatal error loading the DB. Exiting.");
            exit(1);
        }
    }
A
antirez 已提交
2583 2584 2585 2586 2587 2588
}

void redisOutOfMemoryHandler(size_t allocation_size) {
    redisLog(REDIS_WARNING,"Out Of Memory allocating %zu bytes!",
        allocation_size);
    redisPanic("OOM");
A
antirez 已提交
2589 2590
}

2591
void redisSetProcTitle(char *title) {
2592
#ifdef USE_SETPROCTITLE
2593 2594 2595 2596
    setproctitle("%s %s:%d",
        title,
        server.bindaddr ? server.bindaddr : "*",
        server.port);
2597 2598 2599
#else
    REDIS_NOTUSED(title);
#endif
2600 2601
}

2602
int main(int argc, char **argv) {
2603
    struct timeval tv;
2604

A
antirez 已提交
2605
    /* We need to initialize our libraries, and the server configuration. */
2606 2607 2608
#ifdef INIT_SETPROCTITLE_REPLACEMENT
    spt_init(argc, argv);
#endif
2609
    zmalloc_enable_thread_safeness();
A
antirez 已提交
2610
    zmalloc_set_oom_handler(redisOutOfMemoryHandler);
2611 2612 2613
    srand(time(NULL)^getpid());
    gettimeofday(&tv,NULL);
    dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
2614
    server.sentinel_mode = checkForSentinelMode(argc,argv);
2615
    initServerConfig();
2616

2617 2618 2619 2620 2621 2622 2623 2624
    /* We need to init sentinel right now as parsing the configuration file
     * in sentinel mode will have the effect of populating the sentinel
     * data structures with master nodes to monitor. */
    if (server.sentinel_mode) {
        initSentinelConfig();
        initSentinel();
    }

2625 2626 2627 2628 2629 2630
    if (argc >= 2) {
        int j = 1; /* First option to parse in argv[] */
        sds options = sdsempty();
        char *configfile = NULL;

        /* Handle special options --help and --version */
2631 2632
        if (strcmp(argv[1], "-v") == 0 ||
            strcmp(argv[1], "--version") == 0) version();
2633 2634
        if (strcmp(argv[1], "--help") == 0 ||
            strcmp(argv[1], "-h") == 0) usage();
2635 2636
        if (strcmp(argv[1], "--test-memory") == 0) {
            if (argc == 3) {
2637
                memtest(atoi(argv[2]),50);
2638 2639 2640 2641 2642 2643 2644 2645
                exit(0);
            } else {
                fprintf(stderr,"Please specify the amount of memory to test in megabytes.\n");
                fprintf(stderr,"Example: ./redis-server --test-memory 4096\n\n");
                exit(1);
            }
        }

2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665
        /* First argument is the config file name? */
        if (argv[j][0] != '-' || argv[j][1] != '-')
            configfile = argv[j++];
        /* All the other options are parsed and conceptually appended to the
         * configuration file. For instance --port 6380 will generate the
         * string "port 6380\n" to be parsed after the actual file name
         * is parsed, if any. */
        while(j != argc) {
            if (argv[j][0] == '-' && argv[j][1] == '-') {
                /* Option name */
                if (sdslen(options)) options = sdscat(options,"\n");
                options = sdscat(options,argv[j]+2);
                options = sdscat(options," ");
            } else {
                /* Option argument */
                options = sdscatrepr(options,argv[j],strlen(argv[j]));
                options = sdscat(options," ");
            }
            j++;
        }
2666
        resetServerSaveParams();
2667 2668
        loadServerConfig(configfile,options);
        sdsfree(options);
2669
    } else {
2670
        redisLog(REDIS_WARNING, "Warning: no config file specified, using the default config. In order to specify a config file use %s /path/to/%s.conf", argv[0], server.sentinel_mode ? "sentinel" : "redis");
2671 2672 2673 2674
    }
    if (server.daemonize) daemonize();
    initServer();
    if (server.daemonize) createPidFile();
2675
    redisSetProcTitle(argv[0]);
2676
    redisAsciiArt();
2677 2678

    if (!server.sentinel_mode) {
2679
        /* Things only needed when not running in Sentinel mode. */
2680 2681 2682 2683 2684 2685 2686 2687 2688
        redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
    #ifdef __linux__
        linuxOvercommitMemoryWarning();
    #endif
        loadDataFromDisk();
        if (server.ipfd > 0)
            redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
        if (server.sofd > 0)
            redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
2689
    }
2690

2691 2692 2693 2694 2695
    /* Warning the user about suspicious maxmemory setting. */
    if (server.maxmemory > 0 && server.maxmemory < 1024*1024) {
        redisLog(REDIS_WARNING,"WARNING: You specified a maxmemory value that is less than 1MB (current value is %llu bytes). Are you sure this is what you really want?", server.maxmemory);
    }

2696 2697 2698 2699 2700 2701
    aeSetBeforeSleepProc(server.el,beforeSleep);
    aeMain(server.el);
    aeDeleteEventLoop(server.el);
    return 0;
}

2702
/* The End */