redis.c 92.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
/*
 * Copyright (c) 2009-2010, Salvatore Sanfilippo <antirez at gmail dot com>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include "redis.h"
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.
102 103
 * f: force replication of this command, regarless of server.dirty.
 * 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
106 107 108 109
 *    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.
 */
110
struct redisCommand redisCommandTable[] = {
111 112 113
    {"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},
114 115
    {"setex",setexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
    {"psetex",psetexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
    {"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
135 136 137
    {"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},
138 139 140 141 142 143 144 145 146 147 148 149
    {"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},
150
    {"spop",spopCommand,2,"wRs",0,NULL,1,1,1,0,0},
151
    {"srandmember",srandmemberCommand,2,"rR",0,NULL,1,1,1,0,0},
152
    {"sinter",sinterCommand,-2,"rS",0,NULL,1,-1,1,0,0},
153
    {"sinterstore",sinterstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
154
    {"sunion",sunionCommand,-2,"rS",0,NULL,1,-1,1,0,0},
155
    {"sunionstore",sunionstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
156
    {"sdiff",sdiffCommand,-2,"rS",0,NULL,1,-1,1,0,0},
157
    {"sdiffstore",sdiffstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
158
    {"smembers",sinterCommand,2,"rS",0,NULL,1,1,1,0,0},
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
    {"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
181
    {"hincrbyfloat",hincrbyfloatCommand,4,"wm",0,NULL,1,1,1,0,0},
182 183
    {"hdel",hdelCommand,-3,"w",0,NULL,1,1,1,0,0},
    {"hlen",hlenCommand,2,"r",0,NULL,1,1,1,0,0},
184 185
    {"hkeys",hkeysCommand,2,"rS",0,NULL,1,1,1,0,0},
    {"hvals",hvalsCommand,2,"rS",0,NULL,1,1,1,0,0},
186 187 188 189
    {"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
190
    {"incrbyfloat",incrbyfloatCommand,3,"wm",0,NULL,1,1,1,0,0},
191 192 193
    {"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},
194
    {"randomkey",randomkeyCommand,1,"rR",0,NULL,0,0,0,0,0},
195 196 197 198
    {"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},
199 200 201 202
    {"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},
203
    {"keys",keysCommand,2,"rS",0,NULL,0,0,0,0,0},
204
    {"dbsize",dbsizeCommand,1,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
205
    {"auth",authCommand,2,"rs",0,NULL,0,0,0,0,0},
206 207
    {"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
208
    {"save",saveCommand,1,"ars",0,NULL,0,0,0,0,0},
209 210
    {"bgsave",bgsaveCommand,1,"ar",0,NULL,0,0,0,0,0},
    {"bgrewriteaof",bgrewriteaofCommand,1,"ar",0,NULL,0,0,0,0,0},
211
    {"shutdown",shutdownCommand,-1,"ar",0,NULL,0,0,0,0,0},
212 213
    {"lastsave",lastsaveCommand,1,"r",0,NULL,0,0,0,0,0},
    {"type",typeCommand,2,"r",0,NULL,1,1,1,0,0},
214
    {"multi",multiCommand,1,"rs",0,NULL,0,0,0,0,0},
215
    {"exec",execCommand,1,"s",0,NULL,0,0,0,0,0},
216 217
    {"discard",discardCommand,1,"rs",0,NULL,0,0,0,0,0},
    {"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
218 219
    {"flushdb",flushdbCommand,1,"w",0,NULL,0,0,0,0,0},
    {"flushall",flushallCommand,1,"w",0,NULL,0,0,0,0,0},
220
    {"sort",sortCommand,-2,"wmS",0,NULL,1,1,1,0,0},
221
    {"info",infoCommand,-1,"r",0,NULL,0,0,0,0,0},
222
    {"monitor",monitorCommand,1,"ars",0,NULL,0,0,0,0,0},
223 224
    {"ttl",ttlCommand,2,"r",0,NULL,1,1,1,0,0},
    {"pttl",pttlCommand,2,"r",0,NULL,1,1,1,0,0},
225
    {"persist",persistCommand,2,"w",0,NULL,1,1,1,0,0},
A
antirez 已提交
226
    {"slaveof",slaveofCommand,3,"as",0,NULL,0,0,0,0,0},
227
    {"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
228
    {"config",configCommand,-2,"ar",0,NULL,0,0,0,0,0},
229 230 231 232
    {"subscribe",subscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
    {"unsubscribe",unsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
    {"psubscribe",psubscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
    {"punsubscribe",punsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
233
    {"publish",publishCommand,3,"pf",0,NULL,0,0,0,0,0},
234 235
    {"watch",watchCommand,-2,"rs",0,noPreloadGetKeys,1,-1,1,0,0},
    {"unwatch",unwatchCommand,1,"rs",0,NULL,0,0,0,0,0},
236
    {"restore",restoreCommand,4,"awm",0,NULL,1,1,1,0,0},
237
    {"migrate",migrateCommand,6,"aw",0,NULL,0,0,0,0,0},
238 239
    {"dump",dumpCommand,2,"ar",0,NULL,1,1,1,0,0},
    {"object",objectCommand,-2,"r",0,NULL,2,2,2,0,0},
240
    {"client",clientCommand,-2,"ar",0,NULL,0,0,0,0,0},
241 242
    {"eval",evalCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
    {"evalsha",evalShaCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
243
    {"slowlog",slowlogCommand,-2,"r",0,NULL,0,0,0,0,0},
A
antirez 已提交
244
    {"script",scriptCommand,-2,"ras",0,NULL,0,0,0,0,0},
A
antirez 已提交
245 246 247
    {"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}
248 249 250 251
};

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

252 253 254
/* 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
255 256
    const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
    const char *c = ".-*#";
257
    FILE *fp;
258
    char buf[64];
A
antirez 已提交
259
    int rawmode = (level & REDIS_LOG_RAW);
260

A
antirez 已提交
261
    level &= 0xff; /* clear flags */
262
    if (level < server.verbosity) return;
263 264 265 266

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

A
antirez 已提交
267 268 269
    if (rawmode) {
        fprintf(fp,"%s",msg);
    } else {
270 271 272 273 274 275
        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
276 277
        fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg);
    }
J
Jonah H. Harris 已提交
278 279
    fflush(fp);

280
    if (server.logfile) fclose(fp);
J
Jonah H. Harris 已提交
281 282

    if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
283 284
}

285 286 287 288 289 290 291
/* Like redisLogRaw() but with printf-alike support. This is the funciton that
 * is used across the code. The raw version is only used in order to dump
 * the INFO output on crash. */
void redisLog(int level, const char *fmt, ...) {
    va_list ap;
    char msg[REDIS_MAX_LOGMSG_LEN];

A
antirez 已提交
292
    if ((level&0xff) < server.verbosity) return;
293 294 295 296 297 298 299 300

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

    redisLogRaw(level,msg);
}

A
antirez 已提交
301 302 303 304 305 306 307 308 309 310 311 312 313 314
/* 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
315
        STDOUT_FILENO;
A
antirez 已提交
316 317
    if (fd == -1) return;
    ll2string(buf,sizeof(buf),getpid());
318 319 320
    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
321
    ll2string(buf,sizeof(buf),time(NULL));
322 323 324 325 326
    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
327
    if (server.logfile) close(fd);
A
antirez 已提交
328 329
}

330 331 332 333 334 335 336 337 338 339 340
/* Redis generally does not try to recover from out of memory conditions
 * when allocating objects or strings, it is not clear if it will be possible
 * to report this condition to the client since the networking layer itself
 * is based on heap allocation for send buffers, so we simply abort.
 * At least the code will be simpler to read... */
void oom(const char *msg) {
    redisLog(REDIS_WARNING, "%s: Out of memory\n",msg);
    sleep(1);
    abort();
}

A
antirez 已提交
341 342 343 344 345 346 347 348 349 350 351
/* 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;
}

352 353 354 355 356
/* Return the UNIX time in milliseconds */
long long mstime(void) {
    return ustime()/1000;
}

357 358 359 360 361 362 363 364 365 366 367 368
/* 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
}

369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398
/*====================== Hash table type implementation  ==================== */

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

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

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

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

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

399 400 401 402 403 404 405 406 407
/* A case insensitive version used for the command lookup table. */
int dictSdsKeyCaseCompare(void *privdata, const void *key1,
        const void *key2)
{
    DICT_NOTUSED(privdata);

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

408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438
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));
}

439 440 441 442
unsigned int dictSdsCaseHash(const void *key) {
    return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
}

443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483
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;
        }
    }
}

484
/* Sets type hash table */
485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500
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 */
501
    NULL                       /* val destructor */
502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
};

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

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

524 525 526 527 528 529 530 531 532 533
/* 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 */
};

534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
/* Hash type hash table (note that small hashes are represented with zimpaps) */
dictType hashDictType = {
    dictEncObjHash,             /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictEncObjKeyCompare,       /* key compare */
    dictRedisObjectDestructor,  /* key destructor */
    dictRedisObjectDestructor   /* val destructor */
};

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

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

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

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

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

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

    for (j = 0; j < server.dbnum; j++) {
586
        /* Keys dictionary */
587 588 589 590
        if (dictIsRehashing(server.db[j].dict)) {
            dictRehashMilliseconds(server.db[j].dict,1);
            break; /* already used our millisecond for this loop... */
        }
591 592 593 594 595
        /* Expires */
        if (dictIsRehashing(server.db[j].expires)) {
            dictRehashMilliseconds(server.db[j].expires,1);
            break; /* already used our millisecond for this loop... */
        }
596 597 598 599 600 601 602 603 604 605
    }
}

/* 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
606
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
607 608 609 610 611 612 613
        dictEnableResize();
    else
        dictDisableResize();
}

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

614 615 616 617 618
/* 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) {
619
    int j, iteration = 0;
620
    long long start = ustime(), timelimit;
621 622 623 624

    /* We can use at max REDIS_EXPIRELOOKUPS_TIME_PERC percentage of CPU time
     * per iteration. Since this function gets called with a frequency of
     * REDIS_HZ times per second, the following is the max amount of
625 626
     * microseconds we can spend in this function. */
    timelimit = 1000000*REDIS_EXPIRELOOKUPS_TIME_PERC/REDIS_HZ/100;
627
    if (timelimit <= 0) timelimit = 1;
628 629

    for (j = 0; j < server.dbnum; j++) {
630
        int expired;
631 632 633 634 635
        redisDb *db = server.db+j;

        /* Continue to expire if at the end of the cycle more than 25%
         * of the keys were expired. */
        do {
636 637
            unsigned long num = dictSize(db->expires);
            unsigned long slots = dictSlots(db->expires);
638
            long long now = mstime();
639

640 641 642 643 644 645
            /* 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;

646 647
            /* The main collection cycle. Sample random keys among keys
             * with an expire set, checking for expired ones. */
648 649 650 651 652
            expired = 0;
            if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
                num = REDIS_EXPIRELOOKUPS_PER_CRON;
            while (num--) {
                dictEntry *de;
653
                long long t;
654 655

                if ((de = dictGetRandomKey(db->expires)) == NULL) break;
656
                t = dictGetSignedIntegerVal(de);
657
                if (now > t) {
658
                    sds key = dictGetKey(de);
659 660 661 662 663 664 665 666 667
                    robj *keyobj = createStringObject(key,sdslen(key));

                    propagateExpire(db,keyobj);
                    dbDelete(db,keyobj);
                    decrRefCount(keyobj);
                    expired++;
                    server.stat_expiredkeys++;
                }
            }
668 669 670 671
            /* 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++;
672
            if ((iteration & 0xf) == 0 && /* check once every 16 cycles. */
673
                (ustime()-start) > timelimit) return;
674 675 676 677
        } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
    }
}

678
void updateLRUClock(void) {
679
    server.lruclock = (server.unixtime/REDIS_LRU_CLOCK_RESOLUTION) &
680 681
                                                REDIS_LRU_CLOCK_MAX;
}
682

683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707

/* 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
708 709
/* Check for timeouts. Returns non-zero if the client was terminated */
int clientsCronHandleTimeout(redisClient *c) {
710
    time_t now = server.unixtime;
711 712 713 714 715 716 717 718 719 720 721

    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
722
        return 1;
723 724 725 726 727 728
    } 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
729
    return 0;
730 731
}

732
/* The client query buffer is an sds.c string that can end with a lot of
A
antirez 已提交
733 734 735 736
 * free space not used, this function reclaims space if needed.
 *
 * The funciton always returns 0 as it never terminates the client. */
int clientsCronResizeQueryBuffer(redisClient *c) {
737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
    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
755
    return 0;
756 757
}

758
void clientsCron(void) {
759 760
    /* Make sure to process at least 1/(REDIS_HZ*10) of clients per call.
     * Since this function is called REDIS_HZ times per second we are sure that
761 762 763
     * 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. */
764
    int numclients = listLength(server.clients);
765
    int iterations = numclients/(REDIS_HZ*10);
766

767 768
    if (iterations < 50)
        iterations = (numclients < 50) ? numclients : 50;
769 770 771 772 773 774 775 776 777 778
    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
779 780 781 782 783
        /* 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;
784 785 786
    }
}

787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810
/* This is our timer interrupt, called REDIS_HZ times per second.
 * 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).
 * - Software watchdong.
 * - Update some statistic.
 * - Incremental rehashing of the DBs hash tables.
 * - Triggering BGSAVE / AOF rewrite, and handling of terminated children.
 * - Clients timeout of differnet kinds.
 * - Replication reconnection.
 * - Many more...
 *
 * Everything directly called here will be called REDIS_HZ times per second,
 * so in order to throttle execution of things we want to do less frequently
 * a macro is used: run_with_period(milliseconds) { .... }
 */

/* Using the following macro you can run code inside serverCron() with the
 * specified period, specified in milliseconds.
 * The actual resolution depends on REDIS_HZ. */
#define run_with_period(_ms_) if (!(loops % ((_ms_)/(1000/REDIS_HZ))))

811
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
812
    int j, loops = server.cronloops;
813 814 815 816
    REDIS_NOTUSED(eventLoop);
    REDIS_NOTUSED(id);
    REDIS_NOTUSED(clientData);

A
antirez 已提交
817 818 819 820
    /* 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);

821 822 823 824 825
    /* 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
826

827
    run_with_period(100) trackOperationsPerSecond();
828

829
    /* We have just 22 bits per object for LRU information.
830 831
     * So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
     * 2^22 bits with 10 seconds resoluton is more or less 1.5 years.
832
     *
833
     * Note that even if this will wrap after 1.5 years it's not a problem,
834
     * everything will still work but just some object will appear younger
835 836 837 838 839
     * 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.
840
     */
841
    updateLRUClock();
842

843 844 845 846
    /* 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();

847 848 849
    /* 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) {
850
        if (prepareForShutdown(0) == REDIS_OK) exit(0);
851 852 853 854
        redisLog(REDIS_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
    }

    /* Show some info about non-empty databases */
855 856 857 858 859 860 861 862 863 864 865
    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); */
            }
866 867 868 869 870 871 872 873 874
        }
    }

    /* We don't want to resize the hash tables while a bacground saving
     * is in progress: the saving child is created using fork() that is
     * implemented with a copy-on-write semantic in most modern systems, so
     * if we resize the HT while there is the saving child at work actually
     * a lot of memory movements in the parent will cause a lot of pages
     * copied. */
A
antirez 已提交
875
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
876
        tryResizeHashTables();
877 878 879 880
        if (server.activerehashing) incrementallyRehash();
    }

    /* Show information about connected clients */
881
    run_with_period(5000) {
882 883 884
        redisLog(REDIS_VERBOSE,"%d clients connected (%d slaves), %zu bytes in use",
            listLength(server.clients)-listLength(server.slaves),
            listLength(server.slaves),
885
            zmalloc_used_memory());
886 887
    }

888 889
    /* We need to do a few operations on clients asynchronously. */
    clientsCron();
890

891 892
    /* Start a scheduled AOF rewrite if this was requested by the user while
     * a BGSAVE was in progress. */
A
antirez 已提交
893
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
894
        server.aof_rewrite_scheduled)
895 896 897 898
    {
        rewriteAppendOnlyFileBackground();
    }

A
antirez 已提交
899
    /* Check if a background saving or AOF rewrite in progress terminated. */
A
antirez 已提交
900
    if (server.rdb_child_pid != -1 || server.aof_child_pid != -1) {
901 902 903 904
        int statloc;
        pid_t pid;

        if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
905 906 907 908 909
            int exitcode = WEXITSTATUS(statloc);
            int bysignal = 0;
            
            if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);

A
antirez 已提交
910
            if (pid == server.rdb_child_pid) {
911
                backgroundSaveDoneHandler(exitcode,bysignal);
912
            } else {
913
                backgroundRewriteDoneHandler(exitcode,bysignal);
914 915 916
            }
            updateDictResizePolicy();
        }
A
antirez 已提交
917
    } else {
918 919
        /* If there is not a background saving/rewrite in progress check if
         * we have to save/rewrite now */
920 921 922 923
         for (j = 0; j < server.saveparamslen; j++) {
            struct saveparam *sp = server.saveparams+j;

            if (server.dirty >= sp->changes &&
924
                server.unixtime-server.lastsave > sp->seconds) {
925 926
                redisLog(REDIS_NOTICE,"%d changes in %d seconds. Saving...",
                    sp->changes, sp->seconds);
A
antirez 已提交
927
                rdbSaveBackground(server.rdb_filename);
928 929 930
                break;
            }
         }
931 932

         /* Trigger an AOF rewrite if needed */
A
antirez 已提交
933
         if (server.rdb_child_pid == -1 &&
A
antirez 已提交
934
             server.aof_child_pid == -1 &&
935 936
             server.aof_rewrite_perc &&
             server.aof_current_size > server.aof_rewrite_min_size)
937
         {
938 939 940 941
            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) {
942
                redisLog(REDIS_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth);
943 944 945
                rewriteAppendOnlyFileBackground();
            }
         }
946 947
    }

948 949 950 951 952

    /* 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);

953 954 955 956
    /* 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();
957

958 959 960
    /* Close clients that need to be closed asynchronous */
    freeClientsInAsyncFreeQueue();

961 962
    /* Replication cron function -- used to reconnect to master and
     * to detect transfer failures. */
963
    run_with_period(1000) replicationCron();
964

965
    server.cronloops++;
966
    return 1000/REDIS_HZ;
967 968 969 970 971 972 973
}

/* 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);
974 975
    listNode *ln;
    redisClient *c;
976

977 978 979 980 981 982
    /* 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);
983
        c->flags &= ~REDIS_UNBLOCKED;
984 985

        /* Process remaining data in the input buffer. */
986 987
        if (c->querybuf && sdslen(c->querybuf) > 0) {
            server.current_client = c;
988
            processInputBuffer(c);
989 990
            server.current_client = NULL;
        }
991 992
    }

993
    /* Write the AOF buffer on disk */
994
    flushAppendOnlyFile(0);
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
}

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

void createSharedObjects(void) {
    int j;

    shared.crlf = createObject(REDIS_STRING,sdsnew("\r\n"));
    shared.ok = createObject(REDIS_STRING,sdsnew("+OK\r\n"));
    shared.err = createObject(REDIS_STRING,sdsnew("-ERR\r\n"));
    shared.emptybulk = createObject(REDIS_STRING,sdsnew("$0\r\n\r\n"));
    shared.czero = createObject(REDIS_STRING,sdsnew(":0\r\n"));
    shared.cone = createObject(REDIS_STRING,sdsnew(":1\r\n"));
    shared.cnegone = createObject(REDIS_STRING,sdsnew(":-1\r\n"));
    shared.nullbulk = createObject(REDIS_STRING,sdsnew("$-1\r\n"));
    shared.nullmultibulk = createObject(REDIS_STRING,sdsnew("*-1\r\n"));
    shared.emptymultibulk = createObject(REDIS_STRING,sdsnew("*0\r\n"));
    shared.pong = createObject(REDIS_STRING,sdsnew("+PONG\r\n"));
    shared.queued = createObject(REDIS_STRING,sdsnew("+QUEUED\r\n"));
    shared.wrongtypeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR Operation against a key holding the wrong kind of value\r\n"));
    shared.nokeyerr = createObject(REDIS_STRING,sdsnew(
        "-ERR no such key\r\n"));
    shared.syntaxerr = createObject(REDIS_STRING,sdsnew(
        "-ERR syntax error\r\n"));
    shared.sameobjecterr = createObject(REDIS_STRING,sdsnew(
        "-ERR source and destination objects are the same\r\n"));
    shared.outofrangeerr = createObject(REDIS_STRING,sdsnew(
        "-ERR index out of range\r\n"));
A
antirez 已提交
1024 1025
    shared.noscripterr = createObject(REDIS_STRING,sdsnew(
        "-NOSCRIPT No matching script. Please use EVAL.\r\n"));
1026 1027
    shared.loadingerr = createObject(REDIS_STRING,sdsnew(
        "-LOADING Redis is loading the dataset in memory\r\n"));
1028
    shared.slowscripterr = createObject(REDIS_STRING,sdsnew(
1029
        "-BUSY Redis is busy running a script. You can only call SCRIPT KILL or SHUTDOWN NOSAVE.\r\n"));
1030 1031
    shared.masterdownerr = createObject(REDIS_STRING,sdsnew(
        "-MASTERDOWN Link with MASTER is down and slave-serve-stale-data is set to 'no'.\r\n"));
1032
    shared.bgsaveerr = createObject(REDIS_STRING,sdsnew(
1033 1034 1035
        "-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"));
1036 1037
    shared.oomerr = createObject(REDIS_STRING,sdsnew(
        "-OOM command not allowed when used memory > 'maxmemory'.\r\n"));
1038 1039 1040
    shared.space = createObject(REDIS_STRING,sdsnew(" "));
    shared.colon = createObject(REDIS_STRING,sdsnew(":"));
    shared.plus = createObject(REDIS_STRING,sdsnew("+"));
1041 1042 1043 1044 1045

    for (j = 0; j < REDIS_SHARED_SELECT_CMDS; j++) {
        shared.select[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"select %d\r\n", j));
    }
1046 1047 1048 1049 1050 1051
    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("$12\r\npunsubscribe\r\n",19);
1052
    shared.del = createStringObject("DEL",3);
1053 1054
    shared.rpop = createStringObject("RPOP",4);
    shared.lpop = createStringObject("LPOP",4);
1055 1056 1057 1058
    for (j = 0; j < REDIS_SHARED_INTEGERS; j++) {
        shared.integers[j] = createObject(REDIS_STRING,(void*)(long)j);
        shared.integers[j]->encoding = REDIS_ENCODING_INT;
    }
1059 1060 1061 1062 1063 1064
    for (j = 0; j < REDIS_SHARED_BULKHDR_LEN; j++) {
        shared.mbulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"*%d\r\n",j));
        shared.bulkhdr[j] = createObject(REDIS_STRING,
            sdscatprintf(sdsempty(),"$%d\r\n",j));
    }
1065 1066 1067
}

void initServerConfig() {
A
antirez 已提交
1068 1069
    getRandomHexChars(server.runid,REDIS_RUN_ID_SIZE);
    server.runid[REDIS_RUN_ID_SIZE] = '\0';
1070
    server.arch_bits = (sizeof(long) == 8) ? 64 : 32;
1071
    server.port = REDIS_SERVERPORT;
1072
    server.bindaddr = NULL;
1073
    server.unixsocket = NULL;
1074
    server.unixsocketperm = 0;
1075 1076 1077
    server.ipfd = -1;
    server.sofd = -1;
    server.dbnum = REDIS_DEFAULT_DBNUM;
A
antirez 已提交
1078
    server.verbosity = REDIS_NOTICE;
1079
    server.maxidletime = REDIS_MAXIDLETIME;
1080
    server.client_max_querybuf_len = REDIS_MAX_QUERYBUF_LEN;
1081
    server.saveparams = NULL;
1082
    server.loading = 0;
1083
    server.logfile = NULL; /* NULL = log on standard output */
J
Jonah H. Harris 已提交
1084 1085 1086
    server.syslog_enabled = 0;
    server.syslog_ident = zstrdup("redis");
    server.syslog_facility = LOG_LOCAL0;
1087
    server.daemonize = 0;
1088
    server.aof_state = REDIS_AOF_OFF;
1089 1090 1091 1092 1093 1094
    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
1095
    server.aof_last_fsync = time(NULL);
1096
    server.aof_delayed_fsync = 0;
A
antirez 已提交
1097 1098
    server.aof_fd = -1;
    server.aof_selected_db = -1; /* Make sure the first time will not match */
1099
    server.aof_flush_postponed_start = 0;
1100
    server.pidfile = zstrdup("/var/run/redis.pid");
A
antirez 已提交
1101
    server.rdb_filename = zstrdup("dump.rdb");
1102
    server.aof_filename = zstrdup("appendonly.aof");
1103
    server.requirepass = NULL;
A
antirez 已提交
1104
    server.rdb_compression = 1;
1105
    server.rdb_checksum = 1;
1106
    server.activerehashing = 1;
1107
    server.maxclients = REDIS_MAX_CLIENTS;
1108
    server.bpop_blocked_clients = 0;
1109
    server.maxmemory = 0;
1110 1111
    server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU;
    server.maxmemory_samples = 3;
1112 1113
    server.hash_max_ziplist_entries = REDIS_HASH_MAX_ZIPLIST_ENTRIES;
    server.hash_max_ziplist_value = REDIS_HASH_MAX_ZIPLIST_VALUE;
1114 1115
    server.list_max_ziplist_entries = REDIS_LIST_MAX_ZIPLIST_ENTRIES;
    server.list_max_ziplist_value = REDIS_LIST_MAX_ZIPLIST_VALUE;
1116
    server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES;
1117 1118
    server.zset_max_ziplist_entries = REDIS_ZSET_MAX_ZIPLIST_ENTRIES;
    server.zset_max_ziplist_value = REDIS_ZSET_MAX_ZIPLIST_VALUE;
1119
    server.shutdown_asap = 0;
A
1fe4cd5  
antirez 已提交
1120 1121
    server.repl_ping_slave_period = REDIS_REPL_PING_SLAVE_PERIOD;
    server.repl_timeout = REDIS_REPL_TIMEOUT;
1122
    server.lua_caller = NULL;
A
antirez 已提交
1123
    server.lua_time_limit = REDIS_LUA_TIME_LIMIT;
1124
    server.lua_client = NULL;
1125
    server.lua_timedout = 0;
1126

A
antirez 已提交
1127
    updateLRUClock();
1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
    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;
A
antirez 已提交
1138
    server.repl_state = REDIS_REPL_NONE;
1139
    server.repl_syncio_timeout = REDIS_REPL_SYNCIO_TIMEOUT;
1140
    server.repl_serve_stale_data = 1;
1141
    server.repl_slave_ro = 1;
1142
    server.repl_down_since = time(NULL);
1143

1144 1145 1146 1147
    /* 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;
1148 1149
    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;
1150
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_seconds = 60;
1151 1152
    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;
1153 1154
    server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_seconds = 60;

1155 1156 1157 1158 1159
    /* 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;
1160 1161 1162 1163 1164 1165 1166 1167

    /* Command table -- we intiialize it here as it is part of the
     * initial configuration, since command names may be changed via
     * redis.conf using the rename-command directive. */
    server.commands = dictCreate(&commandTableDictType,NULL);
    populateCommandTable();
    server.delCommand = lookupCommandByCString("del");
    server.multiCommand = lookupCommandByCString("multi");
1168
    server.lpushCommand = lookupCommandByCString("lpush");
1169 1170 1171 1172
    
    /* 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
1173

A
antirez 已提交
1174
    /* Debugging */
A
antirez 已提交
1175 1176 1177 1178
    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
1179
    server.watchdog_period = 0;
1180 1181
}

1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203
/* 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) {
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215
            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;
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225
                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);
            }
        }
    }
}

1226 1227 1228 1229 1230
void initServer() {
    int j;

    signal(SIGHUP, SIG_IGN);
    signal(SIGPIPE, SIG_IGN);
1231
    setupSignalHandlers();
1232

J
Jonah H. Harris 已提交
1233 1234 1235 1236 1237
    if (server.syslog_enabled) {
        openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
            server.syslog_facility);
    }

1238
    server.current_client = NULL;
1239
    server.clients = listCreate();
1240
    server.clients_to_close = listCreate();
1241 1242
    server.slaves = listCreate();
    server.monitors = listCreate();
1243
    server.unblocked_clients = listCreate();
1244

1245
    createSharedObjects();
1246 1247
    adjustOpenFilesLimit();
    server.el = aeCreateEventLoop(server.maxclients+1024);
1248
    server.db = zmalloc(sizeof(redisDb)*server.dbnum);
1249

1250
    if (server.port != 0) {
1251
        server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr);
1252
        if (server.ipfd == ANET_ERR) {
1253 1254
            redisLog(REDIS_WARNING, "Opening port %d: %s",
                server.port, server.neterr);
1255 1256
            exit(1);
        }
1257
    }
1258 1259
    if (server.unixsocket != NULL) {
        unlink(server.unixsocket); /* don't care if this fails */
1260
        server.sofd = anetUnixServer(server.neterr,server.unixsocket,server.unixsocketperm);
1261 1262 1263 1264
        if (server.sofd == ANET_ERR) {
            redisLog(REDIS_WARNING, "Opening socket: %s", server.neterr);
            exit(1);
        }
1265
    }
1266 1267
    if (server.ipfd < 0 && server.sofd < 0) {
        redisLog(REDIS_WARNING, "Configured to not listen anywhere, exiting.");
1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
        exit(1);
    }
    for (j = 0; j < server.dbnum; j++) {
        server.db[j].dict = dictCreate(&dbDictType,NULL);
        server.db[j].expires = dictCreate(&keyptrDictType,NULL);
        server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL);
        server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
        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
1282
    server.rdb_child_pid = -1;
A
antirez 已提交
1283
    server.aof_child_pid = -1;
1284
    aofRewriteBufferReset();
A
antirez 已提交
1285
    server.aof_buf = sdsempty();
1286 1287 1288 1289 1290
    server.lastsave = time(NULL);
    server.dirty = 0;
    server.stat_numcommands = 0;
    server.stat_numconnections = 0;
    server.stat_expiredkeys = 0;
1291
    server.stat_evictedkeys = 0;
1292
    server.stat_starttime = time(NULL);
1293 1294
    server.stat_keyspace_misses = 0;
    server.stat_keyspace_hits = 0;
1295
    server.stat_peak_memory = 0;
1296
    server.stat_fork_time = 0;
1297
    server.stat_rejected_conn = 0;
1298 1299 1300 1301
    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;
1302
    server.unixtime = time(NULL);
1303
    server.lastbgsave_status = REDIS_OK;
1304
    server.stop_writes_on_bgsave_err = 1;
1305
    aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL);
1306
    if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE,
1307
        acceptTcpHandler,NULL) == AE_ERR) oom("creating file event");
1308
    if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
1309
        acceptUnixHandler,NULL) == AE_ERR) oom("creating file event");
1310

1311
    if (server.aof_state == REDIS_AOF_ON) {
A
antirez 已提交
1312
        server.aof_fd = open(server.aof_filename,
1313
                               O_WRONLY|O_APPEND|O_CREAT,0644);
A
antirez 已提交
1314
        if (server.aof_fd == -1) {
1315 1316 1317 1318 1319 1320
            redisLog(REDIS_WARNING, "Can't open the append-only file: %s",
                strerror(errno));
            exit(1);
        }
    }

1321 1322 1323 1324 1325 1326 1327 1328 1329 1330
    /* 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
     * at 3.5GB using maxmemory with 'noeviction' policy'. This saves
     * useless crashes of the Redis instance. */
    if (server.arch_bits == 32 && server.maxmemory == 0) {
        redisLog(REDIS_WARNING,"Warning: 32 bit instance detected but no memory limit set. Setting 3.5 GB maxmemory limit with 'noeviction' policy now.");
        server.maxmemory = 3584LL*(1024*1024); /* 3584 MB = 3.5 GB */
        server.maxmemory_policy = REDIS_MAXMEMORY_NO_EVICTION;
    }

1331
    scriptingInit();
1332
    slowlogInit();
1333
    bioInit();
1334 1335
}

1336 1337 1338 1339
/* Populates the Redis Command Table starting from the hard coded list
 * we have on top of redis.c file. */
void populateCommandTable(void) {
    int j;
1340
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1341 1342

    for (j = 0; j < numcommands; j++) {
1343
        struct redisCommand *c = redisCommandTable+j;
1344
        char *f = c->sflags;
1345
        int retval;
1346

1347 1348 1349 1350 1351 1352 1353 1354
        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;
1355 1356
            case 's': c->flags |= REDIS_CMD_NOSCRIPT; break;
            case 'R': c->flags |= REDIS_CMD_RANDOM; break;
1357
            case 'S': c->flags |= REDIS_CMD_SORT_FOR_SCRIPT; break;
1358 1359 1360 1361 1362
            default: redisPanic("Unsupported command flag"); break;
            }
            f++;
        }

1363 1364 1365
        retval = dictAdd(server.commands, sdsnew(c->name), c);
        assert(retval == DICT_OK);
    }
1366 1367
}

1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
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;
    }
}

1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416
/* ========================== 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);
}

1417 1418
/* ====================== Commands lookup and execution ===================== */

1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429
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;
1430 1431
}

A
antirez 已提交
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
/* Propagate the specified command (in the context of the specified database id)
 * to AOF, Slaves and Monitors.
 *
 * 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);
    if (flags & REDIS_PROPAGATE_REPL && listLength(server.slaves))
        replicationFeedSlaves(server.slaves,dbid,argv,argc);
}

1449 1450
/* Used inside commands to schedule the propagation of additional commands
 * after the current command is propagated to AOF / Replication. */
1451 1452 1453
void alsoPropagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
                   int target)
{
1454
    redisOpArrayAppend(&server.also_propagate,cmd,dbid,argv,argc,target);
1455 1456
}

1457
/* Call() is the core of Redis execution of a command */
1458
void call(redisClient *c, int flags) {
1459
    long long dirty, start = ustime(), duration;
1460

A
antirez 已提交
1461 1462 1463
    /* Sent the command to clients in MONITOR mode, only if the commands are
     * not geneated from reading an AOF. */
    if (listLength(server.monitors) && !server.loading)
1464
        replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
A
antirez 已提交
1465 1466

    /* Call the command. */
1467
    redisOpArrayInit(&server.also_propagate);
1468
    dirty = server.dirty;
1469
    c->cmd->proc(c);
1470
    dirty = server.dirty-dirty;
1471
    duration = ustime()-start;
1472 1473 1474 1475 1476 1477

    /* 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
1478 1479
    /* Log the command into the Slow log if needed, and populate the
     * per-command statistics that we show in INFO commandstats. */
1480 1481 1482 1483 1484 1485
    if (flags & REDIS_CALL_SLOWLOG)
        slowlogPushEntryIfNeeded(c->argv,c->argc,duration);
    if (flags & REDIS_CALL_STATS) {
        c->cmd->microseconds += duration;
        c->cmd->calls++;
    }
A
antirez 已提交
1486 1487

    /* Propagate the command into the AOF and replication link */
1488
    if (flags & REDIS_CALL_PROPAGATE) {
A
antirez 已提交
1489 1490 1491 1492 1493 1494 1495 1496
        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);
1497
    }
1498 1499
    /* Commands such as LPUSH or BRPOPLPUSH may propagate an additional
     * PUSH command. */
1500
    if (server.also_propagate.numops) {
1501
        int j;
1502
        redisOp *rop;
1503

1504 1505 1506 1507 1508
        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);
1509
    }
1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
    server.stat_numcommands++;
}

/* If this function gets called we already read a whole
 * command, argments are in the client argv/argc fields.
 * processCommand() execute the command or prepare the
 * server for a bulk read from the client.
 *
 * If 1 is returned the client is still alive and valid and
 * and other operations can be performed by the caller. Otherwise
 * if 0 is returned the client was destroied (i.e. after QUIT). */
int processCommand(redisClient *c) {
P
Pieter Noordhuis 已提交
1522 1523 1524 1525
    /* 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. */
1526
    if (!strcasecmp(c->argv[0]->ptr,"quit")) {
P
Pieter Noordhuis 已提交
1527
        addReply(c,shared.ok);
1528
        c->flags |= REDIS_CLOSE_AFTER_REPLY;
1529
        return REDIS_ERR;
1530 1531 1532
    }

    /* Now lookup the command and check ASAP about trivial error conditions
1533
     * such as wrong arity, bad command name and so forth. */
1534
    c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
1535
    if (!c->cmd) {
1536 1537
        addReplyErrorFormat(c,"unknown command '%s'",
            (char*)c->argv[0]->ptr);
1538
        return REDIS_OK;
1539 1540
    } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) ||
               (c->argc < -c->cmd->arity)) {
1541
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
1542
            c->cmd->name);
1543
        return REDIS_OK;
1544 1545 1546
    }

    /* Check if the user is authenticated */
1547 1548
    if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand)
    {
1549
        addReplyError(c,"operation not permitted");
1550
        return REDIS_OK;
1551 1552
    }

1553 1554 1555 1556 1557
    /* 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. */
1558 1559 1560
    if (server.maxmemory) {
        int retval = freeMemoryIfNeeded();
        if ((c->cmd->flags & REDIS_CMD_DENYOOM) && retval == REDIS_ERR) {
1561
            addReply(c, shared.oomerr);
1562 1563
            return REDIS_OK;
        }
1564 1565
    }

1566
    /* Don't accept write commands if there are problems persisting on disk. */
1567 1568 1569
    if (server.stop_writes_on_bgsave_err &&
        server.saveparamslen > 0
        && server.lastbgsave_status == REDIS_ERR &&
1570 1571 1572 1573
        c->cmd->flags & REDIS_CMD_WRITE)
    {
        addReply(c, shared.bgsaveerr);
        return REDIS_OK;
1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
    }

    /* Don't accept wirte commands if this is a read only slave. But
     * 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;
1584 1585
    }

1586 1587 1588
    /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */
    if ((dictSize(c->pubsub_channels) > 0 || listLength(c->pubsub_patterns) > 0)
        &&
1589 1590 1591 1592
        c->cmd->proc != subscribeCommand &&
        c->cmd->proc != unsubscribeCommand &&
        c->cmd->proc != psubscribeCommand &&
        c->cmd->proc != punsubscribeCommand) {
1593
        addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context");
1594
        return REDIS_OK;
1595 1596
    }

1597 1598
    /* 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
1599
    if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED &&
1600
        server.repl_serve_stale_data == 0 &&
1601
        c->cmd->proc != infoCommand && c->cmd->proc != slaveofCommand)
1602
    {
1603
        addReply(c, shared.masterdownerr);
1604 1605 1606
        return REDIS_OK;
    }

1607
    /* Loading DB? Return an error if the command is not INFO */
1608
    if (server.loading && c->cmd->proc != infoCommand) {
1609 1610 1611 1612
        addReply(c, shared.loadingerr);
        return REDIS_OK;
    }

1613 1614
    /* Lua script too slow? Only allow SHUTDOWN NOSAVE and SCRIPT KILL. */
    if (server.lua_timedout &&
1615
        !(c->cmd->proc == shutdownCommand &&
1616 1617 1618 1619 1620 1621
          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'))
    {
1622 1623 1624 1625
        addReply(c, shared.slowscripterr);
        return REDIS_OK;
    }

1626 1627
    /* Exec the command */
    if (c->flags & REDIS_MULTI &&
1628 1629
        c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
        c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
1630
    {
1631
        queueMultiCommand(c);
1632 1633
        addReply(c,shared.queued);
    } else {
1634
        call(c,REDIS_CALL_FULL);
1635
    }
1636
    return REDIS_OK;
1637 1638 1639 1640
}

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

1641 1642 1643 1644
int prepareForShutdown(int flags) {
    int save = flags & REDIS_SHUTDOWN_SAVE;
    int nosave = flags & REDIS_SHUTDOWN_NOSAVE;

1645
    redisLog(REDIS_WARNING,"User requested shutdown...");
1646 1647 1648
    /* 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
1649
    if (server.rdb_child_pid != -1) {
1650
        redisLog(REDIS_WARNING,"There is a child saving an .rdb. Killing it!");
A
antirez 已提交
1651 1652
        kill(server.rdb_child_pid,SIGKILL);
        rdbRemoveTempFile(server.rdb_child_pid);
1653
    }
1654
    if (server.aof_state != REDIS_AOF_OFF) {
1655 1656
        /* 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
1657
        if (server.aof_child_pid != -1) {
1658 1659
            redisLog(REDIS_WARNING,
                "There is a child rewriting the AOF. Killing it!");
A
antirez 已提交
1660
            kill(server.aof_child_pid,SIGKILL);
1661
        }
1662
        /* Append only file: fsync() the AOF and exit */
1663
        redisLog(REDIS_NOTICE,"Calling fsync() on the AOF file.");
A
antirez 已提交
1664
        aof_fsync(server.aof_fd);
1665
    }
1666
    if ((server.saveparamslen > 0 && !nosave) || save) {
1667
        redisLog(REDIS_NOTICE,"Saving the final RDB snapshot before exiting.");
1668
        /* Snapshotting. Perform a SYNC SAVE and exit */
A
antirez 已提交
1669
        if (rdbSave(server.rdb_filename) != REDIS_OK) {
1670 1671 1672 1673 1674
            /* 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... */
1675
            redisLog(REDIS_WARNING,"Error trying to save the DB, can't exit.");
1676 1677 1678
            return REDIS_ERR;
        }
    }
1679 1680 1681 1682
    if (server.daemonize) {
        redisLog(REDIS_NOTICE,"Removing the pid file.");
        unlink(server.pidfile);
    }
1683 1684 1685
    /* Close the listening sockets. Apparently this allows faster restarts. */
    if (server.ipfd != -1) close(server.ipfd);
    if (server.sofd != -1) close(server.sofd);
1686 1687 1688 1689
    if (server.unixsocket) {
        redisLog(REDIS_NOTICE,"Removing the unix socket file.");
        unlink(server.unixsocket); /* don't care if this fails */
    }
1690

1691
    redisLog(REDIS_WARNING,"Redis is now ready to exit, bye bye...");
1692 1693 1694 1695 1696 1697
    return REDIS_OK;
}

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

void authCommand(redisClient *c) {
1698 1699 1700
    if (!server.requirepass) {
        addReplyError(c,"Client sent AUTH, but no password is set");
    } else if (!strcmp(c->argv[1]->ptr, server.requirepass)) {
1701 1702 1703 1704
      c->authenticated = 1;
      addReply(c,shared.ok);
    } else {
      c->authenticated = 0;
1705
      addReplyError(c,"invalid password");
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716
    }
}

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

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

A
antirez 已提交
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727
void timeCommand(redisClient *c) {
    struct timeval tv;

    /* gettimeofday() can only fail if &tv is a bad addresss so we
     * don't check for errors. */
    gettimeofday(&tv,NULL);
    addReplyMultiBulkLen(c,2);
    addReplyBulkLongLong(c,tv.tv_sec);
    addReplyBulkLongLong(c,tv.tv_usec);
}

1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751
/* 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. */
1752 1753
sds genRedisInfoString(char *section) {
    sds info = sdsempty();
1754
    time_t uptime = server.unixtime-server.stat_starttime;
A
antirez 已提交
1755
    int j, numcommands;
1756
    struct rusage self_ru, c_ru;
1757
    unsigned long lol, bib;
1758 1759 1760 1761 1762
    int allsections = 0, defsections = 0;
    int sections = 0;
    
    if (section) {
        allsections = strcasecmp(section,"all") == 0;
1763
        defsections = strcasecmp(section,"default") == 0;
1764
    }
1765 1766 1767

    getrusage(RUSAGE_SELF, &self_ru);
    getrusage(RUSAGE_CHILDREN, &c_ru);
1768
    getClientsMaxBuffers(&lol,&bib);
1769 1770 1771

    /* Server */
    if (allsections || defsections || !strcasecmp(section,"server")) {
1772 1773
        struct utsname name;

1774
        if (sections++) info = sdscat(info,"\r\n");
1775
        uname(&name);
1776
        info = sdscatprintf(info,
1777 1778 1779 1780
            "# Server\r\n"
            "redis_version:%s\r\n"
            "redis_git_sha1:%s\r\n"
            "redis_git_dirty:%d\r\n"
1781
            "os:%s %s %s\r\n"
1782
            "arch_bits:%d\r\n"
1783
            "multiplexing_api:%s\r\n"
A
antirez 已提交
1784
            "gcc_version:%d.%d.%d\r\n"
1785
            "process_id:%ld\r\n"
A
antirez 已提交
1786
            "run_id:%s\r\n"
1787 1788 1789 1790 1791 1792 1793
            "tcp_port:%d\r\n"
            "uptime_in_seconds:%ld\r\n"
            "uptime_in_days:%ld\r\n"
            "lru_clock:%ld\r\n",
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
1794
            name.sysname, name.release, name.machine,
1795
            server.arch_bits,
1796
            aeGetApiName(),
A
antirez 已提交
1797 1798 1799 1800 1801
#ifdef __GNUC__
            __GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__,
#else
            0,0,0,
#endif
1802
            (long) getpid(),
A
antirez 已提交
1803
            server.runid,
1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814
            server.port,
            uptime,
            uptime/(3600*24),
            (unsigned long) server.lruclock);
    }

    /* Clients */
    if (allsections || defsections || !strcasecmp(section,"clients")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Clients\r\n"
1815
            "connected_clients:%lu\r\n"
1816 1817 1818 1819 1820 1821 1822 1823 1824 1825
            "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")) {
1826 1827 1828 1829 1830
        char hmem[64];
        char peak_hmem[64];

        bytesToHuman(hmem,zmalloc_used_memory());
        bytesToHuman(peak_hmem,server.stat_peak_memory);
1831 1832 1833 1834 1835 1836
        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"
1837 1838
            "used_memory_peak:%zu\r\n"
            "used_memory_peak_human:%s\r\n"
A
antirez 已提交
1839
            "used_memory_lua:%lld\r\n"
1840
            "mem_fragmentation_ratio:%.2f\r\n"
1841
            "mem_allocator:%s\r\n",
1842 1843 1844
            zmalloc_used_memory(),
            hmem,
            zmalloc_get_rss(),
1845 1846
            server.stat_peak_memory,
            peak_hmem,
A
antirez 已提交
1847
            ((long long)lua_gc(server.lua,LUA_GCCOUNT,0))*1024LL,
1848
            zmalloc_get_fragmentation_ratio(),
1849
            ZMALLOC_LIB
1850
            );
1851 1852
    }

1853 1854 1855
    /* Persistence */
    if (allsections || defsections || !strcasecmp(section,"persistence")) {
        if (sections++) info = sdscat(info,"\r\n");
1856
        info = sdscatprintf(info,
1857 1858 1859 1860 1861 1862
            "# Persistence\r\n"
            "loading:%d\r\n"
            "aof_enabled:%d\r\n"
            "changes_since_last_save:%lld\r\n"
            "bgsave_in_progress:%d\r\n"
            "last_save_time:%ld\r\n"
1863
            "last_bgsave_status:%s\r\n"
1864 1865
            "bgrewriteaof_in_progress:%d\r\n"
            "bgrewriteaof_scheduled:%d\r\n",
1866
            server.loading,
1867
            server.aof_state != REDIS_AOF_OFF,
1868
            server.dirty,
A
antirez 已提交
1869
            server.rdb_child_pid != -1,
1870
            server.lastsave,
1871
            server.lastbgsave_status == REDIS_OK ? "ok" : "err",
1872 1873
            server.aof_child_pid != -1,
            server.aof_rewrite_scheduled);
1874

1875
        if (server.aof_state != REDIS_AOF_OFF) {
1876 1877 1878
            info = sdscatprintf(info,
                "aof_current_size:%lld\r\n"
                "aof_base_size:%lld\r\n"
1879 1880
                "aof_pending_rewrite:%d\r\n"
                "aof_buffer_length:%zu\r\n"
1881
                "aof_rewrite_buffer_length:%zu\r\n"
1882 1883
                "aof_pending_bio_fsync:%llu\r\n"
                "aof_delayed_fsync:%lu\r\n",
1884 1885 1886
                (long long) server.aof_current_size,
                (long long) server.aof_rewrite_base_size,
                server.aof_rewrite_scheduled,
A
antirez 已提交
1887
                sdslen(server.aof_buf),
1888
                aofRewriteBufferSize(),
1889 1890
                bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC),
                server.aof_delayed_fsync);
1891 1892
        }

1893 1894 1895 1896 1897 1898 1899 1900 1901
        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;

1902
            elapsed = server.unixtime-server.loading_start_time;
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922
            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
            );
        }
1923
    }
1924 1925 1926 1927

    /* Stats */
    if (allsections || defsections || !strcasecmp(section,"stats")) {
        if (sections++) info = sdscat(info,"\r\n");
1928
        info = sdscatprintf(info,
1929 1930 1931
            "# Stats\r\n"
            "total_connections_received:%lld\r\n"
            "total_commands_processed:%lld\r\n"
1932
            "instantaneous_ops_per_sec:%lld\r\n"
1933
            "rejected_connections:%lld\r\n"
1934 1935 1936 1937 1938
            "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"
1939
            "pubsub_patterns:%lu\r\n"
1940
            "latest_fork_usec:%lld\r\n",
1941 1942
            server.stat_numconnections,
            server.stat_numcommands,
1943
            getOperationsPerSecond(),
1944
            server.stat_rejected_conn,
1945 1946 1947 1948 1949
            server.stat_expiredkeys,
            server.stat_evictedkeys,
            server.stat_keyspace_hits,
            server.stat_keyspace_misses,
            dictSize(server.pubsub_channels),
1950 1951
            listLength(server.pubsub_patterns),
            server.stat_fork_time);
1952
    }
A
antirez 已提交
1953

1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969
    /* Replication */
    if (allsections || defsections || !strcasecmp(section,"replication")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Replication\r\n"
            "role:%s\r\n",
            server.masterhost == NULL ? "master" : "slave");
        if (server.masterhost) {
            info = sdscatprintf(info,
                "master_host:%s\r\n"
                "master_port:%d\r\n"
                "master_link_status:%s\r\n"
                "master_last_io_seconds_ago:%d\r\n"
                "master_sync_in_progress:%d\r\n"
                ,server.masterhost,
                server.masterport,
A
antirez 已提交
1970
                (server.repl_state == REDIS_REPL_CONNECTED) ?
1971 1972
                    "up" : "down",
                server.master ?
1973
                ((int)(server.unixtime-server.master->lastinteraction)) : -1,
A
antirez 已提交
1974
                server.repl_state == REDIS_REPL_TRANSFER
1975 1976
            );

A
antirez 已提交
1977
            if (server.repl_state == REDIS_REPL_TRANSFER) {
1978 1979 1980 1981
                info = sdscatprintf(info,
                    "master_sync_left_bytes:%ld\r\n"
                    "master_sync_last_io_seconds_ago:%d\r\n"
                    ,(long)server.repl_transfer_left,
1982
                    (int)(server.unixtime-server.repl_transfer_lastio)
1983 1984
                );
            }
1985

A
antirez 已提交
1986
            if (server.repl_state != REDIS_REPL_CONNECTED) {
1987 1988
                info = sdscatprintf(info,
                    "master_link_down_since_seconds:%ld\r\n",
1989
                    (long)server.unixtime-server.repl_down_since);
1990
            }
A
antirez 已提交
1991
        }
1992
        info = sdscatprintf(info,
1993
            "connected_slaves:%lu\r\n",
1994
            listLength(server.slaves));
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
        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",
                    slaveid,ip,port,state);
                slaveid++;
            }
        }
A
antirez 已提交
2026 2027
    }

2028 2029
    /* CPU */
    if (allsections || defsections || !strcasecmp(section,"cpu")) {
2030 2031
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
2032
        "# CPU\r\n"
2033 2034
        "used_cpu_sys:%.2f\r\n"
        "used_cpu_user:%.2f\r\n"
2035 2036
        "used_cpu_sys_children:%.2f\r\n"
        "used_cpu_user_children:%.2f\r\n",
2037
        (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000,
2038 2039 2040
        (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);
2041
    }
2042

2043 2044 2045 2046
    /* cmdtime */
    if (allsections || !strcasecmp(section,"commandstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Commandstats\r\n");
2047
        numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
2048
        for (j = 0; j < numcommands; j++) {
2049
            struct redisCommand *c = redisCommandTable+j;
2050

2051 2052 2053 2054 2055
            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));
2056
        }
A
antirez 已提交
2057 2058
    }

2059 2060 2061 2062 2063 2064
    /* 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;
2065

2066 2067 2068 2069 2070 2071
            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);
            }
2072 2073 2074 2075 2076 2077
        }
    }
    return info;
}

void infoCommand(redisClient *c) {
2078 2079 2080 2081 2082 2083 2084
    char *section = c->argc == 2 ? c->argv[1]->ptr : "default";

    if (c->argc > 2) {
        addReply(c,shared.syntaxerr);
        return;
    }
    sds info = genRedisInfoString(section);
2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103
    addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
        (unsigned long)sdslen(info)));
    addReplySds(c,info);
    addReply(c,shared.crlf);
}

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

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

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

/* This function gets called when 'maxmemory' is set on the config file to limit
2104
 * the max memory used by the server, before processing a command.
2105
 *
2106 2107
 * The goal of the function is to free enough memory to keep Redis under the
 * configured memory limit.
2108
 *
2109 2110 2111 2112 2113 2114 2115 2116
 * 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.
2117
 */
2118 2119 2120 2121
int freeMemoryIfNeeded(void) {
    size_t mem_used, mem_tofree, mem_freed;
    int slaves = listLength(server.slaves);

2122 2123
    /* Remove the size of slaves output buffers and AOF buffer from the
     * count of used memory. */
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138
    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;
        }
    }
2139 2140
    if (server.aof_state != REDIS_AOF_OFF) {
        mem_used -= sdslen(server.aof_buf);
2141
        mem_used -= aofRewriteBufferSize();
2142
    }
2143

2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154
    /* 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;
2155

2156
        for (j = 0; j < server.dbnum; j++) {
2157
            long bestval = 0; /* just to prevent warning */
2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176
            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);
2177
                bestkey = dictGetKey(de);
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189
            }

            /* 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);
2190
                    thiskey = dictGetKey(de);
2191 2192 2193 2194
                    /* When policy is volatile-lru we need an additonal lookup
                     * to locate the real key, as dict is set to db->expires. */
                    if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
                        de = dictFind(db->dict, thiskey);
2195
                    o = dictGetVal(de);
2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212
                    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);
2213 2214
                    thiskey = dictGetKey(de);
                    thisval = (long) dictGetVal(de);
2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226

                    /* 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) {
2227 2228
                long long delta;

2229
                robj *keyobj = createStringObject(bestkey,sdslen(bestkey));
2230
                propagateExpire(db,keyobj);
2231 2232 2233 2234 2235 2236 2237 2238 2239
                /* 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();
2240
                dbDelete(db,keyobj);
2241 2242
                delta -= (long long) zmalloc_used_memory();
                mem_freed += delta;
2243
                server.stat_evictedkeys++;
2244
                decrRefCount(keyobj);
2245 2246 2247 2248 2249 2250
                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
2251
                if (slaves) flushSlavesOutputBuffers();
2252 2253
            }
        }
2254
        if (!keys_freed) return REDIS_ERR; /* nothing to free... */
2255
    }
2256
    return REDIS_OK;
2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282
}

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

2283 2284 2285 2286
void createPidFile(void) {
    /* Try to write the pid file in a best-effort way. */
    FILE *fp = fopen(server.pidfile,"w");
    if (fp) {
2287
        fprintf(fp,"%d\n",(int)getpid());
2288 2289 2290 2291
        fclose(fp);
    }
}

2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309
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() {
A
antirez 已提交
2310 2311 2312 2313 2314 2315
    printf("Redis server v=%s sha=%s:%d malloc=%s bits=%d\n",
        REDIS_VERSION,
        redisGitSHA1(),
        atoi(redisGitDirty()) > 0,
        ZMALLOC_LIB,
        sizeof(long) == 4 ? 32 : 64);
2316 2317 2318 2319
    exit(0);
}

void usage() {
2320
    fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf] [options]\n");
2321
    fprintf(stderr,"       ./redis-server - (read config from stdin)\n");
2322
    fprintf(stderr,"       ./redis-server -v or --version\n");
2323 2324
    fprintf(stderr,"       ./redis-server -h or --help\n");
    fprintf(stderr,"       ./redis-server --test-memory <megabytes>\n\n");
2325 2326 2327 2328 2329 2330
    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");
    fprintf(stderr,"       ./redis-server /etc/myredis.conf --loglevel verbose\n");
2331 2332 2333
    exit(1);
}

A
antirez 已提交
2334 2335 2336 2337 2338 2339 2340 2341 2342
void redisAsciiArt(void) {
#include "asciilogo.h"
    char *buf = zmalloc(1024*16);

    snprintf(buf,1024*16,ascii_logo,
        REDIS_VERSION,
        redisGitSHA1(),
        strtol(redisGitDirty(),NULL,10) > 0,
        (sizeof(long) == 8) ? "64" : "32",
2343
        "stand alone",
A
antirez 已提交
2344 2345 2346 2347 2348 2349 2350
        server.port,
        (long) getpid()
    );
    redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf);
    zfree(buf);
}

2351
static void sigtermHandler(int sig) {
2352 2353
    REDIS_NOTUSED(sig);

A
antirez 已提交
2354
    redisLogFromHandler(REDIS_WARNING,"Received SIGTERM, scheduling shutdown...");
2355 2356 2357
    server.shutdown_asap = 1;
}

2358
void setupSignalHandlers(void) {
2359 2360
    struct sigaction act;

2361 2362 2363
    /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used.
     * Otherwise, sa_handler is used. */
    sigemptyset(&act.sa_mask);
2364
    act.sa_flags = 0;
2365
    act.sa_handler = sigtermHandler;
2366
    sigaction(SIGTERM, &act, NULL);
2367

2368 2369
#ifdef HAVE_BACKTRACE
    sigemptyset(&act.sa_mask);
2370
    act.sa_flags = SA_NODEFER | SA_RESETHAND | SA_SIGINFO;
2371 2372 2373 2374 2375 2376 2377
    act.sa_sigaction = sigsegvHandler;
    sigaction(SIGSEGV, &act, NULL);
    sigaction(SIGBUS, &act, NULL);
    sigaction(SIGFPE, &act, NULL);
    sigaction(SIGILL, &act, NULL);
#endif
    return;
2378 2379
}

2380 2381
void memtest(size_t megabytes, int passes);

2382 2383
int main(int argc, char **argv) {
    long long start;
2384
    struct timeval tv;
2385

A
antirez 已提交
2386
    /* We need to initialize our libraries, and the server configuration. */
2387
    zmalloc_enable_thread_safeness();
2388 2389 2390
    srand(time(NULL)^getpid());
    gettimeofday(&tv,NULL);
    dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
2391
    initServerConfig();
2392

2393 2394 2395 2396 2397 2398
    if (argc >= 2) {
        int j = 1; /* First option to parse in argv[] */
        sds options = sdsempty();
        char *configfile = NULL;

        /* Handle special options --help and --version */
2399 2400
        if (strcmp(argv[1], "-v") == 0 ||
            strcmp(argv[1], "--version") == 0) version();
2401 2402
        if (strcmp(argv[1], "--help") == 0 ||
            strcmp(argv[1], "-h") == 0) usage();
2403 2404
        if (strcmp(argv[1], "--test-memory") == 0) {
            if (argc == 3) {
2405
                memtest(atoi(argv[2]),50);
2406 2407 2408 2409 2410 2411 2412 2413
                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);
            }
        }

2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433
        /* 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++;
        }
2434
        resetServerSaveParams();
2435 2436
        loadServerConfig(configfile,options);
        sdsfree(options);
2437 2438 2439 2440 2441 2442 2443
    } else {
        redisLog(REDIS_WARNING,"Warning: no config file specified, using the default config. In order to specify a config file use 'redis-server /path/to/redis.conf'");
    }
    if (server.daemonize) daemonize();
    initServer();
    if (server.daemonize) createPidFile();
    redisAsciiArt();
2444
    redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
2445 2446 2447 2448
#ifdef __linux__
    linuxOvercommitMemoryWarning();
#endif
    start = ustime();
2449
    if (server.aof_state == REDIS_AOF_ON) {
2450
        if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
2451 2452
            redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
    } else {
A
antirez 已提交
2453
        if (rdbLoad(server.rdb_filename) == REDIS_OK) {
2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470
            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);
        }
    }
    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);
    aeSetBeforeSleepProc(server.el,beforeSleep);
    aeMain(server.el);
    aeDeleteEventLoop(server.el);
    return 0;
}

2471
/* The End */