]> git.saurik.com Git - redis.git/blob - src/redis.c
Copyright date fixed in COPYING file.
[redis.git] / src / redis.c
1 /*
2 * Copyright (c) 2009-2012, Salvatore Sanfilippo <antirez at gmail dot com>
3 * All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions are met:
7 *
8 * * Redistributions of source code must retain the above copyright notice,
9 * this list of conditions and the following disclaimer.
10 * * Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * * Neither the name of Redis nor the names of its contributors may be used
14 * to endorse or promote products derived from this software without
15 * specific prior written permission.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
18 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
21 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
22 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
23 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
24 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
25 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
26 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
27 * POSSIBILITY OF SUCH DAMAGE.
28 */
29
30 #include "redis.h"
31 #include "slowlog.h"
32 #include "bio.h"
33
34 #include <time.h>
35 #include <signal.h>
36 #include <sys/wait.h>
37 #include <errno.h>
38 #include <assert.h>
39 #include <ctype.h>
40 #include <stdarg.h>
41 #include <arpa/inet.h>
42 #include <sys/stat.h>
43 #include <fcntl.h>
44 #include <sys/time.h>
45 #include <sys/resource.h>
46 #include <sys/uio.h>
47 #include <limits.h>
48 #include <float.h>
49 #include <math.h>
50 #include <sys/resource.h>
51 #include <sys/utsname.h>
52
53 /* Our shared "common" objects */
54
55 struct sharedObjectsStruct shared;
56
57 /* Global vars that are actually used as constants. The following double
58 * values are used for double on-disk serialization, and are initialized
59 * at runtime to avoid strange compiler optimizations. */
60
61 double R_Zero, R_PosInf, R_NegInf, R_Nan;
62
63 /*================================= Globals ================================= */
64
65 /* Global vars */
66 struct redisServer server; /* server global state */
67 struct redisCommand *commandTable;
68
69 /* Our command table.
70 *
71 * Every entry is composed of the following fields:
72 *
73 * name: a string representing the command name.
74 * function: pointer to the C function implementing the command.
75 * arity: number of arguments, it is possible to use -N to say >= N
76 * sflags: command flags as string. See below for a table of flags.
77 * flags: flags as bitmask. Computed by Redis using the 'sflags' field.
78 * get_keys_proc: an optional function to get key arguments from a command.
79 * This is only used when the following three fields are not
80 * enough to specify what arguments are keys.
81 * first_key_index: first argument that is a key
82 * last_key_index: last argument that is a key
83 * key_step: step to get all the keys from first to last argument. For instance
84 * in MSET the step is two since arguments are key,val,key,val,...
85 * microseconds: microseconds of total execution time for this command.
86 * calls: total number of calls of this command.
87 *
88 * The flags, microseconds and calls fields are computed by Redis and should
89 * always be set to zero.
90 *
91 * Command flags are expressed using strings where every character represents
92 * a flag. Later the populateCommandTable() function will take care of
93 * populating the real 'flags' field using this characters.
94 *
95 * This is the meaning of the flags:
96 *
97 * w: write command (may modify the key space).
98 * r: read command (will never modify the key space).
99 * m: may increase memory usage once called. Don't allow if out of memory.
100 * a: admin command, like SAVE or SHUTDOWN.
101 * p: Pub/Sub related command.
102 * f: force replication of this command, regarless of server.dirty.
103 * s: command not allowed in scripts.
104 * R: random command. Command is not deterministic, that is, the same command
105 * with the same arguments, with the same key space, may have different
106 * results. For instance SPOP and RANDOMKEY are two random commands.
107 * S: Sort command output array if called from script, so that the output
108 * is deterministic.
109 * l: Allow command while loading the database.
110 * t: Allow command while a slave has stale data but is not allowed to
111 * server this data. Normally no command is accepted in this condition
112 * but just a few.
113 * M: Do not automatically propagate the command on MONITOR.
114 */
115 struct redisCommand redisCommandTable[] = {
116 {"get",getCommand,2,"r",0,NULL,1,1,1,0,0},
117 {"set",setCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
118 {"setnx",setnxCommand,3,"wm",0,noPreloadGetKeys,1,1,1,0,0},
119 {"setex",setexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
120 {"psetex",psetexCommand,4,"wm",0,noPreloadGetKeys,1,1,1,0,0},
121 {"append",appendCommand,3,"wm",0,NULL,1,1,1,0,0},
122 {"strlen",strlenCommand,2,"r",0,NULL,1,1,1,0,0},
123 {"del",delCommand,-2,"w",0,noPreloadGetKeys,1,-1,1,0,0},
124 {"exists",existsCommand,2,"r",0,NULL,1,1,1,0,0},
125 {"setbit",setbitCommand,4,"wm",0,NULL,1,1,1,0,0},
126 {"getbit",getbitCommand,3,"r",0,NULL,1,1,1,0,0},
127 {"setrange",setrangeCommand,4,"wm",0,NULL,1,1,1,0,0},
128 {"getrange",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
129 {"substr",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
130 {"incr",incrCommand,2,"wm",0,NULL,1,1,1,0,0},
131 {"decr",decrCommand,2,"wm",0,NULL,1,1,1,0,0},
132 {"mget",mgetCommand,-2,"r",0,NULL,1,-1,1,0,0},
133 {"rpush",rpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
134 {"lpush",lpushCommand,-3,"wm",0,NULL,1,1,1,0,0},
135 {"rpushx",rpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
136 {"lpushx",lpushxCommand,3,"wm",0,NULL,1,1,1,0,0},
137 {"linsert",linsertCommand,5,"wm",0,NULL,1,1,1,0,0},
138 {"rpop",rpopCommand,2,"w",0,NULL,1,1,1,0,0},
139 {"lpop",lpopCommand,2,"w",0,NULL,1,1,1,0,0},
140 {"brpop",brpopCommand,-3,"ws",0,NULL,1,1,1,0,0},
141 {"brpoplpush",brpoplpushCommand,4,"wms",0,NULL,1,2,1,0,0},
142 {"blpop",blpopCommand,-3,"ws",0,NULL,1,-2,1,0,0},
143 {"llen",llenCommand,2,"r",0,NULL,1,1,1,0,0},
144 {"lindex",lindexCommand,3,"r",0,NULL,1,1,1,0,0},
145 {"lset",lsetCommand,4,"wm",0,NULL,1,1,1,0,0},
146 {"lrange",lrangeCommand,4,"r",0,NULL,1,1,1,0,0},
147 {"ltrim",ltrimCommand,4,"w",0,NULL,1,1,1,0,0},
148 {"lrem",lremCommand,4,"w",0,NULL,1,1,1,0,0},
149 {"rpoplpush",rpoplpushCommand,3,"wm",0,NULL,1,2,1,0,0},
150 {"sadd",saddCommand,-3,"wm",0,NULL,1,1,1,0,0},
151 {"srem",sremCommand,-3,"w",0,NULL,1,1,1,0,0},
152 {"smove",smoveCommand,4,"w",0,NULL,1,2,1,0,0},
153 {"sismember",sismemberCommand,3,"r",0,NULL,1,1,1,0,0},
154 {"scard",scardCommand,2,"r",0,NULL,1,1,1,0,0},
155 {"spop",spopCommand,2,"wRs",0,NULL,1,1,1,0,0},
156 {"srandmember",srandmemberCommand,-2,"rR",0,NULL,1,1,1,0,0},
157 {"sinter",sinterCommand,-2,"rS",0,NULL,1,-1,1,0,0},
158 {"sinterstore",sinterstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
159 {"sunion",sunionCommand,-2,"rS",0,NULL,1,-1,1,0,0},
160 {"sunionstore",sunionstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
161 {"sdiff",sdiffCommand,-2,"rS",0,NULL,1,-1,1,0,0},
162 {"sdiffstore",sdiffstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
163 {"smembers",sinterCommand,2,"rS",0,NULL,1,1,1,0,0},
164 {"zadd",zaddCommand,-4,"wm",0,NULL,1,1,1,0,0},
165 {"zincrby",zincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
166 {"zrem",zremCommand,-3,"w",0,NULL,1,1,1,0,0},
167 {"zremrangebyscore",zremrangebyscoreCommand,4,"w",0,NULL,1,1,1,0,0},
168 {"zremrangebyrank",zremrangebyrankCommand,4,"w",0,NULL,1,1,1,0,0},
169 {"zunionstore",zunionstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
170 {"zinterstore",zinterstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
171 {"zrange",zrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
172 {"zrangebyscore",zrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
173 {"zrevrangebyscore",zrevrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
174 {"zcount",zcountCommand,4,"r",0,NULL,1,1,1,0,0},
175 {"zrevrange",zrevrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
176 {"zcard",zcardCommand,2,"r",0,NULL,1,1,1,0,0},
177 {"zscore",zscoreCommand,3,"r",0,NULL,1,1,1,0,0},
178 {"zrank",zrankCommand,3,"r",0,NULL,1,1,1,0,0},
179 {"zrevrank",zrevrankCommand,3,"r",0,NULL,1,1,1,0,0},
180 {"hset",hsetCommand,4,"wm",0,NULL,1,1,1,0,0},
181 {"hsetnx",hsetnxCommand,4,"wm",0,NULL,1,1,1,0,0},
182 {"hget",hgetCommand,3,"r",0,NULL,1,1,1,0,0},
183 {"hmset",hmsetCommand,-4,"wm",0,NULL,1,1,1,0,0},
184 {"hmget",hmgetCommand,-3,"r",0,NULL,1,1,1,0,0},
185 {"hincrby",hincrbyCommand,4,"wm",0,NULL,1,1,1,0,0},
186 {"hincrbyfloat",hincrbyfloatCommand,4,"wm",0,NULL,1,1,1,0,0},
187 {"hdel",hdelCommand,-3,"w",0,NULL,1,1,1,0,0},
188 {"hlen",hlenCommand,2,"r",0,NULL,1,1,1,0,0},
189 {"hkeys",hkeysCommand,2,"rS",0,NULL,1,1,1,0,0},
190 {"hvals",hvalsCommand,2,"rS",0,NULL,1,1,1,0,0},
191 {"hgetall",hgetallCommand,2,"r",0,NULL,1,1,1,0,0},
192 {"hexists",hexistsCommand,3,"r",0,NULL,1,1,1,0,0},
193 {"incrby",incrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
194 {"decrby",decrbyCommand,3,"wm",0,NULL,1,1,1,0,0},
195 {"incrbyfloat",incrbyfloatCommand,3,"wm",0,NULL,1,1,1,0,0},
196 {"getset",getsetCommand,3,"wm",0,NULL,1,1,1,0,0},
197 {"mset",msetCommand,-3,"wm",0,NULL,1,-1,2,0,0},
198 {"msetnx",msetnxCommand,-3,"wm",0,NULL,1,-1,2,0,0},
199 {"randomkey",randomkeyCommand,1,"rR",0,NULL,0,0,0,0,0},
200 {"select",selectCommand,2,"r",0,NULL,0,0,0,0,0},
201 {"move",moveCommand,3,"w",0,NULL,1,1,1,0,0},
202 {"rename",renameCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
203 {"renamenx",renamenxCommand,3,"w",0,renameGetKeys,1,2,1,0,0},
204 {"expire",expireCommand,3,"w",0,NULL,1,1,1,0,0},
205 {"expireat",expireatCommand,3,"w",0,NULL,1,1,1,0,0},
206 {"pexpire",pexpireCommand,3,"w",0,NULL,1,1,1,0,0},
207 {"pexpireat",pexpireatCommand,3,"w",0,NULL,1,1,1,0,0},
208 {"keys",keysCommand,2,"rS",0,NULL,0,0,0,0,0},
209 {"dbsize",dbsizeCommand,1,"r",0,NULL,0,0,0,0,0},
210 {"auth",authCommand,2,"rs",0,NULL,0,0,0,0,0},
211 {"ping",pingCommand,1,"r",0,NULL,0,0,0,0,0},
212 {"echo",echoCommand,2,"r",0,NULL,0,0,0,0,0},
213 {"save",saveCommand,1,"ars",0,NULL,0,0,0,0,0},
214 {"bgsave",bgsaveCommand,1,"ar",0,NULL,0,0,0,0,0},
215 {"bgrewriteaof",bgrewriteaofCommand,1,"ar",0,NULL,0,0,0,0,0},
216 {"shutdown",shutdownCommand,-1,"ar",0,NULL,0,0,0,0,0},
217 {"lastsave",lastsaveCommand,1,"r",0,NULL,0,0,0,0,0},
218 {"type",typeCommand,2,"r",0,NULL,1,1,1,0,0},
219 {"multi",multiCommand,1,"rs",0,NULL,0,0,0,0,0},
220 {"exec",execCommand,1,"sM",0,NULL,0,0,0,0,0},
221 {"discard",discardCommand,1,"rs",0,NULL,0,0,0,0,0},
222 {"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
223 {"replconf",replconfCommand,-1,"ars",0,NULL,0,0,0,0,0},
224 {"flushdb",flushdbCommand,1,"w",0,NULL,0,0,0,0,0},
225 {"flushall",flushallCommand,1,"w",0,NULL,0,0,0,0,0},
226 {"sort",sortCommand,-2,"wm",0,NULL,1,1,1,0,0},
227 {"info",infoCommand,-1,"rlt",0,NULL,0,0,0,0,0},
228 {"monitor",monitorCommand,1,"ars",0,NULL,0,0,0,0,0},
229 {"ttl",ttlCommand,2,"r",0,NULL,1,1,1,0,0},
230 {"pttl",pttlCommand,2,"r",0,NULL,1,1,1,0,0},
231 {"persist",persistCommand,2,"w",0,NULL,1,1,1,0,0},
232 {"slaveof",slaveofCommand,3,"ast",0,NULL,0,0,0,0,0},
233 {"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
234 {"config",configCommand,-2,"ar",0,NULL,0,0,0,0,0},
235 {"subscribe",subscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
236 {"unsubscribe",unsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
237 {"psubscribe",psubscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
238 {"punsubscribe",punsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
239 {"publish",publishCommand,3,"pflt",0,NULL,0,0,0,0,0},
240 {"watch",watchCommand,-2,"rs",0,noPreloadGetKeys,1,-1,1,0,0},
241 {"unwatch",unwatchCommand,1,"rs",0,NULL,0,0,0,0,0},
242 {"cluster",clusterCommand,-2,"ar",0,NULL,0,0,0,0,0},
243 {"restore",restoreCommand,-4,"awm",0,NULL,1,1,1,0,0},
244 {"migrate",migrateCommand,-6,"aw",0,NULL,0,0,0,0,0},
245 {"asking",askingCommand,1,"r",0,NULL,0,0,0,0,0},
246 {"dump",dumpCommand,2,"ar",0,NULL,1,1,1,0,0},
247 {"object",objectCommand,-2,"r",0,NULL,2,2,2,0,0},
248 {"client",clientCommand,-2,"ar",0,NULL,0,0,0,0,0},
249 {"eval",evalCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
250 {"evalsha",evalShaCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
251 {"slowlog",slowlogCommand,-2,"r",0,NULL,0,0,0,0,0},
252 {"script",scriptCommand,-2,"ras",0,NULL,0,0,0,0,0},
253 {"time",timeCommand,1,"rR",0,NULL,0,0,0,0,0},
254 {"bitop",bitopCommand,-4,"wm",0,NULL,2,-1,1,0,0},
255 {"bitcount",bitcountCommand,-2,"r",0,NULL,1,1,1,0,0}
256 };
257
258 /*============================ Utility functions ============================ */
259
260 /* Low level logging. To use only for very big messages, otherwise
261 * redisLog() is to prefer. */
262 void redisLogRaw(int level, const char *msg) {
263 const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
264 const char *c = ".-*#";
265 FILE *fp;
266 char buf[64];
267 int rawmode = (level & REDIS_LOG_RAW);
268
269 level &= 0xff; /* clear flags */
270 if (level < server.verbosity) return;
271
272 fp = (server.logfile == NULL) ? stdout : fopen(server.logfile,"a");
273 if (!fp) return;
274
275 if (rawmode) {
276 fprintf(fp,"%s",msg);
277 } else {
278 int off;
279 struct timeval tv;
280
281 gettimeofday(&tv,NULL);
282 off = strftime(buf,sizeof(buf),"%d %b %H:%M:%S.",localtime(&tv.tv_sec));
283 snprintf(buf+off,sizeof(buf)-off,"%03d",(int)tv.tv_usec/1000);
284 fprintf(fp,"[%d] %s %c %s\n",(int)getpid(),buf,c[level],msg);
285 }
286 fflush(fp);
287
288 if (server.logfile) fclose(fp);
289
290 if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
291 }
292
293 /* Like redisLogRaw() but with printf-alike support. This is the funciton that
294 * is used across the code. The raw version is only used in order to dump
295 * the INFO output on crash. */
296 void redisLog(int level, const char *fmt, ...) {
297 va_list ap;
298 char msg[REDIS_MAX_LOGMSG_LEN];
299
300 if ((level&0xff) < server.verbosity) return;
301
302 va_start(ap, fmt);
303 vsnprintf(msg, sizeof(msg), fmt, ap);
304 va_end(ap);
305
306 redisLogRaw(level,msg);
307 }
308
309 /* Log a fixed message without printf-alike capabilities, in a way that is
310 * safe to call from a signal handler.
311 *
312 * We actually use this only for signals that are not fatal from the point
313 * of view of Redis. Signals that are going to kill the server anyway and
314 * where we need printf-alike features are served by redisLog(). */
315 void redisLogFromHandler(int level, const char *msg) {
316 int fd;
317 char buf[64];
318
319 if ((level&0xff) < server.verbosity ||
320 (server.logfile == NULL && server.daemonize)) return;
321 fd = server.logfile ?
322 open(server.logfile, O_APPEND|O_CREAT|O_WRONLY, 0644) :
323 STDOUT_FILENO;
324 if (fd == -1) return;
325 ll2string(buf,sizeof(buf),getpid());
326 if (write(fd,"[",1) == -1) goto err;
327 if (write(fd,buf,strlen(buf)) == -1) goto err;
328 if (write(fd," | signal handler] (",20) == -1) goto err;
329 ll2string(buf,sizeof(buf),time(NULL));
330 if (write(fd,buf,strlen(buf)) == -1) goto err;
331 if (write(fd,") ",2) == -1) goto err;
332 if (write(fd,msg,strlen(msg)) == -1) goto err;
333 if (write(fd,"\n",1) == -1) goto err;
334 err:
335 if (server.logfile) close(fd);
336 }
337
338 /* Return the UNIX time in microseconds */
339 long long ustime(void) {
340 struct timeval tv;
341 long long ust;
342
343 gettimeofday(&tv, NULL);
344 ust = ((long long)tv.tv_sec)*1000000;
345 ust += tv.tv_usec;
346 return ust;
347 }
348
349 /* Return the UNIX time in milliseconds */
350 long long mstime(void) {
351 return ustime()/1000;
352 }
353
354 /* After an RDB dump or AOF rewrite we exit from children using _exit() instead of
355 * exit(), because the latter may interact with the same file objects used by
356 * the parent process. However if we are testing the coverage normal exit() is
357 * used in order to obtain the right coverage information. */
358 void exitFromChild(int retcode) {
359 #ifdef COVERAGE_TEST
360 exit(retcode);
361 #else
362 _exit(retcode);
363 #endif
364 }
365
366 /*====================== Hash table type implementation ==================== */
367
368 /* This is an hash table type that uses the SDS dynamic strings libary as
369 * keys and radis objects as values (objects can hold SDS strings,
370 * lists, sets). */
371
372 void dictVanillaFree(void *privdata, void *val)
373 {
374 DICT_NOTUSED(privdata);
375 zfree(val);
376 }
377
378 void dictListDestructor(void *privdata, void *val)
379 {
380 DICT_NOTUSED(privdata);
381 listRelease((list*)val);
382 }
383
384 int dictSdsKeyCompare(void *privdata, const void *key1,
385 const void *key2)
386 {
387 int l1,l2;
388 DICT_NOTUSED(privdata);
389
390 l1 = sdslen((sds)key1);
391 l2 = sdslen((sds)key2);
392 if (l1 != l2) return 0;
393 return memcmp(key1, key2, l1) == 0;
394 }
395
396 /* A case insensitive version used for the command lookup table. */
397 int dictSdsKeyCaseCompare(void *privdata, const void *key1,
398 const void *key2)
399 {
400 DICT_NOTUSED(privdata);
401
402 return strcasecmp(key1, key2) == 0;
403 }
404
405 void dictRedisObjectDestructor(void *privdata, void *val)
406 {
407 DICT_NOTUSED(privdata);
408
409 if (val == NULL) return; /* Values of swapped out keys as set to NULL */
410 decrRefCount(val);
411 }
412
413 void dictSdsDestructor(void *privdata, void *val)
414 {
415 DICT_NOTUSED(privdata);
416
417 sdsfree(val);
418 }
419
420 int dictObjKeyCompare(void *privdata, const void *key1,
421 const void *key2)
422 {
423 const robj *o1 = key1, *o2 = key2;
424 return dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
425 }
426
427 unsigned int dictObjHash(const void *key) {
428 const robj *o = key;
429 return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
430 }
431
432 unsigned int dictSdsHash(const void *key) {
433 return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
434 }
435
436 unsigned int dictSdsCaseHash(const void *key) {
437 return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
438 }
439
440 int dictEncObjKeyCompare(void *privdata, const void *key1,
441 const void *key2)
442 {
443 robj *o1 = (robj*) key1, *o2 = (robj*) key2;
444 int cmp;
445
446 if (o1->encoding == REDIS_ENCODING_INT &&
447 o2->encoding == REDIS_ENCODING_INT)
448 return o1->ptr == o2->ptr;
449
450 o1 = getDecodedObject(o1);
451 o2 = getDecodedObject(o2);
452 cmp = dictSdsKeyCompare(privdata,o1->ptr,o2->ptr);
453 decrRefCount(o1);
454 decrRefCount(o2);
455 return cmp;
456 }
457
458 unsigned int dictEncObjHash(const void *key) {
459 robj *o = (robj*) key;
460
461 if (o->encoding == REDIS_ENCODING_RAW) {
462 return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
463 } else {
464 if (o->encoding == REDIS_ENCODING_INT) {
465 char buf[32];
466 int len;
467
468 len = ll2string(buf,32,(long)o->ptr);
469 return dictGenHashFunction((unsigned char*)buf, len);
470 } else {
471 unsigned int hash;
472
473 o = getDecodedObject(o);
474 hash = dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
475 decrRefCount(o);
476 return hash;
477 }
478 }
479 }
480
481 /* Sets type hash table */
482 dictType setDictType = {
483 dictEncObjHash, /* hash function */
484 NULL, /* key dup */
485 NULL, /* val dup */
486 dictEncObjKeyCompare, /* key compare */
487 dictRedisObjectDestructor, /* key destructor */
488 NULL /* val destructor */
489 };
490
491 /* Sorted sets hash (note: a skiplist is used in addition to the hash table) */
492 dictType zsetDictType = {
493 dictEncObjHash, /* hash function */
494 NULL, /* key dup */
495 NULL, /* val dup */
496 dictEncObjKeyCompare, /* key compare */
497 dictRedisObjectDestructor, /* key destructor */
498 NULL /* val destructor */
499 };
500
501 /* Db->dict, keys are sds strings, vals are Redis objects. */
502 dictType dbDictType = {
503 dictSdsHash, /* hash function */
504 NULL, /* key dup */
505 NULL, /* val dup */
506 dictSdsKeyCompare, /* key compare */
507 dictSdsDestructor, /* key destructor */
508 dictRedisObjectDestructor /* val destructor */
509 };
510
511 /* Db->expires */
512 dictType keyptrDictType = {
513 dictSdsHash, /* hash function */
514 NULL, /* key dup */
515 NULL, /* val dup */
516 dictSdsKeyCompare, /* key compare */
517 NULL, /* key destructor */
518 NULL /* val destructor */
519 };
520
521 /* Command table. sds string -> command struct pointer. */
522 dictType commandTableDictType = {
523 dictSdsCaseHash, /* hash function */
524 NULL, /* key dup */
525 NULL, /* val dup */
526 dictSdsKeyCaseCompare, /* key compare */
527 dictSdsDestructor, /* key destructor */
528 NULL /* val destructor */
529 };
530
531 /* Hash type hash table (note that small hashes are represented with zimpaps) */
532 dictType hashDictType = {
533 dictEncObjHash, /* hash function */
534 NULL, /* key dup */
535 NULL, /* val dup */
536 dictEncObjKeyCompare, /* key compare */
537 dictRedisObjectDestructor, /* key destructor */
538 dictRedisObjectDestructor /* val destructor */
539 };
540
541 /* Keylist hash table type has unencoded redis objects as keys and
542 * lists as values. It's used for blocking operations (BLPOP) and to
543 * map swapped keys to a list of clients waiting for this keys to be loaded. */
544 dictType keylistDictType = {
545 dictObjHash, /* hash function */
546 NULL, /* key dup */
547 NULL, /* val dup */
548 dictObjKeyCompare, /* key compare */
549 dictRedisObjectDestructor, /* key destructor */
550 dictListDestructor /* val destructor */
551 };
552
553 /* Cluster nodes hash table, mapping nodes addresses 1.2.3.4:6379 to
554 * clusterNode structures. */
555 dictType clusterNodesDictType = {
556 dictSdsHash, /* hash function */
557 NULL, /* key dup */
558 NULL, /* val dup */
559 dictSdsKeyCompare, /* key compare */
560 dictSdsDestructor, /* key destructor */
561 NULL /* val destructor */
562 };
563
564 int htNeedsResize(dict *dict) {
565 long long size, used;
566
567 size = dictSlots(dict);
568 used = dictSize(dict);
569 return (size && used && size > DICT_HT_INITIAL_SIZE &&
570 (used*100/size < REDIS_HT_MINFILL));
571 }
572
573 /* If the percentage of used slots in the HT reaches REDIS_HT_MINFILL
574 * we resize the hash table to save memory */
575 void tryResizeHashTables(void) {
576 int j;
577
578 for (j = 0; j < server.dbnum; j++) {
579 if (htNeedsResize(server.db[j].dict))
580 dictResize(server.db[j].dict);
581 if (htNeedsResize(server.db[j].expires))
582 dictResize(server.db[j].expires);
583 }
584 }
585
586 /* Our hash table implementation performs rehashing incrementally while
587 * we write/read from the hash table. Still if the server is idle, the hash
588 * table will use two tables for a long time. So we try to use 1 millisecond
589 * of CPU time at every serverCron() loop in order to rehash some key. */
590 void incrementallyRehash(void) {
591 int j;
592
593 for (j = 0; j < server.dbnum; j++) {
594 /* Keys dictionary */
595 if (dictIsRehashing(server.db[j].dict)) {
596 dictRehashMilliseconds(server.db[j].dict,1);
597 break; /* already used our millisecond for this loop... */
598 }
599 /* Expires */
600 if (dictIsRehashing(server.db[j].expires)) {
601 dictRehashMilliseconds(server.db[j].expires,1);
602 break; /* already used our millisecond for this loop... */
603 }
604 }
605 }
606
607 /* This function is called once a background process of some kind terminates,
608 * as we want to avoid resizing the hash tables when there is a child in order
609 * to play well with copy-on-write (otherwise when a resize happens lots of
610 * memory pages are copied). The goal of this function is to update the ability
611 * for dict.c to resize the hash tables accordingly to the fact we have o not
612 * running childs. */
613 void updateDictResizePolicy(void) {
614 if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
615 dictEnableResize();
616 else
617 dictDisableResize();
618 }
619
620 /* ======================= Cron: called every 100 ms ======================== */
621
622 /* Try to expire a few timed out keys. The algorithm used is adaptive and
623 * will use few CPU cycles if there are few expiring keys, otherwise
624 * it will get more aggressive to avoid that too much memory is used by
625 * keys that can be removed from the keyspace. */
626 void activeExpireCycle(void) {
627 int j, iteration = 0;
628 long long start = ustime(), timelimit;
629
630 /* We can use at max REDIS_EXPIRELOOKUPS_TIME_PERC percentage of CPU time
631 * per iteration. Since this function gets called with a frequency of
632 * REDIS_HZ times per second, the following is the max amount of
633 * microseconds we can spend in this function. */
634 timelimit = 1000000*REDIS_EXPIRELOOKUPS_TIME_PERC/REDIS_HZ/100;
635 if (timelimit <= 0) timelimit = 1;
636
637 for (j = 0; j < server.dbnum; j++) {
638 int expired;
639 redisDb *db = server.db+j;
640
641 /* Continue to expire if at the end of the cycle more than 25%
642 * of the keys were expired. */
643 do {
644 unsigned long num = dictSize(db->expires);
645 unsigned long slots = dictSlots(db->expires);
646 long long now = mstime();
647
648 /* When there are less than 1% filled slots getting random
649 * keys is expensive, so stop here waiting for better times...
650 * The dictionary will be resized asap. */
651 if (num && slots > DICT_HT_INITIAL_SIZE &&
652 (num*100/slots < 1)) break;
653
654 /* The main collection cycle. Sample random keys among keys
655 * with an expire set, checking for expired ones. */
656 expired = 0;
657 if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
658 num = REDIS_EXPIRELOOKUPS_PER_CRON;
659 while (num--) {
660 dictEntry *de;
661 long long t;
662
663 if ((de = dictGetRandomKey(db->expires)) == NULL) break;
664 t = dictGetSignedIntegerVal(de);
665 if (now > t) {
666 sds key = dictGetKey(de);
667 robj *keyobj = createStringObject(key,sdslen(key));
668
669 propagateExpire(db,keyobj);
670 dbDelete(db,keyobj);
671 decrRefCount(keyobj);
672 expired++;
673 server.stat_expiredkeys++;
674 }
675 }
676 /* We can't block forever here even if there are many keys to
677 * expire. So after a given amount of milliseconds return to the
678 * caller waiting for the other active expire cycle. */
679 iteration++;
680 if ((iteration & 0xf) == 0 && /* check once every 16 cycles. */
681 (ustime()-start) > timelimit) return;
682 } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
683 }
684 }
685
686 void updateLRUClock(void) {
687 server.lruclock = (server.unixtime/REDIS_LRU_CLOCK_RESOLUTION) &
688 REDIS_LRU_CLOCK_MAX;
689 }
690
691
692 /* Add a sample to the operations per second array of samples. */
693 void trackOperationsPerSecond(void) {
694 long long t = mstime() - server.ops_sec_last_sample_time;
695 long long ops = server.stat_numcommands - server.ops_sec_last_sample_ops;
696 long long ops_sec;
697
698 ops_sec = t > 0 ? (ops*1000/t) : 0;
699
700 server.ops_sec_samples[server.ops_sec_idx] = ops_sec;
701 server.ops_sec_idx = (server.ops_sec_idx+1) % REDIS_OPS_SEC_SAMPLES;
702 server.ops_sec_last_sample_time = mstime();
703 server.ops_sec_last_sample_ops = server.stat_numcommands;
704 }
705
706 /* Return the mean of all the samples. */
707 long long getOperationsPerSecond(void) {
708 int j;
709 long long sum = 0;
710
711 for (j = 0; j < REDIS_OPS_SEC_SAMPLES; j++)
712 sum += server.ops_sec_samples[j];
713 return sum / REDIS_OPS_SEC_SAMPLES;
714 }
715
716 /* Check for timeouts. Returns non-zero if the client was terminated */
717 int clientsCronHandleTimeout(redisClient *c) {
718 time_t now = server.unixtime;
719
720 if (server.maxidletime &&
721 !(c->flags & REDIS_SLAVE) && /* no timeout for slaves */
722 !(c->flags & REDIS_MASTER) && /* no timeout for masters */
723 !(c->flags & REDIS_BLOCKED) && /* no timeout for BLPOP */
724 dictSize(c->pubsub_channels) == 0 && /* no timeout for pubsub */
725 listLength(c->pubsub_patterns) == 0 &&
726 (now - c->lastinteraction > server.maxidletime))
727 {
728 redisLog(REDIS_VERBOSE,"Closing idle client");
729 freeClient(c);
730 return 1;
731 } else if (c->flags & REDIS_BLOCKED) {
732 if (c->bpop.timeout != 0 && c->bpop.timeout < now) {
733 addReply(c,shared.nullmultibulk);
734 unblockClientWaitingData(c);
735 }
736 }
737 return 0;
738 }
739
740 /* The client query buffer is an sds.c string that can end with a lot of
741 * free space not used, this function reclaims space if needed.
742 *
743 * The funciton always returns 0 as it never terminates the client. */
744 int clientsCronResizeQueryBuffer(redisClient *c) {
745 size_t querybuf_size = sdsAllocSize(c->querybuf);
746 time_t idletime = server.unixtime - c->lastinteraction;
747
748 /* There are two conditions to resize the query buffer:
749 * 1) Query buffer is > BIG_ARG and too big for latest peak.
750 * 2) Client is inactive and the buffer is bigger than 1k. */
751 if (((querybuf_size > REDIS_MBULK_BIG_ARG) &&
752 (querybuf_size/(c->querybuf_peak+1)) > 2) ||
753 (querybuf_size > 1024 && idletime > 2))
754 {
755 /* Only resize the query buffer if it is actually wasting space. */
756 if (sdsavail(c->querybuf) > 1024) {
757 c->querybuf = sdsRemoveFreeSpace(c->querybuf);
758 }
759 }
760 /* Reset the peak again to capture the peak memory usage in the next
761 * cycle. */
762 c->querybuf_peak = 0;
763 return 0;
764 }
765
766 void clientsCron(void) {
767 /* Make sure to process at least 1/(REDIS_HZ*10) of clients per call.
768 * Since this function is called REDIS_HZ times per second we are sure that
769 * in the worst case we process all the clients in 10 seconds.
770 * In normal conditions (a reasonable number of clients) we process
771 * all the clients in a shorter time. */
772 int numclients = listLength(server.clients);
773 int iterations = numclients/(REDIS_HZ*10);
774
775 if (iterations < 50)
776 iterations = (numclients < 50) ? numclients : 50;
777 while(listLength(server.clients) && iterations--) {
778 redisClient *c;
779 listNode *head;
780
781 /* Rotate the list, take the current head, process.
782 * This way if the client must be removed from the list it's the
783 * first element and we don't incur into O(N) computation. */
784 listRotate(server.clients);
785 head = listFirst(server.clients);
786 c = listNodeValue(head);
787 /* The following functions do different service checks on the client.
788 * The protocol is that they return non-zero if the client was
789 * terminated. */
790 if (clientsCronHandleTimeout(c)) continue;
791 if (clientsCronResizeQueryBuffer(c)) continue;
792 }
793 }
794
795 /* This is our timer interrupt, called REDIS_HZ times per second.
796 * Here is where we do a number of things that need to be done asynchronously.
797 * For instance:
798 *
799 * - Active expired keys collection (it is also performed in a lazy way on
800 * lookup).
801 * - Software watchdong.
802 * - Update some statistic.
803 * - Incremental rehashing of the DBs hash tables.
804 * - Triggering BGSAVE / AOF rewrite, and handling of terminated children.
805 * - Clients timeout of differnet kinds.
806 * - Replication reconnection.
807 * - Many more...
808 *
809 * Everything directly called here will be called REDIS_HZ times per second,
810 * so in order to throttle execution of things we want to do less frequently
811 * a macro is used: run_with_period(milliseconds) { .... }
812 */
813
814 int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
815 int j;
816 REDIS_NOTUSED(eventLoop);
817 REDIS_NOTUSED(id);
818 REDIS_NOTUSED(clientData);
819
820 /* Software watchdog: deliver the SIGALRM that will reach the signal
821 * handler if we don't return here fast enough. */
822 if (server.watchdog_period) watchdogScheduleSignal(server.watchdog_period);
823
824 /* We take a cached value of the unix time in the global state because
825 * with virtual memory and aging there is to store the current time
826 * in objects at every object access, and accuracy is not needed.
827 * To access a global var is faster than calling time(NULL) */
828 server.unixtime = time(NULL);
829
830 run_with_period(100) trackOperationsPerSecond();
831
832 /* We have just 22 bits per object for LRU information.
833 * So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
834 * 2^22 bits with 10 seconds resoluton is more or less 1.5 years.
835 *
836 * Note that even if this will wrap after 1.5 years it's not a problem,
837 * everything will still work but just some object will appear younger
838 * to Redis. But for this to happen a given object should never be touched
839 * for 1.5 years.
840 *
841 * Note that you can change the resolution altering the
842 * REDIS_LRU_CLOCK_RESOLUTION define.
843 */
844 updateLRUClock();
845
846 /* Record the max memory used since the server was started. */
847 if (zmalloc_used_memory() > server.stat_peak_memory)
848 server.stat_peak_memory = zmalloc_used_memory();
849
850 /* We received a SIGTERM, shutting down here in a safe way, as it is
851 * not ok doing so inside the signal handler. */
852 if (server.shutdown_asap) {
853 if (prepareForShutdown(0) == REDIS_OK) exit(0);
854 redisLog(REDIS_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
855 }
856
857 /* Show some info about non-empty databases */
858 run_with_period(5000) {
859 for (j = 0; j < server.dbnum; j++) {
860 long long size, used, vkeys;
861
862 size = dictSlots(server.db[j].dict);
863 used = dictSize(server.db[j].dict);
864 vkeys = dictSize(server.db[j].expires);
865 if (used || vkeys) {
866 redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
867 /* dictPrintStats(server.dict); */
868 }
869 }
870 }
871
872 /* We don't want to resize the hash tables while a bacground saving
873 * is in progress: the saving child is created using fork() that is
874 * implemented with a copy-on-write semantic in most modern systems, so
875 * if we resize the HT while there is the saving child at work actually
876 * a lot of memory movements in the parent will cause a lot of pages
877 * copied. */
878 if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
879 tryResizeHashTables();
880 if (server.activerehashing) incrementallyRehash();
881 }
882
883 /* Show information about connected clients */
884 if (!server.sentinel_mode) {
885 run_with_period(5000) {
886 redisLog(REDIS_VERBOSE,
887 "%d clients connected (%d slaves), %zu bytes in use",
888 listLength(server.clients)-listLength(server.slaves),
889 listLength(server.slaves),
890 zmalloc_used_memory());
891 }
892 }
893
894 /* We need to do a few operations on clients asynchronously. */
895 clientsCron();
896
897 /* Start a scheduled AOF rewrite if this was requested by the user while
898 * a BGSAVE was in progress. */
899 if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
900 server.aof_rewrite_scheduled)
901 {
902 rewriteAppendOnlyFileBackground();
903 }
904
905 /* Check if a background saving or AOF rewrite in progress terminated. */
906 if (server.rdb_child_pid != -1 || server.aof_child_pid != -1) {
907 int statloc;
908 pid_t pid;
909
910 if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
911 int exitcode = WEXITSTATUS(statloc);
912 int bysignal = 0;
913
914 if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);
915
916 if (pid == server.rdb_child_pid) {
917 backgroundSaveDoneHandler(exitcode,bysignal);
918 } else if (pid == server.aof_child_pid) {
919 backgroundRewriteDoneHandler(exitcode,bysignal);
920 } else {
921 redisLog(REDIS_WARNING,
922 "Warning, detected child with unmatched pid: %ld",
923 (long)pid);
924 }
925 updateDictResizePolicy();
926 }
927 } else {
928 /* If there is not a background saving/rewrite in progress check if
929 * we have to save/rewrite now */
930 for (j = 0; j < server.saveparamslen; j++) {
931 struct saveparam *sp = server.saveparams+j;
932
933 if (server.dirty >= sp->changes &&
934 server.unixtime-server.lastsave > sp->seconds) {
935 redisLog(REDIS_NOTICE,"%d changes in %d seconds. Saving...",
936 sp->changes, sp->seconds);
937 rdbSaveBackground(server.rdb_filename);
938 break;
939 }
940 }
941
942 /* Trigger an AOF rewrite if needed */
943 if (server.rdb_child_pid == -1 &&
944 server.aof_child_pid == -1 &&
945 server.aof_rewrite_perc &&
946 server.aof_current_size > server.aof_rewrite_min_size)
947 {
948 long long base = server.aof_rewrite_base_size ?
949 server.aof_rewrite_base_size : 1;
950 long long growth = (server.aof_current_size*100/base) - 100;
951 if (growth >= server.aof_rewrite_perc) {
952 redisLog(REDIS_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth);
953 rewriteAppendOnlyFileBackground();
954 }
955 }
956 }
957
958
959 /* If we postponed an AOF buffer flush, let's try to do it every time the
960 * cron function is called. */
961 if (server.aof_flush_postponed_start) flushAppendOnlyFile(0);
962
963 /* Expire a few keys per cycle, only if this is a master.
964 * On slaves we wait for DEL operations synthesized by the master
965 * in order to guarantee a strict consistency. */
966 if (server.masterhost == NULL) activeExpireCycle();
967
968 /* Close clients that need to be closed asynchronous */
969 freeClientsInAsyncFreeQueue();
970
971 /* Replication cron function -- used to reconnect to master and
972 * to detect transfer failures. */
973 run_with_period(1000) replicationCron();
974
975 /* Run other sub-systems specific cron jobs */
976 run_with_period(1000) {
977 if (server.cluster_enabled) clusterCron();
978 }
979
980 /* Run the sentinel timer if we are in sentinel mode. */
981 run_with_period(100) {
982 if (server.sentinel_mode) sentinelTimer();
983 }
984
985 server.cronloops++;
986 return 1000/REDIS_HZ;
987 }
988
989 /* This function gets called every time Redis is entering the
990 * main loop of the event driven library, that is, before to sleep
991 * for ready file descriptors. */
992 void beforeSleep(struct aeEventLoop *eventLoop) {
993 REDIS_NOTUSED(eventLoop);
994 listNode *ln;
995 redisClient *c;
996
997 /* Try to process pending commands for clients that were just unblocked. */
998 while (listLength(server.unblocked_clients)) {
999 ln = listFirst(server.unblocked_clients);
1000 redisAssert(ln != NULL);
1001 c = ln->value;
1002 listDelNode(server.unblocked_clients,ln);
1003 c->flags &= ~REDIS_UNBLOCKED;
1004
1005 /* Process remaining data in the input buffer. */
1006 if (c->querybuf && sdslen(c->querybuf) > 0) {
1007 server.current_client = c;
1008 processInputBuffer(c);
1009 server.current_client = NULL;
1010 }
1011 }
1012
1013 /* Write the AOF buffer on disk */
1014 flushAppendOnlyFile(0);
1015 }
1016
1017 /* =========================== Server initialization ======================== */
1018
1019 void createSharedObjects(void) {
1020 int j;
1021
1022 shared.crlf = createObject(REDIS_STRING,sdsnew("\r\n"));
1023 shared.ok = createObject(REDIS_STRING,sdsnew("+OK\r\n"));
1024 shared.err = createObject(REDIS_STRING,sdsnew("-ERR\r\n"));
1025 shared.emptybulk = createObject(REDIS_STRING,sdsnew("$0\r\n\r\n"));
1026 shared.czero = createObject(REDIS_STRING,sdsnew(":0\r\n"));
1027 shared.cone = createObject(REDIS_STRING,sdsnew(":1\r\n"));
1028 shared.cnegone = createObject(REDIS_STRING,sdsnew(":-1\r\n"));
1029 shared.nullbulk = createObject(REDIS_STRING,sdsnew("$-1\r\n"));
1030 shared.nullmultibulk = createObject(REDIS_STRING,sdsnew("*-1\r\n"));
1031 shared.emptymultibulk = createObject(REDIS_STRING,sdsnew("*0\r\n"));
1032 shared.pong = createObject(REDIS_STRING,sdsnew("+PONG\r\n"));
1033 shared.queued = createObject(REDIS_STRING,sdsnew("+QUEUED\r\n"));
1034 shared.wrongtypeerr = createObject(REDIS_STRING,sdsnew(
1035 "-WRONGTYPE Operation against a key holding the wrong kind of value\r\n"));
1036 shared.nokeyerr = createObject(REDIS_STRING,sdsnew(
1037 "-ERR no such key\r\n"));
1038 shared.syntaxerr = createObject(REDIS_STRING,sdsnew(
1039 "-ERR syntax error\r\n"));
1040 shared.sameobjecterr = createObject(REDIS_STRING,sdsnew(
1041 "-ERR source and destination objects are the same\r\n"));
1042 shared.outofrangeerr = createObject(REDIS_STRING,sdsnew(
1043 "-ERR index out of range\r\n"));
1044 shared.noscripterr = createObject(REDIS_STRING,sdsnew(
1045 "-NOSCRIPT No matching script. Please use EVAL.\r\n"));
1046 shared.loadingerr = createObject(REDIS_STRING,sdsnew(
1047 "-LOADING Redis is loading the dataset in memory\r\n"));
1048 shared.slowscripterr = createObject(REDIS_STRING,sdsnew(
1049 "-BUSY Redis is busy running a script. You can only call SCRIPT KILL or SHUTDOWN NOSAVE.\r\n"));
1050 shared.masterdownerr = createObject(REDIS_STRING,sdsnew(
1051 "-MASTERDOWN Link with MASTER is down and slave-serve-stale-data is set to 'no'.\r\n"));
1052 shared.bgsaveerr = createObject(REDIS_STRING,sdsnew(
1053 "-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"));
1054 shared.roslaveerr = createObject(REDIS_STRING,sdsnew(
1055 "-READONLY You can't write against a read only slave.\r\n"));
1056 shared.oomerr = createObject(REDIS_STRING,sdsnew(
1057 "-OOM command not allowed when used memory > 'maxmemory'.\r\n"));
1058 shared.space = createObject(REDIS_STRING,sdsnew(" "));
1059 shared.colon = createObject(REDIS_STRING,sdsnew(":"));
1060 shared.plus = createObject(REDIS_STRING,sdsnew("+"));
1061
1062 for (j = 0; j < REDIS_SHARED_SELECT_CMDS; j++) {
1063 shared.select[j] = createObject(REDIS_STRING,
1064 sdscatprintf(sdsempty(),"select %d\r\n", j));
1065 }
1066 shared.messagebulk = createStringObject("$7\r\nmessage\r\n",13);
1067 shared.pmessagebulk = createStringObject("$8\r\npmessage\r\n",14);
1068 shared.subscribebulk = createStringObject("$9\r\nsubscribe\r\n",15);
1069 shared.unsubscribebulk = createStringObject("$11\r\nunsubscribe\r\n",18);
1070 shared.psubscribebulk = createStringObject("$10\r\npsubscribe\r\n",17);
1071 shared.punsubscribebulk = createStringObject("$12\r\npunsubscribe\r\n",19);
1072 shared.del = createStringObject("DEL",3);
1073 shared.rpop = createStringObject("RPOP",4);
1074 shared.lpop = createStringObject("LPOP",4);
1075 shared.lpush = createStringObject("LPUSH",5);
1076 for (j = 0; j < REDIS_SHARED_INTEGERS; j++) {
1077 shared.integers[j] = createObject(REDIS_STRING,(void*)(long)j);
1078 shared.integers[j]->encoding = REDIS_ENCODING_INT;
1079 }
1080 for (j = 0; j < REDIS_SHARED_BULKHDR_LEN; j++) {
1081 shared.mbulkhdr[j] = createObject(REDIS_STRING,
1082 sdscatprintf(sdsempty(),"*%d\r\n",j));
1083 shared.bulkhdr[j] = createObject(REDIS_STRING,
1084 sdscatprintf(sdsempty(),"$%d\r\n",j));
1085 }
1086 }
1087
1088 void initServerConfig() {
1089 getRandomHexChars(server.runid,REDIS_RUN_ID_SIZE);
1090 server.runid[REDIS_RUN_ID_SIZE] = '\0';
1091 server.arch_bits = (sizeof(long) == 8) ? 64 : 32;
1092 server.port = REDIS_SERVERPORT;
1093 server.bindaddr = NULL;
1094 server.unixsocket = NULL;
1095 server.unixsocketperm = 0;
1096 server.ipfd = -1;
1097 server.sofd = -1;
1098 server.dbnum = REDIS_DEFAULT_DBNUM;
1099 server.verbosity = REDIS_NOTICE;
1100 server.maxidletime = REDIS_MAXIDLETIME;
1101 server.client_max_querybuf_len = REDIS_MAX_QUERYBUF_LEN;
1102 server.saveparams = NULL;
1103 server.loading = 0;
1104 server.logfile = NULL; /* NULL = log on standard output */
1105 server.syslog_enabled = 0;
1106 server.syslog_ident = zstrdup("redis");
1107 server.syslog_facility = LOG_LOCAL0;
1108 server.daemonize = 0;
1109 server.aof_state = REDIS_AOF_OFF;
1110 server.aof_fsync = AOF_FSYNC_EVERYSEC;
1111 server.aof_no_fsync_on_rewrite = 0;
1112 server.aof_rewrite_perc = REDIS_AOF_REWRITE_PERC;
1113 server.aof_rewrite_min_size = REDIS_AOF_REWRITE_MIN_SIZE;
1114 server.aof_rewrite_base_size = 0;
1115 server.aof_rewrite_scheduled = 0;
1116 server.aof_last_fsync = time(NULL);
1117 server.aof_rewrite_time_last = -1;
1118 server.aof_rewrite_time_start = -1;
1119 server.aof_lastbgrewrite_status = REDIS_OK;
1120 server.aof_delayed_fsync = 0;
1121 server.aof_fd = -1;
1122 server.aof_selected_db = -1; /* Make sure the first time will not match */
1123 server.aof_flush_postponed_start = 0;
1124 server.pidfile = zstrdup("/var/run/redis.pid");
1125 server.rdb_filename = zstrdup("dump.rdb");
1126 server.aof_filename = zstrdup("appendonly.aof");
1127 server.requirepass = NULL;
1128 server.rdb_compression = 1;
1129 server.rdb_checksum = 1;
1130 server.activerehashing = 1;
1131 server.maxclients = REDIS_MAX_CLIENTS;
1132 server.bpop_blocked_clients = 0;
1133 server.maxmemory = 0;
1134 server.maxmemory_policy = REDIS_MAXMEMORY_VOLATILE_LRU;
1135 server.maxmemory_samples = 3;
1136 server.hash_max_ziplist_entries = REDIS_HASH_MAX_ZIPLIST_ENTRIES;
1137 server.hash_max_ziplist_value = REDIS_HASH_MAX_ZIPLIST_VALUE;
1138 server.list_max_ziplist_entries = REDIS_LIST_MAX_ZIPLIST_ENTRIES;
1139 server.list_max_ziplist_value = REDIS_LIST_MAX_ZIPLIST_VALUE;
1140 server.set_max_intset_entries = REDIS_SET_MAX_INTSET_ENTRIES;
1141 server.zset_max_ziplist_entries = REDIS_ZSET_MAX_ZIPLIST_ENTRIES;
1142 server.zset_max_ziplist_value = REDIS_ZSET_MAX_ZIPLIST_VALUE;
1143 server.shutdown_asap = 0;
1144 server.repl_ping_slave_period = REDIS_REPL_PING_SLAVE_PERIOD;
1145 server.repl_timeout = REDIS_REPL_TIMEOUT;
1146 server.cluster_enabled = 0;
1147 server.cluster.configfile = zstrdup("nodes.conf");
1148 server.lua_caller = NULL;
1149 server.lua_time_limit = REDIS_LUA_TIME_LIMIT;
1150 server.lua_client = NULL;
1151 server.lua_timedout = 0;
1152
1153 updateLRUClock();
1154 resetServerSaveParams();
1155
1156 appendServerSaveParams(60*60,1); /* save after 1 hour and 1 change */
1157 appendServerSaveParams(300,100); /* save after 5 minutes and 100 changes */
1158 appendServerSaveParams(60,10000); /* save after 1 minute and 10000 changes */
1159 /* Replication related */
1160 server.masterauth = NULL;
1161 server.masterhost = NULL;
1162 server.masterport = 6379;
1163 server.master = NULL;
1164 server.repl_state = REDIS_REPL_NONE;
1165 server.repl_syncio_timeout = REDIS_REPL_SYNCIO_TIMEOUT;
1166 server.repl_serve_stale_data = 1;
1167 server.repl_slave_ro = 1;
1168 server.repl_down_since = time(NULL);
1169 server.slave_priority = REDIS_DEFAULT_SLAVE_PRIORITY;
1170
1171 /* Client output buffer limits */
1172 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].hard_limit_bytes = 0;
1173 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_bytes = 0;
1174 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_NORMAL].soft_limit_seconds = 0;
1175 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].hard_limit_bytes = 1024*1024*256;
1176 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_bytes = 1024*1024*64;
1177 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_SLAVE].soft_limit_seconds = 60;
1178 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].hard_limit_bytes = 1024*1024*32;
1179 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_bytes = 1024*1024*8;
1180 server.client_obuf_limits[REDIS_CLIENT_LIMIT_CLASS_PUBSUB].soft_limit_seconds = 60;
1181
1182 /* Double constants initialization */
1183 R_Zero = 0.0;
1184 R_PosInf = 1.0/R_Zero;
1185 R_NegInf = -1.0/R_Zero;
1186 R_Nan = R_Zero/R_Zero;
1187
1188 /* Command table -- we intiialize it here as it is part of the
1189 * initial configuration, since command names may be changed via
1190 * redis.conf using the rename-command directive. */
1191 server.commands = dictCreate(&commandTableDictType,NULL);
1192 populateCommandTable();
1193 server.delCommand = lookupCommandByCString("del");
1194 server.multiCommand = lookupCommandByCString("multi");
1195 server.lpushCommand = lookupCommandByCString("lpush");
1196 server.lpopCommand = lookupCommandByCString("lpop");
1197 server.rpopCommand = lookupCommandByCString("rpop");
1198
1199 /* Slow log */
1200 server.slowlog_log_slower_than = REDIS_SLOWLOG_LOG_SLOWER_THAN;
1201 server.slowlog_max_len = REDIS_SLOWLOG_MAX_LEN;
1202
1203 /* Debugging */
1204 server.assert_failed = "<no assertion failed>";
1205 server.assert_file = "<no file>";
1206 server.assert_line = 0;
1207 server.bug_report_start = 0;
1208 server.watchdog_period = 0;
1209 }
1210
1211 /* This function will try to raise the max number of open files accordingly to
1212 * the configured max number of clients. It will also account for 32 additional
1213 * file descriptors as we need a few more for persistence, listening
1214 * sockets, log files and so forth.
1215 *
1216 * If it will not be possible to set the limit accordingly to the configured
1217 * max number of clients, the function will do the reverse setting
1218 * server.maxclients to the value that we can actually handle. */
1219 void adjustOpenFilesLimit(void) {
1220 rlim_t maxfiles = server.maxclients+32;
1221 struct rlimit limit;
1222
1223 if (getrlimit(RLIMIT_NOFILE,&limit) == -1) {
1224 redisLog(REDIS_WARNING,"Unable to obtain the current NOFILE limit (%s), assuming 1024 and setting the max clients configuration accordingly.",
1225 strerror(errno));
1226 server.maxclients = 1024-32;
1227 } else {
1228 rlim_t oldlimit = limit.rlim_cur;
1229
1230 /* Set the max number of files if the current limit is not enough
1231 * for our needs. */
1232 if (oldlimit < maxfiles) {
1233 rlim_t f;
1234
1235 f = maxfiles;
1236 while(f > oldlimit) {
1237 limit.rlim_cur = f;
1238 limit.rlim_max = f;
1239 if (setrlimit(RLIMIT_NOFILE,&limit) != -1) break;
1240 f -= 128;
1241 }
1242 if (f < oldlimit) f = oldlimit;
1243 if (f != maxfiles) {
1244 server.maxclients = f-32;
1245 redisLog(REDIS_WARNING,"Unable to set the max number of files limit to %d (%s), setting the max clients configuration to %d.",
1246 (int) maxfiles, strerror(errno), (int) server.maxclients);
1247 } else {
1248 redisLog(REDIS_NOTICE,"Max number of open files set to %d",
1249 (int) maxfiles);
1250 }
1251 }
1252 }
1253 }
1254
1255 void initServer() {
1256 int j;
1257
1258 signal(SIGHUP, SIG_IGN);
1259 signal(SIGPIPE, SIG_IGN);
1260 setupSignalHandlers();
1261
1262 if (server.syslog_enabled) {
1263 openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
1264 server.syslog_facility);
1265 }
1266
1267 server.current_client = NULL;
1268 server.clients = listCreate();
1269 server.clients_to_close = listCreate();
1270 server.slaves = listCreate();
1271 server.monitors = listCreate();
1272 server.unblocked_clients = listCreate();
1273 server.ready_keys = listCreate();
1274
1275 createSharedObjects();
1276 adjustOpenFilesLimit();
1277 server.el = aeCreateEventLoop(server.maxclients+1024);
1278 server.db = zmalloc(sizeof(redisDb)*server.dbnum);
1279
1280 if (server.port != 0) {
1281 server.ipfd = anetTcpServer(server.neterr,server.port,server.bindaddr);
1282 if (server.ipfd == ANET_ERR) {
1283 redisLog(REDIS_WARNING, "Opening port %d: %s",
1284 server.port, server.neterr);
1285 exit(1);
1286 }
1287 }
1288 if (server.unixsocket != NULL) {
1289 unlink(server.unixsocket); /* don't care if this fails */
1290 server.sofd = anetUnixServer(server.neterr,server.unixsocket,server.unixsocketperm);
1291 if (server.sofd == ANET_ERR) {
1292 redisLog(REDIS_WARNING, "Opening socket: %s", server.neterr);
1293 exit(1);
1294 }
1295 }
1296 if (server.ipfd < 0 && server.sofd < 0) {
1297 redisLog(REDIS_WARNING, "Configured to not listen anywhere, exiting.");
1298 exit(1);
1299 }
1300 for (j = 0; j < server.dbnum; j++) {
1301 server.db[j].dict = dictCreate(&dbDictType,NULL);
1302 server.db[j].expires = dictCreate(&keyptrDictType,NULL);
1303 server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL);
1304 server.db[j].ready_keys = dictCreate(&setDictType,NULL);
1305 server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
1306 server.db[j].id = j;
1307 }
1308 server.pubsub_channels = dictCreate(&keylistDictType,NULL);
1309 server.pubsub_patterns = listCreate();
1310 listSetFreeMethod(server.pubsub_patterns,freePubsubPattern);
1311 listSetMatchMethod(server.pubsub_patterns,listMatchPubsubPattern);
1312 server.cronloops = 0;
1313 server.rdb_child_pid = -1;
1314 server.aof_child_pid = -1;
1315 aofRewriteBufferReset();
1316 server.aof_buf = sdsempty();
1317 server.lastsave = time(NULL);
1318 server.rdb_save_time_last = -1;
1319 server.rdb_save_time_start = -1;
1320 server.dirty = 0;
1321 server.stat_numcommands = 0;
1322 server.stat_numconnections = 0;
1323 server.stat_expiredkeys = 0;
1324 server.stat_evictedkeys = 0;
1325 server.stat_starttime = time(NULL);
1326 server.stat_keyspace_misses = 0;
1327 server.stat_keyspace_hits = 0;
1328 server.stat_peak_memory = 0;
1329 server.stat_fork_time = 0;
1330 server.stat_rejected_conn = 0;
1331 memset(server.ops_sec_samples,0,sizeof(server.ops_sec_samples));
1332 server.ops_sec_idx = 0;
1333 server.ops_sec_last_sample_time = mstime();
1334 server.ops_sec_last_sample_ops = 0;
1335 server.unixtime = time(NULL);
1336 server.lastbgsave_status = REDIS_OK;
1337 server.stop_writes_on_bgsave_err = 1;
1338 aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL);
1339 if (server.ipfd > 0 && aeCreateFileEvent(server.el,server.ipfd,AE_READABLE,
1340 acceptTcpHandler,NULL) == AE_ERR) redisPanic("Unrecoverable error creating server.ipfd file event.");
1341 if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
1342 acceptUnixHandler,NULL) == AE_ERR) redisPanic("Unrecoverable error creating server.sofd file event.");
1343
1344 if (server.aof_state == REDIS_AOF_ON) {
1345 server.aof_fd = open(server.aof_filename,
1346 O_WRONLY|O_APPEND|O_CREAT,0644);
1347 if (server.aof_fd == -1) {
1348 redisLog(REDIS_WARNING, "Can't open the append-only file: %s",
1349 strerror(errno));
1350 exit(1);
1351 }
1352 }
1353
1354 /* 32 bit instances are limited to 4GB of address space, so if there is
1355 * no explicit limit in the user provided configuration we set a limit
1356 * at 3 GB using maxmemory with 'noeviction' policy'. This avoids
1357 * useless crashes of the Redis instance for out of memory. */
1358 if (server.arch_bits == 32 && server.maxmemory == 0) {
1359 redisLog(REDIS_WARNING,"Warning: 32 bit instance detected but no memory limit set. Setting 3 GB maxmemory limit with 'noeviction' policy now.");
1360 server.maxmemory = 3072LL*(1024*1024); /* 3 GB */
1361 server.maxmemory_policy = REDIS_MAXMEMORY_NO_EVICTION;
1362 }
1363
1364 if (server.cluster_enabled) clusterInit();
1365 scriptingInit();
1366 slowlogInit();
1367 bioInit();
1368 }
1369
1370 /* Populates the Redis Command Table starting from the hard coded list
1371 * we have on top of redis.c file. */
1372 void populateCommandTable(void) {
1373 int j;
1374 int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1375
1376 for (j = 0; j < numcommands; j++) {
1377 struct redisCommand *c = redisCommandTable+j;
1378 char *f = c->sflags;
1379 int retval;
1380
1381 while(*f != '\0') {
1382 switch(*f) {
1383 case 'w': c->flags |= REDIS_CMD_WRITE; break;
1384 case 'r': c->flags |= REDIS_CMD_READONLY; break;
1385 case 'm': c->flags |= REDIS_CMD_DENYOOM; break;
1386 case 'a': c->flags |= REDIS_CMD_ADMIN; break;
1387 case 'p': c->flags |= REDIS_CMD_PUBSUB; break;
1388 case 'f': c->flags |= REDIS_CMD_FORCE_REPLICATION; break;
1389 case 's': c->flags |= REDIS_CMD_NOSCRIPT; break;
1390 case 'R': c->flags |= REDIS_CMD_RANDOM; break;
1391 case 'S': c->flags |= REDIS_CMD_SORT_FOR_SCRIPT; break;
1392 case 'l': c->flags |= REDIS_CMD_LOADING; break;
1393 case 't': c->flags |= REDIS_CMD_STALE; break;
1394 case 'M': c->flags |= REDIS_CMD_SKIP_MONITOR; break;
1395 default: redisPanic("Unsupported command flag"); break;
1396 }
1397 f++;
1398 }
1399
1400 retval = dictAdd(server.commands, sdsnew(c->name), c);
1401 assert(retval == DICT_OK);
1402 }
1403 }
1404
1405 void resetCommandTableStats(void) {
1406 int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1407 int j;
1408
1409 for (j = 0; j < numcommands; j++) {
1410 struct redisCommand *c = redisCommandTable+j;
1411
1412 c->microseconds = 0;
1413 c->calls = 0;
1414 }
1415 }
1416
1417 /* ========================== Redis OP Array API ============================ */
1418
1419 void redisOpArrayInit(redisOpArray *oa) {
1420 oa->ops = NULL;
1421 oa->numops = 0;
1422 }
1423
1424 int redisOpArrayAppend(redisOpArray *oa, struct redisCommand *cmd, int dbid,
1425 robj **argv, int argc, int target)
1426 {
1427 redisOp *op;
1428
1429 oa->ops = zrealloc(oa->ops,sizeof(redisOp)*(oa->numops+1));
1430 op = oa->ops+oa->numops;
1431 op->cmd = cmd;
1432 op->dbid = dbid;
1433 op->argv = argv;
1434 op->argc = argc;
1435 op->target = target;
1436 oa->numops++;
1437 return oa->numops;
1438 }
1439
1440 void redisOpArrayFree(redisOpArray *oa) {
1441 while(oa->numops) {
1442 int j;
1443 redisOp *op;
1444
1445 oa->numops--;
1446 op = oa->ops+oa->numops;
1447 for (j = 0; j < op->argc; j++)
1448 decrRefCount(op->argv[j]);
1449 zfree(op->argv);
1450 }
1451 zfree(oa->ops);
1452 }
1453
1454 /* ====================== Commands lookup and execution ===================== */
1455
1456 struct redisCommand *lookupCommand(sds name) {
1457 return dictFetchValue(server.commands, name);
1458 }
1459
1460 struct redisCommand *lookupCommandByCString(char *s) {
1461 struct redisCommand *cmd;
1462 sds name = sdsnew(s);
1463
1464 cmd = dictFetchValue(server.commands, name);
1465 sdsfree(name);
1466 return cmd;
1467 }
1468
1469 /* Propagate the specified command (in the context of the specified database id)
1470 * to AOF and Slaves.
1471 *
1472 * flags are an xor between:
1473 * + REDIS_PROPAGATE_NONE (no propagation of command at all)
1474 * + REDIS_PROPAGATE_AOF (propagate into the AOF file if is enabled)
1475 * + REDIS_PROPAGATE_REPL (propagate into the replication link)
1476 */
1477 void propagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
1478 int flags)
1479 {
1480 if (server.aof_state != REDIS_AOF_OFF && flags & REDIS_PROPAGATE_AOF)
1481 feedAppendOnlyFile(cmd,dbid,argv,argc);
1482 if (flags & REDIS_PROPAGATE_REPL && listLength(server.slaves))
1483 replicationFeedSlaves(server.slaves,dbid,argv,argc);
1484 }
1485
1486 /* Used inside commands to schedule the propagation of additional commands
1487 * after the current command is propagated to AOF / Replication. */
1488 void alsoPropagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
1489 int target)
1490 {
1491 redisOpArrayAppend(&server.also_propagate,cmd,dbid,argv,argc,target);
1492 }
1493
1494 /* Call() is the core of Redis execution of a command */
1495 void call(redisClient *c, int flags) {
1496 long long dirty, start = ustime(), duration;
1497
1498 /* Sent the command to clients in MONITOR mode, only if the commands are
1499 * not geneated from reading an AOF. */
1500 if (listLength(server.monitors) &&
1501 !server.loading &&
1502 !(c->cmd->flags & REDIS_CMD_SKIP_MONITOR))
1503 {
1504 replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
1505 }
1506
1507 /* Call the command. */
1508 redisOpArrayInit(&server.also_propagate);
1509 dirty = server.dirty;
1510 c->cmd->proc(c);
1511 dirty = server.dirty-dirty;
1512 duration = ustime()-start;
1513
1514 /* When EVAL is called loading the AOF we don't want commands called
1515 * from Lua to go into the slowlog or to populate statistics. */
1516 if (server.loading && c->flags & REDIS_LUA_CLIENT)
1517 flags &= ~(REDIS_CALL_SLOWLOG | REDIS_CALL_STATS);
1518
1519 /* Log the command into the Slow log if needed, and populate the
1520 * per-command statistics that we show in INFO commandstats. */
1521 if (flags & REDIS_CALL_SLOWLOG)
1522 slowlogPushEntryIfNeeded(c->argv,c->argc,duration);
1523 if (flags & REDIS_CALL_STATS) {
1524 c->cmd->microseconds += duration;
1525 c->cmd->calls++;
1526 }
1527
1528 /* Propagate the command into the AOF and replication link */
1529 if (flags & REDIS_CALL_PROPAGATE) {
1530 int flags = REDIS_PROPAGATE_NONE;
1531
1532 if (c->cmd->flags & REDIS_CMD_FORCE_REPLICATION)
1533 flags |= REDIS_PROPAGATE_REPL;
1534 if (dirty)
1535 flags |= (REDIS_PROPAGATE_REPL | REDIS_PROPAGATE_AOF);
1536 if (flags != REDIS_PROPAGATE_NONE)
1537 propagate(c->cmd,c->db->id,c->argv,c->argc,flags);
1538 }
1539 /* Commands such as LPUSH or BRPOPLPUSH may propagate an additional
1540 * PUSH command. */
1541 if (server.also_propagate.numops) {
1542 int j;
1543 redisOp *rop;
1544
1545 for (j = 0; j < server.also_propagate.numops; j++) {
1546 rop = &server.also_propagate.ops[j];
1547 propagate(rop->cmd, rop->dbid, rop->argv, rop->argc, rop->target);
1548 }
1549 redisOpArrayFree(&server.also_propagate);
1550 }
1551 server.stat_numcommands++;
1552 }
1553
1554 /* If this function gets called we already read a whole
1555 * command, arguments are in the client argv/argc fields.
1556 * processCommand() execute the command or prepare the
1557 * server for a bulk read from the client.
1558 *
1559 * If 1 is returned the client is still alive and valid and
1560 * and other operations can be performed by the caller. Otherwise
1561 * if 0 is returned the client was destroied (i.e. after QUIT). */
1562 int processCommand(redisClient *c) {
1563 /* The QUIT command is handled separately. Normal command procs will
1564 * go through checking for replication and QUIT will cause trouble
1565 * when FORCE_REPLICATION is enabled and would be implemented in
1566 * a regular command proc. */
1567 if (!strcasecmp(c->argv[0]->ptr,"quit")) {
1568 addReply(c,shared.ok);
1569 c->flags |= REDIS_CLOSE_AFTER_REPLY;
1570 return REDIS_ERR;
1571 }
1572
1573 /* Now lookup the command and check ASAP about trivial error conditions
1574 * such as wrong arity, bad command name and so forth. */
1575 c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
1576 if (!c->cmd) {
1577 addReplyErrorFormat(c,"unknown command '%s'",
1578 (char*)c->argv[0]->ptr);
1579 return REDIS_OK;
1580 } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) ||
1581 (c->argc < -c->cmd->arity)) {
1582 addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
1583 c->cmd->name);
1584 return REDIS_OK;
1585 }
1586
1587 /* Check if the user is authenticated */
1588 if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand)
1589 {
1590 addReplyError(c,"operation not permitted");
1591 return REDIS_OK;
1592 }
1593
1594 /* If cluster is enabled, redirect here */
1595 if (server.cluster_enabled &&
1596 !(c->cmd->getkeys_proc == NULL && c->cmd->firstkey == 0)) {
1597 int hashslot;
1598
1599 if (server.cluster.state != REDIS_CLUSTER_OK) {
1600 addReplyError(c,"The cluster is down. Check with CLUSTER INFO for more information");
1601 return REDIS_OK;
1602 } else {
1603 int ask;
1604 clusterNode *n = getNodeByQuery(c,c->cmd,c->argv,c->argc,&hashslot,&ask);
1605 if (n == NULL) {
1606 addReplyError(c,"Multi keys request invalid in cluster");
1607 return REDIS_OK;
1608 } else if (n != server.cluster.myself) {
1609 addReplySds(c,sdscatprintf(sdsempty(),
1610 "-%s %d %s:%d\r\n", ask ? "ASK" : "MOVED",
1611 hashslot,n->ip,n->port));
1612 return REDIS_OK;
1613 }
1614 }
1615 }
1616
1617 /* Handle the maxmemory directive.
1618 *
1619 * First we try to free some memory if possible (if there are volatile
1620 * keys in the dataset). If there are not the only thing we can do
1621 * is returning an error. */
1622 if (server.maxmemory) {
1623 int retval = freeMemoryIfNeeded();
1624 if ((c->cmd->flags & REDIS_CMD_DENYOOM) && retval == REDIS_ERR) {
1625 addReply(c, shared.oomerr);
1626 return REDIS_OK;
1627 }
1628 }
1629
1630 /* Don't accept write commands if there are problems persisting on disk. */
1631 if (server.stop_writes_on_bgsave_err &&
1632 server.saveparamslen > 0
1633 && server.lastbgsave_status == REDIS_ERR &&
1634 c->cmd->flags & REDIS_CMD_WRITE)
1635 {
1636 addReply(c, shared.bgsaveerr);
1637 return REDIS_OK;
1638 }
1639
1640 /* Don't accept write commands if this is a read only slave. But
1641 * accept write commands if this is our master. */
1642 if (server.masterhost && server.repl_slave_ro &&
1643 !(c->flags & REDIS_MASTER) &&
1644 c->cmd->flags & REDIS_CMD_WRITE)
1645 {
1646 addReply(c, shared.roslaveerr);
1647 return REDIS_OK;
1648 }
1649
1650 /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */
1651 if ((dictSize(c->pubsub_channels) > 0 || listLength(c->pubsub_patterns) > 0)
1652 &&
1653 c->cmd->proc != subscribeCommand &&
1654 c->cmd->proc != unsubscribeCommand &&
1655 c->cmd->proc != psubscribeCommand &&
1656 c->cmd->proc != punsubscribeCommand) {
1657 addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / QUIT allowed in this context");
1658 return REDIS_OK;
1659 }
1660
1661 /* Only allow INFO and SLAVEOF when slave-serve-stale-data is no and
1662 * we are a slave with a broken link with master. */
1663 if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED &&
1664 server.repl_serve_stale_data == 0 &&
1665 !(c->cmd->flags & REDIS_CMD_STALE))
1666 {
1667 addReply(c, shared.masterdownerr);
1668 return REDIS_OK;
1669 }
1670
1671 /* Loading DB? Return an error if the command has not the
1672 * REDIS_CMD_LOADING flag. */
1673 if (server.loading && !(c->cmd->flags & REDIS_CMD_LOADING)) {
1674 addReply(c, shared.loadingerr);
1675 return REDIS_OK;
1676 }
1677
1678 /* Lua script too slow? Only allow commands with REDIS_CMD_STALE flag. */
1679 if (server.lua_timedout &&
1680 c->cmd->proc != authCommand &&
1681 !(c->cmd->proc == shutdownCommand &&
1682 c->argc == 2 &&
1683 tolower(((char*)c->argv[1]->ptr)[0]) == 'n') &&
1684 !(c->cmd->proc == scriptCommand &&
1685 c->argc == 2 &&
1686 tolower(((char*)c->argv[1]->ptr)[0]) == 'k'))
1687 {
1688 addReply(c, shared.slowscripterr);
1689 return REDIS_OK;
1690 }
1691
1692 /* Exec the command */
1693 if (c->flags & REDIS_MULTI &&
1694 c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
1695 c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
1696 {
1697 queueMultiCommand(c);
1698 addReply(c,shared.queued);
1699 } else {
1700 call(c,REDIS_CALL_FULL);
1701 if (listLength(server.ready_keys))
1702 handleClientsBlockedOnLists();
1703 }
1704 return REDIS_OK;
1705 }
1706
1707 /*================================== Shutdown =============================== */
1708
1709 int prepareForShutdown(int flags) {
1710 int save = flags & REDIS_SHUTDOWN_SAVE;
1711 int nosave = flags & REDIS_SHUTDOWN_NOSAVE;
1712
1713 redisLog(REDIS_WARNING,"User requested shutdown...");
1714 /* Kill the saving child if there is a background saving in progress.
1715 We want to avoid race conditions, for instance our saving child may
1716 overwrite the synchronous saving did by SHUTDOWN. */
1717 if (server.rdb_child_pid != -1) {
1718 redisLog(REDIS_WARNING,"There is a child saving an .rdb. Killing it!");
1719 kill(server.rdb_child_pid,SIGKILL);
1720 rdbRemoveTempFile(server.rdb_child_pid);
1721 }
1722 if (server.aof_state != REDIS_AOF_OFF) {
1723 /* Kill the AOF saving child as the AOF we already have may be longer
1724 * but contains the full dataset anyway. */
1725 if (server.aof_child_pid != -1) {
1726 redisLog(REDIS_WARNING,
1727 "There is a child rewriting the AOF. Killing it!");
1728 kill(server.aof_child_pid,SIGKILL);
1729 }
1730 /* Append only file: fsync() the AOF and exit */
1731 redisLog(REDIS_NOTICE,"Calling fsync() on the AOF file.");
1732 aof_fsync(server.aof_fd);
1733 }
1734 if ((server.saveparamslen > 0 && !nosave) || save) {
1735 redisLog(REDIS_NOTICE,"Saving the final RDB snapshot before exiting.");
1736 /* Snapshotting. Perform a SYNC SAVE and exit */
1737 if (rdbSave(server.rdb_filename) != REDIS_OK) {
1738 /* Ooops.. error saving! The best we can do is to continue
1739 * operating. Note that if there was a background saving process,
1740 * in the next cron() Redis will be notified that the background
1741 * saving aborted, handling special stuff like slaves pending for
1742 * synchronization... */
1743 redisLog(REDIS_WARNING,"Error trying to save the DB, can't exit.");
1744 return REDIS_ERR;
1745 }
1746 }
1747 if (server.daemonize) {
1748 redisLog(REDIS_NOTICE,"Removing the pid file.");
1749 unlink(server.pidfile);
1750 }
1751 /* Close the listening sockets. Apparently this allows faster restarts. */
1752 if (server.ipfd != -1) close(server.ipfd);
1753 if (server.sofd != -1) close(server.sofd);
1754 if (server.unixsocket) {
1755 redisLog(REDIS_NOTICE,"Removing the unix socket file.");
1756 unlink(server.unixsocket); /* don't care if this fails */
1757 }
1758
1759 redisLog(REDIS_WARNING,"Redis is now ready to exit, bye bye...");
1760 return REDIS_OK;
1761 }
1762
1763 /*================================== Commands =============================== */
1764
1765 /* Return zero if strings are the same, non-zero if they are not.
1766 * The comparison is performed in a way that prevents an attacker to obtain
1767 * information about the nature of the strings just monitoring the execution
1768 * time of the function.
1769 *
1770 * Note that limiting the comparison length to strings up to 512 bytes we
1771 * can avoid leaking any information about the password length and any
1772 * possible branch misprediction related leak.
1773 */
1774 int time_independent_strcmp(char *a, char *b) {
1775 char bufa[REDIS_AUTHPASS_MAX_LEN], bufb[REDIS_AUTHPASS_MAX_LEN];
1776 /* The above two strlen perform len(a) + len(b) operations where either
1777 * a or b are fixed (our password) length, and the difference is only
1778 * relative to the length of the user provided string, so no information
1779 * leak is possible in the following two lines of code. */
1780 int alen = strlen(a);
1781 int blen = strlen(b);
1782 int j;
1783 int diff = 0;
1784
1785 /* We can't compare strings longer than our static buffers.
1786 * Note that this will never pass the first test in practical circumstances
1787 * so there is no info leak. */
1788 if (alen > sizeof(bufa) || blen > sizeof(bufb)) return 1;
1789
1790 memset(bufa,0,sizeof(bufa)); /* Constant time. */
1791 memset(bufb,0,sizeof(bufb)); /* Constant time. */
1792 /* Again the time of the following two copies is proportional to
1793 * len(a) + len(b) so no info is leaked. */
1794 memcpy(bufa,a,alen);
1795 memcpy(bufb,b,blen);
1796
1797 /* Always compare all the chars in the two buffers without
1798 * conditional expressions. */
1799 for (j = 0; j < sizeof(bufa); j++) {
1800 diff |= (bufa[j] ^ bufb[j]);
1801 }
1802 /* Length must be equal as well. */
1803 diff |= alen ^ blen;
1804 return diff; /* If zero strings are the same. */
1805 }
1806
1807 void authCommand(redisClient *c) {
1808 if (!server.requirepass) {
1809 addReplyError(c,"Client sent AUTH, but no password is set");
1810 } else if (!time_independent_strcmp(c->argv[1]->ptr, server.requirepass)) {
1811 c->authenticated = 1;
1812 addReply(c,shared.ok);
1813 } else {
1814 c->authenticated = 0;
1815 addReplyError(c,"invalid password");
1816 }
1817 }
1818
1819 void pingCommand(redisClient *c) {
1820 addReply(c,shared.pong);
1821 }
1822
1823 void echoCommand(redisClient *c) {
1824 addReplyBulk(c,c->argv[1]);
1825 }
1826
1827 void timeCommand(redisClient *c) {
1828 struct timeval tv;
1829
1830 /* gettimeofday() can only fail if &tv is a bad addresss so we
1831 * don't check for errors. */
1832 gettimeofday(&tv,NULL);
1833 addReplyMultiBulkLen(c,2);
1834 addReplyBulkLongLong(c,tv.tv_sec);
1835 addReplyBulkLongLong(c,tv.tv_usec);
1836 }
1837
1838 /* Convert an amount of bytes into a human readable string in the form
1839 * of 100B, 2G, 100M, 4K, and so forth. */
1840 void bytesToHuman(char *s, unsigned long long n) {
1841 double d;
1842
1843 if (n < 1024) {
1844 /* Bytes */
1845 sprintf(s,"%lluB",n);
1846 return;
1847 } else if (n < (1024*1024)) {
1848 d = (double)n/(1024);
1849 sprintf(s,"%.2fK",d);
1850 } else if (n < (1024LL*1024*1024)) {
1851 d = (double)n/(1024*1024);
1852 sprintf(s,"%.2fM",d);
1853 } else if (n < (1024LL*1024*1024*1024)) {
1854 d = (double)n/(1024LL*1024*1024);
1855 sprintf(s,"%.2fG",d);
1856 }
1857 }
1858
1859 /* Create the string returned by the INFO command. This is decoupled
1860 * by the INFO command itself as we need to report the same information
1861 * on memory corruption problems. */
1862 sds genRedisInfoString(char *section) {
1863 sds info = sdsempty();
1864 time_t uptime = server.unixtime-server.stat_starttime;
1865 int j, numcommands;
1866 struct rusage self_ru, c_ru;
1867 unsigned long lol, bib;
1868 int allsections = 0, defsections = 0;
1869 int sections = 0;
1870
1871 if (section) {
1872 allsections = strcasecmp(section,"all") == 0;
1873 defsections = strcasecmp(section,"default") == 0;
1874 }
1875
1876 getrusage(RUSAGE_SELF, &self_ru);
1877 getrusage(RUSAGE_CHILDREN, &c_ru);
1878 getClientsMaxBuffers(&lol,&bib);
1879
1880 /* Server */
1881 if (allsections || defsections || !strcasecmp(section,"server")) {
1882 struct utsname name;
1883 char *mode;
1884
1885 if (server.cluster_enabled) mode = "cluster";
1886 else if (server.sentinel_mode) mode = "sentinel";
1887 else mode = "standalone";
1888
1889 if (sections++) info = sdscat(info,"\r\n");
1890 uname(&name);
1891 info = sdscatprintf(info,
1892 "# Server\r\n"
1893 "redis_version:%s\r\n"
1894 "redis_git_sha1:%s\r\n"
1895 "redis_git_dirty:%d\r\n"
1896 "redis_mode:%s\r\n"
1897 "os:%s %s %s\r\n"
1898 "arch_bits:%d\r\n"
1899 "multiplexing_api:%s\r\n"
1900 "gcc_version:%d.%d.%d\r\n"
1901 "process_id:%ld\r\n"
1902 "run_id:%s\r\n"
1903 "tcp_port:%d\r\n"
1904 "uptime_in_seconds:%ld\r\n"
1905 "uptime_in_days:%ld\r\n"
1906 "lru_clock:%ld\r\n",
1907 REDIS_VERSION,
1908 redisGitSHA1(),
1909 strtol(redisGitDirty(),NULL,10) > 0,
1910 mode,
1911 name.sysname, name.release, name.machine,
1912 server.arch_bits,
1913 aeGetApiName(),
1914 #ifdef __GNUC__
1915 __GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__,
1916 #else
1917 0,0,0,
1918 #endif
1919 (long) getpid(),
1920 server.runid,
1921 server.port,
1922 uptime,
1923 uptime/(3600*24),
1924 (unsigned long) server.lruclock);
1925 }
1926
1927 /* Clients */
1928 if (allsections || defsections || !strcasecmp(section,"clients")) {
1929 if (sections++) info = sdscat(info,"\r\n");
1930 info = sdscatprintf(info,
1931 "# Clients\r\n"
1932 "connected_clients:%lu\r\n"
1933 "client_longest_output_list:%lu\r\n"
1934 "client_biggest_input_buf:%lu\r\n"
1935 "blocked_clients:%d\r\n",
1936 listLength(server.clients)-listLength(server.slaves),
1937 lol, bib,
1938 server.bpop_blocked_clients);
1939 }
1940
1941 /* Memory */
1942 if (allsections || defsections || !strcasecmp(section,"memory")) {
1943 char hmem[64];
1944 char peak_hmem[64];
1945
1946 bytesToHuman(hmem,zmalloc_used_memory());
1947 bytesToHuman(peak_hmem,server.stat_peak_memory);
1948 if (sections++) info = sdscat(info,"\r\n");
1949 info = sdscatprintf(info,
1950 "# Memory\r\n"
1951 "used_memory:%zu\r\n"
1952 "used_memory_human:%s\r\n"
1953 "used_memory_rss:%zu\r\n"
1954 "used_memory_peak:%zu\r\n"
1955 "used_memory_peak_human:%s\r\n"
1956 "used_memory_lua:%lld\r\n"
1957 "mem_fragmentation_ratio:%.2f\r\n"
1958 "mem_allocator:%s\r\n",
1959 zmalloc_used_memory(),
1960 hmem,
1961 zmalloc_get_rss(),
1962 server.stat_peak_memory,
1963 peak_hmem,
1964 ((long long)lua_gc(server.lua,LUA_GCCOUNT,0))*1024LL,
1965 zmalloc_get_fragmentation_ratio(),
1966 ZMALLOC_LIB
1967 );
1968 }
1969
1970 /* Persistence */
1971 if (allsections || defsections || !strcasecmp(section,"persistence")) {
1972 if (sections++) info = sdscat(info,"\r\n");
1973 info = sdscatprintf(info,
1974 "# Persistence\r\n"
1975 "loading:%d\r\n"
1976 "rdb_changes_since_last_save:%lld\r\n"
1977 "rdb_bgsave_in_progress:%d\r\n"
1978 "rdb_last_save_time:%ld\r\n"
1979 "rdb_last_bgsave_status:%s\r\n"
1980 "rdb_last_bgsave_time_sec:%ld\r\n"
1981 "rdb_current_bgsave_time_sec:%ld\r\n"
1982 "aof_enabled:%d\r\n"
1983 "aof_rewrite_in_progress:%d\r\n"
1984 "aof_rewrite_scheduled:%d\r\n"
1985 "aof_last_rewrite_time_sec:%ld\r\n"
1986 "aof_current_rewrite_time_sec:%ld\r\n"
1987 "aof_last_bgrewrite_status:%s\r\n",
1988 server.loading,
1989 server.dirty,
1990 server.rdb_child_pid != -1,
1991 server.lastsave,
1992 (server.lastbgsave_status == REDIS_OK) ? "ok" : "err",
1993 server.rdb_save_time_last,
1994 (server.rdb_child_pid == -1) ?
1995 -1 : time(NULL)-server.rdb_save_time_start,
1996 server.aof_state != REDIS_AOF_OFF,
1997 server.aof_child_pid != -1,
1998 server.aof_rewrite_scheduled,
1999 server.aof_rewrite_time_last,
2000 (server.aof_child_pid == -1) ?
2001 -1 : time(NULL)-server.aof_rewrite_time_start,
2002 (server.aof_lastbgrewrite_status == REDIS_OK) ? "ok" : "err");
2003
2004 if (server.aof_state != REDIS_AOF_OFF) {
2005 info = sdscatprintf(info,
2006 "aof_current_size:%lld\r\n"
2007 "aof_base_size:%lld\r\n"
2008 "aof_pending_rewrite:%d\r\n"
2009 "aof_buffer_length:%zu\r\n"
2010 "aof_rewrite_buffer_length:%lu\r\n"
2011 "aof_pending_bio_fsync:%llu\r\n"
2012 "aof_delayed_fsync:%lu\r\n",
2013 (long long) server.aof_current_size,
2014 (long long) server.aof_rewrite_base_size,
2015 server.aof_rewrite_scheduled,
2016 sdslen(server.aof_buf),
2017 aofRewriteBufferSize(),
2018 bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC),
2019 server.aof_delayed_fsync);
2020 }
2021
2022 if (server.loading) {
2023 double perc;
2024 time_t eta, elapsed;
2025 off_t remaining_bytes = server.loading_total_bytes-
2026 server.loading_loaded_bytes;
2027
2028 perc = ((double)server.loading_loaded_bytes /
2029 server.loading_total_bytes) * 100;
2030
2031 elapsed = server.unixtime-server.loading_start_time;
2032 if (elapsed == 0) {
2033 eta = 1; /* A fake 1 second figure if we don't have
2034 enough info */
2035 } else {
2036 eta = (elapsed*remaining_bytes)/server.loading_loaded_bytes;
2037 }
2038
2039 info = sdscatprintf(info,
2040 "loading_start_time:%ld\r\n"
2041 "loading_total_bytes:%llu\r\n"
2042 "loading_loaded_bytes:%llu\r\n"
2043 "loading_loaded_perc:%.2f\r\n"
2044 "loading_eta_seconds:%ld\r\n"
2045 ,(unsigned long) server.loading_start_time,
2046 (unsigned long long) server.loading_total_bytes,
2047 (unsigned long long) server.loading_loaded_bytes,
2048 perc,
2049 eta
2050 );
2051 }
2052 }
2053
2054 /* Stats */
2055 if (allsections || defsections || !strcasecmp(section,"stats")) {
2056 if (sections++) info = sdscat(info,"\r\n");
2057 info = sdscatprintf(info,
2058 "# Stats\r\n"
2059 "total_connections_received:%lld\r\n"
2060 "total_commands_processed:%lld\r\n"
2061 "instantaneous_ops_per_sec:%lld\r\n"
2062 "rejected_connections:%lld\r\n"
2063 "expired_keys:%lld\r\n"
2064 "evicted_keys:%lld\r\n"
2065 "keyspace_hits:%lld\r\n"
2066 "keyspace_misses:%lld\r\n"
2067 "pubsub_channels:%ld\r\n"
2068 "pubsub_patterns:%lu\r\n"
2069 "latest_fork_usec:%lld\r\n",
2070 server.stat_numconnections,
2071 server.stat_numcommands,
2072 getOperationsPerSecond(),
2073 server.stat_rejected_conn,
2074 server.stat_expiredkeys,
2075 server.stat_evictedkeys,
2076 server.stat_keyspace_hits,
2077 server.stat_keyspace_misses,
2078 dictSize(server.pubsub_channels),
2079 listLength(server.pubsub_patterns),
2080 server.stat_fork_time);
2081 }
2082
2083 /* Replication */
2084 if (allsections || defsections || !strcasecmp(section,"replication")) {
2085 if (sections++) info = sdscat(info,"\r\n");
2086 info = sdscatprintf(info,
2087 "# Replication\r\n"
2088 "role:%s\r\n",
2089 server.masterhost == NULL ? "master" : "slave");
2090 if (server.masterhost) {
2091 info = sdscatprintf(info,
2092 "master_host:%s\r\n"
2093 "master_port:%d\r\n"
2094 "master_link_status:%s\r\n"
2095 "master_last_io_seconds_ago:%d\r\n"
2096 "master_sync_in_progress:%d\r\n"
2097 ,server.masterhost,
2098 server.masterport,
2099 (server.repl_state == REDIS_REPL_CONNECTED) ?
2100 "up" : "down",
2101 server.master ?
2102 ((int)(server.unixtime-server.master->lastinteraction)) : -1,
2103 server.repl_state == REDIS_REPL_TRANSFER
2104 );
2105
2106 if (server.repl_state == REDIS_REPL_TRANSFER) {
2107 info = sdscatprintf(info,
2108 "master_sync_left_bytes:%lld\r\n"
2109 "master_sync_last_io_seconds_ago:%d\r\n"
2110 , (long long)
2111 (server.repl_transfer_size - server.repl_transfer_read),
2112 (int)(server.unixtime-server.repl_transfer_lastio)
2113 );
2114 }
2115
2116 if (server.repl_state != REDIS_REPL_CONNECTED) {
2117 info = sdscatprintf(info,
2118 "master_link_down_since_seconds:%ld\r\n",
2119 (long)server.unixtime-server.repl_down_since);
2120 }
2121 info = sdscatprintf(info,
2122 "slave_priority:%d\r\n"
2123 "slave_read_only:%d\r\n",
2124 server.slave_priority,
2125 server.repl_slave_ro);
2126 }
2127 info = sdscatprintf(info,
2128 "connected_slaves:%lu\r\n",
2129 listLength(server.slaves));
2130 if (listLength(server.slaves)) {
2131 int slaveid = 0;
2132 listNode *ln;
2133 listIter li;
2134
2135 listRewind(server.slaves,&li);
2136 while((ln = listNext(&li))) {
2137 redisClient *slave = listNodeValue(ln);
2138 char *state = NULL;
2139 char ip[32];
2140 int port;
2141
2142 if (anetPeerToString(slave->fd,ip,&port) == -1) continue;
2143 switch(slave->replstate) {
2144 case REDIS_REPL_WAIT_BGSAVE_START:
2145 case REDIS_REPL_WAIT_BGSAVE_END:
2146 state = "wait_bgsave";
2147 break;
2148 case REDIS_REPL_SEND_BULK:
2149 state = "send_bulk";
2150 break;
2151 case REDIS_REPL_ONLINE:
2152 state = "online";
2153 break;
2154 }
2155 if (state == NULL) continue;
2156 info = sdscatprintf(info,"slave%d:%s,%d,%s\r\n",
2157 slaveid,ip,slave->slave_listening_port,state);
2158 slaveid++;
2159 }
2160 }
2161 }
2162
2163 /* CPU */
2164 if (allsections || defsections || !strcasecmp(section,"cpu")) {
2165 if (sections++) info = sdscat(info,"\r\n");
2166 info = sdscatprintf(info,
2167 "# CPU\r\n"
2168 "used_cpu_sys:%.2f\r\n"
2169 "used_cpu_user:%.2f\r\n"
2170 "used_cpu_sys_children:%.2f\r\n"
2171 "used_cpu_user_children:%.2f\r\n",
2172 (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000,
2173 (float)self_ru.ru_utime.tv_sec+(float)self_ru.ru_utime.tv_usec/1000000,
2174 (float)c_ru.ru_stime.tv_sec+(float)c_ru.ru_stime.tv_usec/1000000,
2175 (float)c_ru.ru_utime.tv_sec+(float)c_ru.ru_utime.tv_usec/1000000);
2176 }
2177
2178 /* cmdtime */
2179 if (allsections || !strcasecmp(section,"commandstats")) {
2180 if (sections++) info = sdscat(info,"\r\n");
2181 info = sdscatprintf(info, "# Commandstats\r\n");
2182 numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
2183 for (j = 0; j < numcommands; j++) {
2184 struct redisCommand *c = redisCommandTable+j;
2185
2186 if (!c->calls) continue;
2187 info = sdscatprintf(info,
2188 "cmdstat_%s:calls=%lld,usec=%lld,usec_per_call=%.2f\r\n",
2189 c->name, c->calls, c->microseconds,
2190 (c->calls == 0) ? 0 : ((float)c->microseconds/c->calls));
2191 }
2192 }
2193
2194 /* Cluster */
2195 if (allsections || defsections || !strcasecmp(section,"cluster")) {
2196 if (sections++) info = sdscat(info,"\r\n");
2197 info = sdscatprintf(info,
2198 "# Cluster\r\n"
2199 "cluster_enabled:%d\r\n",
2200 server.cluster_enabled);
2201 }
2202
2203 /* Key space */
2204 if (allsections || defsections || !strcasecmp(section,"keyspace")) {
2205 if (sections++) info = sdscat(info,"\r\n");
2206 info = sdscatprintf(info, "# Keyspace\r\n");
2207 for (j = 0; j < server.dbnum; j++) {
2208 long long keys, vkeys;
2209
2210 keys = dictSize(server.db[j].dict);
2211 vkeys = dictSize(server.db[j].expires);
2212 if (keys || vkeys) {
2213 info = sdscatprintf(info, "db%d:keys=%lld,expires=%lld\r\n",
2214 j, keys, vkeys);
2215 }
2216 }
2217 }
2218 return info;
2219 }
2220
2221 void infoCommand(redisClient *c) {
2222 char *section = c->argc == 2 ? c->argv[1]->ptr : "default";
2223
2224 if (c->argc > 2) {
2225 addReply(c,shared.syntaxerr);
2226 return;
2227 }
2228 sds info = genRedisInfoString(section);
2229 addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
2230 (unsigned long)sdslen(info)));
2231 addReplySds(c,info);
2232 addReply(c,shared.crlf);
2233 }
2234
2235 void monitorCommand(redisClient *c) {
2236 /* ignore MONITOR if already slave or in monitor mode */
2237 if (c->flags & REDIS_SLAVE) return;
2238
2239 c->flags |= (REDIS_SLAVE|REDIS_MONITOR);
2240 c->slaveseldb = 0;
2241 listAddNodeTail(server.monitors,c);
2242 addReply(c,shared.ok);
2243 }
2244
2245 /* ============================ Maxmemory directive ======================== */
2246
2247 /* This function gets called when 'maxmemory' is set on the config file to limit
2248 * the max memory used by the server, before processing a command.
2249 *
2250 * The goal of the function is to free enough memory to keep Redis under the
2251 * configured memory limit.
2252 *
2253 * The function starts calculating how many bytes should be freed to keep
2254 * Redis under the limit, and enters a loop selecting the best keys to
2255 * evict accordingly to the configured policy.
2256 *
2257 * If all the bytes needed to return back under the limit were freed the
2258 * function returns REDIS_OK, otherwise REDIS_ERR is returned, and the caller
2259 * should block the execution of commands that will result in more memory
2260 * used by the server.
2261 */
2262 int freeMemoryIfNeeded(void) {
2263 size_t mem_used, mem_tofree, mem_freed;
2264 int slaves = listLength(server.slaves);
2265
2266 /* Remove the size of slaves output buffers and AOF buffer from the
2267 * count of used memory. */
2268 mem_used = zmalloc_used_memory();
2269 if (slaves) {
2270 listIter li;
2271 listNode *ln;
2272
2273 listRewind(server.slaves,&li);
2274 while((ln = listNext(&li))) {
2275 redisClient *slave = listNodeValue(ln);
2276 unsigned long obuf_bytes = getClientOutputBufferMemoryUsage(slave);
2277 if (obuf_bytes > mem_used)
2278 mem_used = 0;
2279 else
2280 mem_used -= obuf_bytes;
2281 }
2282 }
2283 if (server.aof_state != REDIS_AOF_OFF) {
2284 mem_used -= sdslen(server.aof_buf);
2285 mem_used -= aofRewriteBufferSize();
2286 }
2287
2288 /* Check if we are over the memory limit. */
2289 if (mem_used <= server.maxmemory) return REDIS_OK;
2290
2291 if (server.maxmemory_policy == REDIS_MAXMEMORY_NO_EVICTION)
2292 return REDIS_ERR; /* We need to free memory, but policy forbids. */
2293
2294 /* Compute how much memory we need to free. */
2295 mem_tofree = mem_used - server.maxmemory;
2296 mem_freed = 0;
2297 while (mem_freed < mem_tofree) {
2298 int j, k, keys_freed = 0;
2299
2300 for (j = 0; j < server.dbnum; j++) {
2301 long bestval = 0; /* just to prevent warning */
2302 sds bestkey = NULL;
2303 struct dictEntry *de;
2304 redisDb *db = server.db+j;
2305 dict *dict;
2306
2307 if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
2308 server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM)
2309 {
2310 dict = server.db[j].dict;
2311 } else {
2312 dict = server.db[j].expires;
2313 }
2314 if (dictSize(dict) == 0) continue;
2315
2316 /* volatile-random and allkeys-random policy */
2317 if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM ||
2318 server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_RANDOM)
2319 {
2320 de = dictGetRandomKey(dict);
2321 bestkey = dictGetKey(de);
2322 }
2323
2324 /* volatile-lru and allkeys-lru policy */
2325 else if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
2326 server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
2327 {
2328 for (k = 0; k < server.maxmemory_samples; k++) {
2329 sds thiskey;
2330 long thisval;
2331 robj *o;
2332
2333 de = dictGetRandomKey(dict);
2334 thiskey = dictGetKey(de);
2335 /* When policy is volatile-lru we need an additional lookup
2336 * to locate the real key, as dict is set to db->expires. */
2337 if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU)
2338 de = dictFind(db->dict, thiskey);
2339 o = dictGetVal(de);
2340 thisval = estimateObjectIdleTime(o);
2341
2342 /* Higher idle time is better candidate for deletion */
2343 if (bestkey == NULL || thisval > bestval) {
2344 bestkey = thiskey;
2345 bestval = thisval;
2346 }
2347 }
2348 }
2349
2350 /* volatile-ttl */
2351 else if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_TTL) {
2352 for (k = 0; k < server.maxmemory_samples; k++) {
2353 sds thiskey;
2354 long thisval;
2355
2356 de = dictGetRandomKey(dict);
2357 thiskey = dictGetKey(de);
2358 thisval = (long) dictGetVal(de);
2359
2360 /* Expire sooner (minor expire unix timestamp) is better
2361 * candidate for deletion */
2362 if (bestkey == NULL || thisval < bestval) {
2363 bestkey = thiskey;
2364 bestval = thisval;
2365 }
2366 }
2367 }
2368
2369 /* Finally remove the selected key. */
2370 if (bestkey) {
2371 long long delta;
2372
2373 robj *keyobj = createStringObject(bestkey,sdslen(bestkey));
2374 propagateExpire(db,keyobj);
2375 /* We compute the amount of memory freed by dbDelete() alone.
2376 * It is possible that actually the memory needed to propagate
2377 * the DEL in AOF and replication link is greater than the one
2378 * we are freeing removing the key, but we can't account for
2379 * that otherwise we would never exit the loop.
2380 *
2381 * AOF and Output buffer memory will be freed eventually so
2382 * we only care about memory used by the key space. */
2383 delta = (long long) zmalloc_used_memory();
2384 dbDelete(db,keyobj);
2385 delta -= (long long) zmalloc_used_memory();
2386 mem_freed += delta;
2387 server.stat_evictedkeys++;
2388 decrRefCount(keyobj);
2389 keys_freed++;
2390
2391 /* When the memory to free starts to be big enough, we may
2392 * start spending so much time here that is impossible to
2393 * deliver data to the slaves fast enough, so we force the
2394 * transmission here inside the loop. */
2395 if (slaves) flushSlavesOutputBuffers();
2396 }
2397 }
2398 if (!keys_freed) return REDIS_ERR; /* nothing to free... */
2399 }
2400 return REDIS_OK;
2401 }
2402
2403 /* =================================== Main! ================================ */
2404
2405 #ifdef __linux__
2406 int linuxOvercommitMemoryValue(void) {
2407 FILE *fp = fopen("/proc/sys/vm/overcommit_memory","r");
2408 char buf[64];
2409
2410 if (!fp) return -1;
2411 if (fgets(buf,64,fp) == NULL) {
2412 fclose(fp);
2413 return -1;
2414 }
2415 fclose(fp);
2416
2417 return atoi(buf);
2418 }
2419
2420 void linuxOvercommitMemoryWarning(void) {
2421 if (linuxOvercommitMemoryValue() == 0) {
2422 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.");
2423 }
2424 }
2425 #endif /* __linux__ */
2426
2427 void createPidFile(void) {
2428 /* Try to write the pid file in a best-effort way. */
2429 FILE *fp = fopen(server.pidfile,"w");
2430 if (fp) {
2431 fprintf(fp,"%d\n",(int)getpid());
2432 fclose(fp);
2433 }
2434 }
2435
2436 void daemonize(void) {
2437 int fd;
2438
2439 if (fork() != 0) exit(0); /* parent exits */
2440 setsid(); /* create a new session */
2441
2442 /* Every output goes to /dev/null. If Redis is daemonized but
2443 * the 'logfile' is set to 'stdout' in the configuration file
2444 * it will not log at all. */
2445 if ((fd = open("/dev/null", O_RDWR, 0)) != -1) {
2446 dup2(fd, STDIN_FILENO);
2447 dup2(fd, STDOUT_FILENO);
2448 dup2(fd, STDERR_FILENO);
2449 if (fd > STDERR_FILENO) close(fd);
2450 }
2451 }
2452
2453 void version() {
2454 printf("Redis server v=%s sha=%s:%d malloc=%s bits=%d\n",
2455 REDIS_VERSION,
2456 redisGitSHA1(),
2457 atoi(redisGitDirty()) > 0,
2458 ZMALLOC_LIB,
2459 sizeof(long) == 4 ? 32 : 64);
2460 exit(0);
2461 }
2462
2463 void usage() {
2464 fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf] [options]\n");
2465 fprintf(stderr," ./redis-server - (read config from stdin)\n");
2466 fprintf(stderr," ./redis-server -v or --version\n");
2467 fprintf(stderr," ./redis-server -h or --help\n");
2468 fprintf(stderr," ./redis-server --test-memory <megabytes>\n\n");
2469 fprintf(stderr,"Examples:\n");
2470 fprintf(stderr," ./redis-server (run the server with default conf)\n");
2471 fprintf(stderr," ./redis-server /etc/redis/6379.conf\n");
2472 fprintf(stderr," ./redis-server --port 7777\n");
2473 fprintf(stderr," ./redis-server --port 7777 --slaveof 127.0.0.1 8888\n");
2474 fprintf(stderr," ./redis-server /etc/myredis.conf --loglevel verbose\n\n");
2475 fprintf(stderr,"Sentinel mode:\n");
2476 fprintf(stderr," ./redis-server /etc/sentinel.conf --sentinel\n");
2477 exit(1);
2478 }
2479
2480 void redisAsciiArt(void) {
2481 #include "asciilogo.h"
2482 char *buf = zmalloc(1024*16);
2483 char *mode = "stand alone";
2484
2485 if (server.cluster_enabled) mode = "cluster";
2486 else if (server.sentinel_mode) mode = "sentinel";
2487
2488 snprintf(buf,1024*16,ascii_logo,
2489 REDIS_VERSION,
2490 redisGitSHA1(),
2491 strtol(redisGitDirty(),NULL,10) > 0,
2492 (sizeof(long) == 8) ? "64" : "32",
2493 mode, server.port,
2494 (long) getpid()
2495 );
2496 redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf);
2497 zfree(buf);
2498 }
2499
2500 static void sigtermHandler(int sig) {
2501 REDIS_NOTUSED(sig);
2502
2503 redisLogFromHandler(REDIS_WARNING,"Received SIGTERM, scheduling shutdown...");
2504 server.shutdown_asap = 1;
2505 }
2506
2507 void setupSignalHandlers(void) {
2508 struct sigaction act;
2509
2510 /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used.
2511 * Otherwise, sa_handler is used. */
2512 sigemptyset(&act.sa_mask);
2513 act.sa_flags = 0;
2514 act.sa_handler = sigtermHandler;
2515 sigaction(SIGTERM, &act, NULL);
2516
2517 #ifdef HAVE_BACKTRACE
2518 sigemptyset(&act.sa_mask);
2519 act.sa_flags = SA_NODEFER | SA_RESETHAND | SA_SIGINFO;
2520 act.sa_sigaction = sigsegvHandler;
2521 sigaction(SIGSEGV, &act, NULL);
2522 sigaction(SIGBUS, &act, NULL);
2523 sigaction(SIGFPE, &act, NULL);
2524 sigaction(SIGILL, &act, NULL);
2525 #endif
2526 return;
2527 }
2528
2529 void memtest(size_t megabytes, int passes);
2530
2531 /* Returns 1 if there is --sentinel among the arguments or if
2532 * argv[0] is exactly "redis-sentinel". */
2533 int checkForSentinelMode(int argc, char **argv) {
2534 int j;
2535
2536 if (strstr(argv[0],"redis-sentinel") != NULL) return 1;
2537 for (j = 1; j < argc; j++)
2538 if (!strcmp(argv[j],"--sentinel")) return 1;
2539 return 0;
2540 }
2541
2542 /* Function called at startup to load RDB or AOF file in memory. */
2543 void loadDataFromDisk(void) {
2544 long long start = ustime();
2545 if (server.aof_state == REDIS_AOF_ON) {
2546 if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
2547 redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
2548 } else {
2549 if (rdbLoad(server.rdb_filename) == REDIS_OK) {
2550 redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",
2551 (float)(ustime()-start)/1000000);
2552 } else if (errno != ENOENT) {
2553 redisLog(REDIS_WARNING,"Fatal error loading the DB. Exiting.");
2554 exit(1);
2555 }
2556 }
2557 }
2558
2559 void redisOutOfMemoryHandler(size_t allocation_size) {
2560 redisLog(REDIS_WARNING,"Out Of Memory allocating %zu bytes!",
2561 allocation_size);
2562 redisPanic("OOM");
2563 }
2564
2565 int main(int argc, char **argv) {
2566 struct timeval tv;
2567
2568 /* We need to initialize our libraries, and the server configuration. */
2569 zmalloc_enable_thread_safeness();
2570 zmalloc_set_oom_handler(redisOutOfMemoryHandler);
2571 srand(time(NULL)^getpid());
2572 gettimeofday(&tv,NULL);
2573 dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
2574 server.sentinel_mode = checkForSentinelMode(argc,argv);
2575 initServerConfig();
2576
2577 /* We need to init sentinel right now as parsing the configuration file
2578 * in sentinel mode will have the effect of populating the sentinel
2579 * data structures with master nodes to monitor. */
2580 if (server.sentinel_mode) {
2581 initSentinelConfig();
2582 initSentinel();
2583 }
2584
2585 if (argc >= 2) {
2586 int j = 1; /* First option to parse in argv[] */
2587 sds options = sdsempty();
2588 char *configfile = NULL;
2589
2590 /* Handle special options --help and --version */
2591 if (strcmp(argv[1], "-v") == 0 ||
2592 strcmp(argv[1], "--version") == 0) version();
2593 if (strcmp(argv[1], "--help") == 0 ||
2594 strcmp(argv[1], "-h") == 0) usage();
2595 if (strcmp(argv[1], "--test-memory") == 0) {
2596 if (argc == 3) {
2597 memtest(atoi(argv[2]),50);
2598 exit(0);
2599 } else {
2600 fprintf(stderr,"Please specify the amount of memory to test in megabytes.\n");
2601 fprintf(stderr,"Example: ./redis-server --test-memory 4096\n\n");
2602 exit(1);
2603 }
2604 }
2605
2606 /* First argument is the config file name? */
2607 if (argv[j][0] != '-' || argv[j][1] != '-')
2608 configfile = argv[j++];
2609 /* All the other options are parsed and conceptually appended to the
2610 * configuration file. For instance --port 6380 will generate the
2611 * string "port 6380\n" to be parsed after the actual file name
2612 * is parsed, if any. */
2613 while(j != argc) {
2614 if (argv[j][0] == '-' && argv[j][1] == '-') {
2615 /* Option name */
2616 if (sdslen(options)) options = sdscat(options,"\n");
2617 options = sdscat(options,argv[j]+2);
2618 options = sdscat(options," ");
2619 } else {
2620 /* Option argument */
2621 options = sdscatrepr(options,argv[j],strlen(argv[j]));
2622 options = sdscat(options," ");
2623 }
2624 j++;
2625 }
2626 resetServerSaveParams();
2627 loadServerConfig(configfile,options);
2628 sdsfree(options);
2629 } else {
2630 redisLog(REDIS_WARNING, "Warning: no config file specified, using the default config. In order to specify a config file use %s /path/to/%s.conf", argv[0], server.sentinel_mode ? "sentinel" : "redis");
2631 }
2632 if (server.daemonize) daemonize();
2633 initServer();
2634 if (server.daemonize) createPidFile();
2635 redisAsciiArt();
2636
2637 if (!server.sentinel_mode) {
2638 /* Things only needed when not running in Sentinel mode. */
2639 redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
2640 #ifdef __linux__
2641 linuxOvercommitMemoryWarning();
2642 #endif
2643 loadDataFromDisk();
2644 if (server.ipfd > 0)
2645 redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
2646 if (server.sofd > 0)
2647 redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
2648 }
2649
2650 /* Warning the user about suspicious maxmemory setting. */
2651 if (server.maxmemory > 0 && server.maxmemory < 1024*1024) {
2652 redisLog(REDIS_WARNING,"WARNING: You specified a maxmemory value that is less than 1MB (current value is %llu bytes). Are you sure this is what you really want?", server.maxmemory);
2653 }
2654
2655 aeSetBeforeSleepProc(server.el,beforeSleep);
2656 aeMain(server.el);
2657 aeDeleteEventLoop(server.el);
2658 return 0;
2659 }
2660
2661 /* The End */