server.c 151.7 KB
Newer Older
1
/*
A
antirez 已提交
2
 * Copyright (c) 2009-2016, Salvatore Sanfilippo <antirez at gmail dot com>
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

30
#include "server.h"
31
#include "cluster.h"
32
#include "slowlog.h"
33
#include "bio.h"
34
#include "latency.h"
35
#include "atomicvar.h"
36 37 38 39 40 41 42 43 44 45 46 47 48 49

#include <time.h>
#include <signal.h>
#include <sys/wait.h>
#include <errno.h>
#include <assert.h>
#include <ctype.h>
#include <stdarg.h>
#include <arpa/inet.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <sys/time.h>
#include <sys/resource.h>
#include <sys/uio.h>
50
#include <sys/un.h>
51 52 53
#include <limits.h>
#include <float.h>
#include <math.h>
54
#include <sys/resource.h>
55
#include <sys/utsname.h>
56
#include <locale.h>
C
clark.kang 已提交
57
#include <sys/socket.h>
58 59 60 61 62

/* Our shared "common" objects */

struct sharedObjectsStruct shared;

63
/* Global vars that are actually used as constants. The following double
64 65 66 67 68 69 70 71
 * values are used for double on-disk serialization, and are initialized
 * at runtime to avoid strange compiler optimizations. */

double R_Zero, R_PosInf, R_NegInf, R_Nan;

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

/* Global vars */
72 73
struct redisServer server; /* Server global state */
volatile unsigned long lru_clock; /* Server global current LRU time. */
74

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
/* Our command table.
 *
 * Every entry is composed of the following fields:
 *
 * name: a string representing the command name.
 * function: pointer to the C function implementing the command.
 * arity: number of arguments, it is possible to use -N to say >= N
 * sflags: command flags as string. See below for a table of flags.
 * flags: flags as bitmask. Computed by Redis using the 'sflags' field.
 * get_keys_proc: an optional function to get key arguments from a command.
 *                This is only used when the following three fields are not
 *                enough to specify what arguments are keys.
 * first_key_index: first argument that is a key
 * last_key_index: last argument that is a key
 * key_step: step to get all the keys from first to last argument. For instance
 *           in MSET the step is two since arguments are key,val,key,val,...
 * microseconds: microseconds of total execution time for this command.
 * calls: total number of calls of this command.
 *
 * The flags, microseconds and calls fields are computed by Redis and should
 * always be set to zero.
 *
 * Command flags are expressed using strings where every character represents
 * a flag. Later the populateCommandTable() function will take care of
 * populating the real 'flags' field using this characters.
100 101 102 103 104 105 106 107
 *
 * This is the meaning of the flags:
 *
 * w: write command (may modify the key space).
 * r: read command  (will never modify the key space).
 * m: may increase memory usage once called. Don't allow if out of memory.
 * a: admin command, like SAVE or SHUTDOWN.
 * p: Pub/Sub related command.
G
guiquanz 已提交
108
 * f: force replication of this command, regardless of server.dirty.
109
 * s: command not allowed in scripts.
110
 * R: random command. Command is not deterministic, that is, the same command
111
 *    with the same arguments, with the same key space, may have different
112 113 114
 *    results. For instance SPOP and RANDOMKEY are two random commands.
 * S: Sort command output array if called from script, so that the output
 *    is deterministic.
115 116 117 118
 * l: Allow command while loading the database.
 * t: Allow command while a slave has stale data but is not allowed to
 *    server this data. Normally no command is accepted in this condition
 *    but just a few.
119
 * M: Do not automatically propagate the command on MONITOR.
120 121
 * k: Perform an implicit ASKING for this command, so the command will be
 *    accepted in cluster mode if the slot is marked as 'importing'.
122 123 124 125
 * F: Fast command: O(1) or O(log(N)) command that should never delay
 *    its execution as long as the kernel scheduler is giving us time.
 *    Note that commands that may trigger a DEL as a side effect (like SET)
 *    are not fast commands.
126
 */
127
struct redisCommand redisCommandTable[] = {
128
    {"module",moduleCommand,-2,"as",0,NULL,0,0,0,0,0},
129
    {"get",getCommand,2,"rF",0,NULL,1,1,1,0,0},
130
    {"set",setCommand,-3,"wm",0,NULL,1,1,1,0,0},
131
    {"setnx",setnxCommand,3,"wmF",0,NULL,1,1,1,0,0},
132 133
    {"setex",setexCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"psetex",psetexCommand,4,"wm",0,NULL,1,1,1,0,0},
134
    {"append",appendCommand,3,"wm",0,NULL,1,1,1,0,0},
135
    {"strlen",strlenCommand,2,"rF",0,NULL,1,1,1,0,0},
136
    {"del",delCommand,-2,"w",0,NULL,1,-1,1,0,0},
137
    {"unlink",unlinkCommand,-2,"wF",0,NULL,1,-1,1,0,0},
A
antirez 已提交
138
    {"exists",existsCommand,-2,"rF",0,NULL,1,-1,1,0,0},
139
    {"setbit",setbitCommand,4,"wm",0,NULL,1,1,1,0,0},
140
    {"getbit",getbitCommand,3,"rF",0,NULL,1,1,1,0,0},
141
    {"bitfield",bitfieldCommand,-2,"wm",0,NULL,1,1,1,0,0},
142 143 144
    {"setrange",setrangeCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"getrange",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"substr",getrangeCommand,4,"r",0,NULL,1,1,1,0,0},
145 146
    {"incr",incrCommand,2,"wmF",0,NULL,1,1,1,0,0},
    {"decr",decrCommand,2,"wmF",0,NULL,1,1,1,0,0},
147
    {"mget",mgetCommand,-2,"rF",0,NULL,1,-1,1,0,0},
148 149
    {"rpush",rpushCommand,-3,"wmF",0,NULL,1,1,1,0,0},
    {"lpush",lpushCommand,-3,"wmF",0,NULL,1,1,1,0,0},
P
Pierre Chapuis 已提交
150 151
    {"rpushx",rpushxCommand,-3,"wmF",0,NULL,1,1,1,0,0},
    {"lpushx",lpushxCommand,-3,"wmF",0,NULL,1,1,1,0,0},
152
    {"linsert",linsertCommand,5,"wm",0,NULL,1,1,1,0,0},
153 154
    {"rpop",rpopCommand,2,"wF",0,NULL,1,1,1,0,0},
    {"lpop",lpopCommand,2,"wF",0,NULL,1,1,1,0,0},
155
    {"brpop",brpopCommand,-3,"ws",0,NULL,1,-2,1,0,0},
A
antirez 已提交
156 157
    {"brpoplpush",brpoplpushCommand,4,"wms",0,NULL,1,2,1,0,0},
    {"blpop",blpopCommand,-3,"ws",0,NULL,1,-2,1,0,0},
158
    {"llen",llenCommand,2,"rF",0,NULL,1,1,1,0,0},
159 160 161 162 163 164
    {"lindex",lindexCommand,3,"r",0,NULL,1,1,1,0,0},
    {"lset",lsetCommand,4,"wm",0,NULL,1,1,1,0,0},
    {"lrange",lrangeCommand,4,"r",0,NULL,1,1,1,0,0},
    {"ltrim",ltrimCommand,4,"w",0,NULL,1,1,1,0,0},
    {"lrem",lremCommand,4,"w",0,NULL,1,1,1,0,0},
    {"rpoplpush",rpoplpushCommand,3,"wm",0,NULL,1,2,1,0,0},
165 166 167 168 169
    {"sadd",saddCommand,-3,"wmF",0,NULL,1,1,1,0,0},
    {"srem",sremCommand,-3,"wF",0,NULL,1,1,1,0,0},
    {"smove",smoveCommand,4,"wF",0,NULL,1,2,1,0,0},
    {"sismember",sismemberCommand,3,"rF",0,NULL,1,1,1,0,0},
    {"scard",scardCommand,2,"rF",0,NULL,1,1,1,0,0},
I
Itamar Haber 已提交
170
    {"spop",spopCommand,-2,"wRF",0,NULL,1,1,1,0,0},
171
    {"srandmember",srandmemberCommand,-2,"rR",0,NULL,1,1,1,0,0},
172
    {"sinter",sinterCommand,-2,"rS",0,NULL,1,-1,1,0,0},
173
    {"sinterstore",sinterstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
174
    {"sunion",sunionCommand,-2,"rS",0,NULL,1,-1,1,0,0},
175
    {"sunionstore",sunionstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
176
    {"sdiff",sdiffCommand,-2,"rS",0,NULL,1,-1,1,0,0},
177
    {"sdiffstore",sdiffstoreCommand,-3,"wm",0,NULL,1,-1,1,0,0},
178
    {"smembers",sinterCommand,2,"rS",0,NULL,1,1,1,0,0},
A
antirez 已提交
179
    {"sscan",sscanCommand,-3,"rR",0,NULL,1,1,1,0,0},
180 181 182
    {"zadd",zaddCommand,-4,"wmF",0,NULL,1,1,1,0,0},
    {"zincrby",zincrbyCommand,4,"wmF",0,NULL,1,1,1,0,0},
    {"zrem",zremCommand,-3,"wF",0,NULL,1,1,1,0,0},
183 184
    {"zremrangebyscore",zremrangebyscoreCommand,4,"w",0,NULL,1,1,1,0,0},
    {"zremrangebyrank",zremrangebyrankCommand,4,"w",0,NULL,1,1,1,0,0},
A
antirez 已提交
185
    {"zremrangebylex",zremrangebylexCommand,4,"w",0,NULL,1,1,1,0,0},
186 187 188 189 190
    {"zunionstore",zunionstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zinterstore",zinterstoreCommand,-4,"wm",0,zunionInterGetKeys,0,0,0,0,0},
    {"zrange",zrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrangebyscore",zrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrevrangebyscore",zrevrangebyscoreCommand,-4,"r",0,NULL,1,1,1,0,0},
191 192
    {"zrangebylex",zrangebylexCommand,-4,"r",0,NULL,1,1,1,0,0},
    {"zrevrangebylex",zrevrangebylexCommand,-4,"r",0,NULL,1,1,1,0,0},
193 194
    {"zcount",zcountCommand,4,"rF",0,NULL,1,1,1,0,0},
    {"zlexcount",zlexcountCommand,4,"rF",0,NULL,1,1,1,0,0},
195
    {"zrevrange",zrevrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
196 197 198 199
    {"zcard",zcardCommand,2,"rF",0,NULL,1,1,1,0,0},
    {"zscore",zscoreCommand,3,"rF",0,NULL,1,1,1,0,0},
    {"zrank",zrankCommand,3,"rF",0,NULL,1,1,1,0,0},
    {"zrevrank",zrevrankCommand,3,"rF",0,NULL,1,1,1,0,0},
A
antirez 已提交
200
    {"zscan",zscanCommand,-3,"rR",0,NULL,1,1,1,0,0},
201
    {"hset",hsetCommand,-4,"wmF",0,NULL,1,1,1,0,0},
202 203
    {"hsetnx",hsetnxCommand,4,"wmF",0,NULL,1,1,1,0,0},
    {"hget",hgetCommand,3,"rF",0,NULL,1,1,1,0,0},
204 205
    {"hmset",hsetCommand,-4,"wmF",0,NULL,1,1,1,0,0},
    {"hmget",hmgetCommand,-3,"rF",0,NULL,1,1,1,0,0},
206 207 208 209
    {"hincrby",hincrbyCommand,4,"wmF",0,NULL,1,1,1,0,0},
    {"hincrbyfloat",hincrbyfloatCommand,4,"wmF",0,NULL,1,1,1,0,0},
    {"hdel",hdelCommand,-3,"wF",0,NULL,1,1,1,0,0},
    {"hlen",hlenCommand,2,"rF",0,NULL,1,1,1,0,0},
A
antirez 已提交
210
    {"hstrlen",hstrlenCommand,3,"rF",0,NULL,1,1,1,0,0},
211 212
    {"hkeys",hkeysCommand,2,"rS",0,NULL,1,1,1,0,0},
    {"hvals",hvalsCommand,2,"rS",0,NULL,1,1,1,0,0},
213
    {"hgetall",hgetallCommand,2,"r",0,NULL,1,1,1,0,0},
214
    {"hexists",hexistsCommand,3,"rF",0,NULL,1,1,1,0,0},
A
antirez 已提交
215
    {"hscan",hscanCommand,-3,"rR",0,NULL,1,1,1,0,0},
216 217 218
    {"incrby",incrbyCommand,3,"wmF",0,NULL,1,1,1,0,0},
    {"decrby",decrbyCommand,3,"wmF",0,NULL,1,1,1,0,0},
    {"incrbyfloat",incrbyfloatCommand,3,"wmF",0,NULL,1,1,1,0,0},
219 220 221
    {"getset",getsetCommand,3,"wm",0,NULL,1,1,1,0,0},
    {"mset",msetCommand,-3,"wm",0,NULL,1,-1,2,0,0},
    {"msetnx",msetnxCommand,-3,"wm",0,NULL,1,-1,2,0,0},
222
    {"randomkey",randomkeyCommand,1,"rR",0,NULL,0,0,0,0,0},
223
    {"select",selectCommand,2,"lF",0,NULL,0,0,0,0,0},
A
antirez 已提交
224
    {"swapdb",swapdbCommand,3,"wF",0,NULL,0,0,0,0,0},
225
    {"move",moveCommand,3,"wF",0,NULL,1,1,1,0,0},
226
    {"rename",renameCommand,3,"w",0,NULL,1,2,1,0,0},
227 228 229 230 231
    {"renamenx",renamenxCommand,3,"wF",0,NULL,1,2,1,0,0},
    {"expire",expireCommand,3,"wF",0,NULL,1,1,1,0,0},
    {"expireat",expireatCommand,3,"wF",0,NULL,1,1,1,0,0},
    {"pexpire",pexpireCommand,3,"wF",0,NULL,1,1,1,0,0},
    {"pexpireat",pexpireatCommand,3,"wF",0,NULL,1,1,1,0,0},
232
    {"keys",keysCommand,2,"rS",0,NULL,0,0,0,0,0},
233
    {"scan",scanCommand,-2,"rR",0,NULL,0,0,0,0,0},
234
    {"dbsize",dbsizeCommand,1,"rF",0,NULL,0,0,0,0,0},
235 236 237 238
    {"auth",authCommand,2,"sltF",0,NULL,0,0,0,0,0},
    {"ping",pingCommand,-1,"tF",0,NULL,0,0,0,0,0},
    {"echo",echoCommand,2,"F",0,NULL,0,0,0,0,0},
    {"save",saveCommand,1,"as",0,NULL,0,0,0,0,0},
239
    {"bgsave",bgsaveCommand,-1,"a",0,NULL,0,0,0,0,0},
240 241 242
    {"bgrewriteaof",bgrewriteaofCommand,1,"a",0,NULL,0,0,0,0,0},
    {"shutdown",shutdownCommand,-1,"alt",0,NULL,0,0,0,0,0},
    {"lastsave",lastsaveCommand,1,"RF",0,NULL,0,0,0,0,0},
243
    {"type",typeCommand,2,"rF",0,NULL,1,1,1,0,0},
244
    {"multi",multiCommand,1,"sF",0,NULL,0,0,0,0,0},
245
    {"exec",execCommand,1,"sM",0,NULL,0,0,0,0,0},
246
    {"discard",discardCommand,1,"sF",0,NULL,0,0,0,0,0},
247
    {"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
248
    {"psync",syncCommand,3,"ars",0,NULL,0,0,0,0,0},
249
    {"replconf",replconfCommand,-1,"aslt",0,NULL,0,0,0,0,0},
250 251
    {"flushdb",flushdbCommand,-1,"w",0,NULL,0,0,0,0,0},
    {"flushall",flushallCommand,-1,"w",0,NULL,0,0,0,0,0},
252
    {"sort",sortCommand,-2,"wm",0,sortGetKeys,1,1,1,0,0},
253 254
    {"info",infoCommand,-1,"lt",0,NULL,0,0,0,0,0},
    {"monitor",monitorCommand,1,"as",0,NULL,0,0,0,0,0},
255
    {"ttl",ttlCommand,2,"rF",0,NULL,1,1,1,0,0},
256
    {"touch",touchCommand,-2,"rF",0,NULL,1,1,1,0,0},
257 258
    {"pttl",pttlCommand,2,"rF",0,NULL,1,1,1,0,0},
    {"persist",persistCommand,2,"wF",0,NULL,1,1,1,0,0},
259
    {"slaveof",slaveofCommand,3,"ast",0,NULL,0,0,0,0,0},
260
    {"role",roleCommand,1,"lst",0,NULL,0,0,0,0,0},
261
    {"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
A
antirez 已提交
262
    {"config",configCommand,-2,"lat",0,NULL,0,0,0,0,0},
263 264 265 266 267 268 269 270 271
    {"subscribe",subscribeCommand,-2,"pslt",0,NULL,0,0,0,0,0},
    {"unsubscribe",unsubscribeCommand,-1,"pslt",0,NULL,0,0,0,0,0},
    {"psubscribe",psubscribeCommand,-2,"pslt",0,NULL,0,0,0,0,0},
    {"punsubscribe",punsubscribeCommand,-1,"pslt",0,NULL,0,0,0,0,0},
    {"publish",publishCommand,3,"pltF",0,NULL,0,0,0,0,0},
    {"pubsub",pubsubCommand,-2,"pltR",0,NULL,0,0,0,0,0},
    {"watch",watchCommand,-2,"sF",0,NULL,1,-1,1,0,0},
    {"unwatch",unwatchCommand,1,"sF",0,NULL,0,0,0,0,0},
    {"cluster",clusterCommand,-2,"a",0,NULL,0,0,0,0,0},
272 273
    {"restore",restoreCommand,-4,"wm",0,NULL,1,1,1,0,0},
    {"restore-asking",restoreCommand,-4,"wmk",0,NULL,1,1,1,0,0},
274
    {"migrate",migrateCommand,-6,"w",0,migrateGetKeys,0,0,0,0,0},
275 276 277
    {"asking",askingCommand,1,"F",0,NULL,0,0,0,0,0},
    {"readonly",readonlyCommand,1,"F",0,NULL,0,0,0,0,0},
    {"readwrite",readwriteCommand,1,"F",0,NULL,0,0,0,0,0},
278
    {"dump",dumpCommand,2,"r",0,NULL,1,1,1,0,0},
I
Itamar Haber 已提交
279
    {"object",objectCommand,-2,"r",0,NULL,2,2,2,0,0},
280
    {"memory",memoryCommand,-2,"r",0,NULL,0,0,0,0,0},
281
    {"client",clientCommand,-2,"as",0,NULL,0,0,0,0,0},
282 283
    {"eval",evalCommand,-3,"s",0,evalGetKeys,0,0,0,0,0},
    {"evalsha",evalShaCommand,-3,"s",0,evalGetKeys,0,0,0,0,0},
284 285 286
    {"slowlog",slowlogCommand,-2,"a",0,NULL,0,0,0,0,0},
    {"script",scriptCommand,-2,"s",0,NULL,0,0,0,0,0},
    {"time",timeCommand,1,"RF",0,NULL,0,0,0,0,0},
A
antirez 已提交
287
    {"bitop",bitopCommand,-4,"wm",0,NULL,2,-1,1,0,0},
288
    {"bitcount",bitcountCommand,-2,"r",0,NULL,1,1,1,0,0},
A
antirez 已提交
289
    {"bitpos",bitposCommand,-3,"r",0,NULL,1,1,1,0,0},
290
    {"wait",waitCommand,3,"s",0,NULL,0,0,0,0,0},
A
antirez 已提交
291
    {"command",commandCommand,0,"lt",0,NULL,0,0,0,0,0},
292
    {"geoadd",geoaddCommand,-5,"wm",0,NULL,1,1,1,0,0},
293
    {"georadius",georadiusCommand,-6,"w",0,georadiusGetKeys,1,1,1,0,0},
294 295 296
    {"georadius_ro",georadiusroCommand,-6,"r",0,georadiusGetKeys,1,1,1,0,0},
    {"georadiusbymember",georadiusbymemberCommand,-5,"w",0,georadiusGetKeys,1,1,1,0,0},
    {"georadiusbymember_ro",georadiusbymemberroCommand,-5,"r",0,georadiusGetKeys,1,1,1,0,0},
297 298 299
    {"geohash",geohashCommand,-2,"r",0,NULL,1,1,1,0,0},
    {"geopos",geoposCommand,-2,"r",0,NULL,1,1,1,0,0},
    {"geodist",geodistCommand,-4,"r",0,NULL,1,1,1,0,0},
300
    {"pfselftest",pfselftestCommand,1,"a",0,NULL,0,0,0,0,0},
301
    {"pfadd",pfaddCommand,-2,"wmF",0,NULL,1,1,1,0,0},
M
MOON_CLJ 已提交
302
    {"pfcount",pfcountCommand,-2,"r",0,NULL,1,-1,1,0,0},
A
antirez 已提交
303
    {"pfmerge",pfmergeCommand,-2,"wm",0,NULL,1,-1,1,0,0},
A
antirez 已提交
304
    {"pfdebug",pfdebugCommand,-3,"w",0,NULL,0,0,0,0,0},
305
    {"xadd",xaddCommand,-5,"wmF",0,NULL,1,1,1,0,0},
306
    {"xrange",xrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
307
    {"xrevrange",xrevrangeCommand,-4,"r",0,NULL,1,1,1,0,0},
A
antirez 已提交
308
    {"xlen",xlenCommand,2,"rF",0,NULL,1,1,1,0,0},
A
antirez 已提交
309
    {"xread",xreadCommand,-3,"rs",0,xreadGetKeys,1,1,1,0,0},
310 311
    {"post",securityWarningCommand,-1,"lt",0,NULL,0,0,0,0,0},
    {"host:",securityWarningCommand,-1,"lt",0,NULL,0,0,0,0,0},
312
    {"latency",latencyCommand,-2,"aslt",0,NULL,0,0,0,0,0}
313 314 315 316
};

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

317
/* Low level logging. To use only for very big messages, otherwise
A
antirez 已提交
318 319
 * serverLog() is to prefer. */
void serverLogRaw(int level, const char *msg) {
J
Jonah H. Harris 已提交
320 321
    const int syslogLevelMap[] = { LOG_DEBUG, LOG_INFO, LOG_NOTICE, LOG_WARNING };
    const char *c = ".-*#";
322
    FILE *fp;
323
    char buf[64];
A
antirez 已提交
324
    int rawmode = (level & LL_RAW);
325
    int log_to_stdout = server.logfile[0] == '\0';
326

A
antirez 已提交
327
    level &= 0xff; /* clear flags */
328
    if (level < server.verbosity) return;
329

330
    fp = log_to_stdout ? stdout : fopen(server.logfile,"a");
331 332
    if (!fp) return;

A
antirez 已提交
333 334 335
    if (rawmode) {
        fprintf(fp,"%s",msg);
    } else {
336 337
        int off;
        struct timeval tv;
A
antirez 已提交
338 339
        int role_char;
        pid_t pid = getpid();
340 341 342 343

        gettimeofday(&tv,NULL);
        off = strftime(buf,sizeof(buf),"%d %b %H:%M:%S.",localtime(&tv.tv_sec));
        snprintf(buf+off,sizeof(buf)-off,"%03d",(int)tv.tv_usec/1000);
A
antirez 已提交
344 345 346 347 348 349 350 351 352
        if (server.sentinel_mode) {
            role_char = 'X'; /* Sentinel. */
        } else if (pid != server.pid) {
            role_char = 'C'; /* RDB / AOF writing child. */
        } else {
            role_char = (server.masterhost ? 'S':'M'); /* Slave or Master. */
        }
        fprintf(fp,"%d:%c %s %c %s\n",
            (int)getpid(),role_char, buf,c[level],msg);
A
antirez 已提交
353
    }
J
Jonah H. Harris 已提交
354 355
    fflush(fp);

356
    if (!log_to_stdout) fclose(fp);
J
Jonah H. Harris 已提交
357
    if (server.syslog_enabled) syslog(syslogLevelMap[level], "%s", msg);
358 359
}

A
antirez 已提交
360
/* Like serverLogRaw() but with printf-alike support. This is the function that
361 362
 * is used across the code. The raw version is only used in order to dump
 * the INFO output on crash. */
A
antirez 已提交
363
void serverLog(int level, const char *fmt, ...) {
364
    va_list ap;
A
antirez 已提交
365
    char msg[LOG_MAX_LEN];
366

A
antirez 已提交
367
    if ((level&0xff) < server.verbosity) return;
368 369 370 371 372

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

A
antirez 已提交
373
    serverLogRaw(level,msg);
374 375
}

A
antirez 已提交
376 377 378 379 380
/* Log a fixed message without printf-alike capabilities, in a way that is
 * safe to call from a signal handler.
 *
 * We actually use this only for signals that are not fatal from the point
 * of view of Redis. Signals that are going to kill the server anyway and
A
antirez 已提交
381 382
 * where we need printf-alike features are served by serverLog(). */
void serverLogFromHandler(int level, const char *msg) {
A
antirez 已提交
383
    int fd;
384
    int log_to_stdout = server.logfile[0] == '\0';
A
antirez 已提交
385 386
    char buf[64];

387 388
    if ((level&0xff) < server.verbosity || (log_to_stdout && server.daemonize))
        return;
389
    fd = log_to_stdout ? STDOUT_FILENO :
390
                         open(server.logfile, O_APPEND|O_CREAT|O_WRONLY, 0644);
A
antirez 已提交
391 392
    if (fd == -1) return;
    ll2string(buf,sizeof(buf),getpid());
393
    if (write(fd,buf,strlen(buf)) == -1) goto err;
394
    if (write(fd,":signal-handler (",17) == -1) goto err;
A
antirez 已提交
395
    ll2string(buf,sizeof(buf),time(NULL));
396 397 398 399 400
    if (write(fd,buf,strlen(buf)) == -1) goto err;
    if (write(fd,") ",2) == -1) goto err;
    if (write(fd,msg,strlen(msg)) == -1) goto err;
    if (write(fd,"\n",1) == -1) goto err;
err:
401
    if (!log_to_stdout) close(fd);
A
antirez 已提交
402 403
}

A
antirez 已提交
404 405 406 407 408 409 410 411 412 413 414
/* Return the UNIX time in microseconds */
long long ustime(void) {
    struct timeval tv;
    long long ust;

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

415
/* Return the UNIX time in milliseconds */
416
mstime_t mstime(void) {
417 418 419
    return ustime()/1000;
}

420 421 422 423 424 425 426 427 428 429 430 431
/* After an RDB dump or AOF rewrite we exit from children using _exit() instead of
 * exit(), because the latter may interact with the same file objects used by
 * the parent process. However if we are testing the coverage normal exit() is
 * used in order to obtain the right coverage information. */
void exitFromChild(int retcode) {
#ifdef COVERAGE_TEST
    exit(retcode);
#else
    _exit(retcode);
#endif
}

432 433
/*====================== Hash table type implementation  ==================== */

434
/* This is a hash table type that uses the SDS dynamic strings library as
T
T.J. Schuck 已提交
435
 * keys and redis objects as values (objects can hold SDS strings,
436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
 * lists, sets). */

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

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

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

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

A
antirez 已提交
462 463
/* A case insensitive version used for the command lookup table and other
 * places where case insensitive non binary-safe comparison is needed. */
464 465 466 467 468 469 470 471
int dictSdsKeyCaseCompare(void *privdata, const void *key1,
        const void *key2)
{
    DICT_NOTUSED(privdata);

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

472
void dictObjectDestructor(void *privdata, void *val)
473 474 475
{
    DICT_NOTUSED(privdata);

476
    if (val == NULL) return; /* Lazy freeing will set value to NULL. */
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493
    decrRefCount(val);
}

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

    sdsfree(val);
}

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

494
uint64_t dictObjHash(const void *key) {
495 496 497 498
    const robj *o = key;
    return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
}

499
uint64_t dictSdsHash(const void *key) {
500 501 502
    return dictGenHashFunction((unsigned char*)key, sdslen((char*)key));
}

503
uint64_t dictSdsCaseHash(const void *key) {
504 505 506
    return dictGenCaseHashFunction((unsigned char*)key, sdslen((char*)key));
}

507 508 509 510 511 512
int dictEncObjKeyCompare(void *privdata, const void *key1,
        const void *key2)
{
    robj *o1 = (robj*) key1, *o2 = (robj*) key2;
    int cmp;

513 514
    if (o1->encoding == OBJ_ENCODING_INT &&
        o2->encoding == OBJ_ENCODING_INT)
515 516 517 518 519 520 521 522 523 524
            return o1->ptr == o2->ptr;

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

525
uint64_t dictEncObjHash(const void *key) {
526 527
    robj *o = (robj*) key;

528
    if (sdsEncodedObject(o)) {
529 530
        return dictGenHashFunction(o->ptr, sdslen((sds)o->ptr));
    } else {
531
        if (o->encoding == OBJ_ENCODING_INT) {
532 533 534 535 536 537
            char buf[32];
            int len;

            len = ll2string(buf,32,(long)o->ptr);
            return dictGenHashFunction((unsigned char*)buf, len);
        } else {
538
            uint64_t hash;
539 540 541 542 543 544 545 546 547

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

548 549 550
/* Generic hash table type where keys are Redis Objects, Values
 * dummy pointers. */
dictType objectKeyPointerValueDictType = {
551 552 553 554
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
555
    dictObjectDestructor,      /* key destructor */
556 557 558
    NULL                       /* val destructor */
};

559 560 561 562 563 564 565 566 567 568 569
/* Like objectKeyPointerValueDictType(), but values can be destroyed, if
 * not NULL, calling zfree(). */
dictType objectKeyHeapPointerValueDictType = {
    dictEncObjHash,            /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictEncObjKeyCompare,      /* key compare */
    dictObjectDestructor,      /* key destructor */
    dictVanillaFree            /* val destructor */
};

570 571 572 573 574 575 576 577 578 579
/* Set dictionary type. Keys are SDS strings, values are ot used. */
dictType setDictType = {
    dictSdsHash,               /* hash function */
    NULL,                      /* key dup */
    NULL,                      /* val dup */
    dictSdsKeyCompare,         /* key compare */
    dictSdsDestructor,         /* key destructor */
    NULL                       /* val destructor */
};

580 581
/* Sorted sets hash (note: a skiplist is used in addition to the hash table) */
dictType zsetDictType = {
582
    dictSdsHash,               /* hash function */
583 584
    NULL,                      /* key dup */
    NULL,                      /* val dup */
585 586
    dictSdsKeyCompare,         /* key compare */
    NULL,                      /* Note: SDS string shared & freed by skiplist */
587
    NULL                       /* val destructor */
588 589 590 591 592 593 594 595 596
};

/* Db->dict, keys are sds strings, vals are Redis objects. */
dictType dbDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
597
    dictObjectDestructor   /* val destructor */
598 599
};

A
antirez 已提交
600 601 602 603 604 605 606
/* server.lua_scripts sha (as sds string) -> scripts (as robj) cache. */
dictType shaScriptObjectDictType = {
    dictSdsCaseHash,            /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCaseCompare,      /* key compare */
    dictSdsDestructor,          /* key destructor */
607
    dictObjectDestructor        /* val destructor */
A
antirez 已提交
608 609
};

610 611
/* Db->expires */
dictType keyptrDictType = {
612 613 614 615 616 617
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    NULL,                       /* key destructor */
    NULL                        /* val destructor */
618 619
};

620 621
/* Command table. sds string -> command struct pointer. */
dictType commandTableDictType = {
622 623 624 625 626 627
    dictSdsCaseHash,            /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCaseCompare,      /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
628 629
};

630
/* Hash type hash table (note that small hashes are represented with ziplists) */
631
dictType hashDictType = {
632
    dictSdsHash,                /* hash function */
633 634
    NULL,                       /* key dup */
    NULL,                       /* val dup */
635
    dictSdsKeyCompare,          /* key compare */
636 637
    dictSdsDestructor,          /* key destructor */
    dictSdsDestructor           /* val destructor */
638 639 640 641 642 643 644 645 646 647
};

/* Keylist hash table type has unencoded redis objects as keys and
 * lists as values. It's used for blocking operations (BLPOP) and to
 * map swapped keys to a list of clients waiting for this keys to be loaded. */
dictType keylistDictType = {
    dictObjHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictObjKeyCompare,          /* key compare */
648
    dictObjectDestructor,       /* key destructor */
649 650 651
    dictListDestructor          /* val destructor */
};

A
antirez 已提交
652 653 654 655 656 657 658 659 660 661 662
/* Cluster nodes hash table, mapping nodes addresses 1.2.3.4:6379 to
 * clusterNode structures. */
dictType clusterNodesDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

663 664 665 666 667 668 669 670 671 672 673 674
/* Cluster re-addition blacklist. This maps node IDs to the time
 * we can re-add this node. The goal is to avoid readding a removed
 * node for some time. */
dictType clusterNodesBlackListDictType = {
    dictSdsCaseHash,            /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCaseCompare,      /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

A
antirez 已提交
675 676 677 678 679 680 681 682 683 684 685 686
/* Cluster re-addition blacklist. This maps node IDs to the time
 * we can re-add this node. The goal is to avoid readding a removed
 * node for some time. */
dictType modulesDictType = {
    dictSdsCaseHash,            /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCaseCompare,      /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

A
antirez 已提交
687 688 689 690 691 692 693 694 695 696
/* Migrate cache dict type. */
dictType migrateCacheDictType = {
    dictSdsHash,                /* hash function */
    NULL,                       /* key dup */
    NULL,                       /* val dup */
    dictSdsKeyCompare,          /* key compare */
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

697 698 699 700
/* Replication cached script dict (server.repl_scriptcache_dict).
 * Keys are sds SHA1 strings, while values are not used at all in the current
 * implementation. */
dictType replScriptCacheDictType = {
701
    dictSdsCaseHash,            /* hash function */
702 703
    NULL,                       /* key dup */
    NULL,                       /* val dup */
704
    dictSdsKeyCaseCompare,      /* key compare */
705 706 707 708
    dictSdsDestructor,          /* key destructor */
    NULL                        /* val destructor */
};

709 710 711 712 713
int htNeedsResize(dict *dict) {
    long long size, used;

    size = dictSlots(dict);
    used = dictSize(dict);
714
    return (size > DICT_HT_INITIAL_SIZE &&
A
antirez 已提交
715
            (used*100/size < HASHTABLE_MIN_FILL));
716 717
}

A
antirez 已提交
718
/* If the percentage of used slots in the HT reaches HASHTABLE_MIN_FILL
719
 * we resize the hash table to save memory */
720 721 722 723 724
void tryResizeHashTables(int dbid) {
    if (htNeedsResize(server.db[dbid].dict))
        dictResize(server.db[dbid].dict);
    if (htNeedsResize(server.db[dbid].expires))
        dictResize(server.db[dbid].expires);
725 726 727 728 729
}

/* Our hash table implementation performs rehashing incrementally while
 * we write/read from the hash table. Still if the server is idle, the hash
 * table will use two tables for a long time. So we try to use 1 millisecond
730 731 732 733 734 735 736 737 738 739 740 741 742 743
 * of CPU time at every call of this function to perform some rehahsing.
 *
 * The function returns 1 if some rehashing was performed, otherwise 0
 * is returned. */
int incrementallyRehash(int dbid) {
    /* Keys dictionary */
    if (dictIsRehashing(server.db[dbid].dict)) {
        dictRehashMilliseconds(server.db[dbid].dict,1);
        return 1; /* already used our millisecond for this loop... */
    }
    /* Expires */
    if (dictIsRehashing(server.db[dbid].expires)) {
        dictRehashMilliseconds(server.db[dbid].expires,1);
        return 1; /* already used our millisecond for this loop... */
744
    }
745
    return 0;
746 747 748 749 750 751 752 753 754
}

/* This function is called once a background process of some kind terminates,
 * as we want to avoid resizing the hash tables when there is a child in order
 * to play well with copy-on-write (otherwise when a resize happens lots of
 * memory pages are copied). The goal of this function is to update the ability
 * for dict.c to resize the hash tables accordingly to the fact we have o not
 * running childs. */
void updateDictResizePolicy(void) {
A
antirez 已提交
755
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
756 757 758 759 760 761 762
        dictEnableResize();
    else
        dictDisableResize();
}

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

763
/* Add a sample to the operations per second array of samples. */
764 765 766 767
void trackInstantaneousMetric(int metric, long long current_reading) {
    long long t = mstime() - server.inst_metric[metric].last_sample_time;
    long long ops = current_reading -
                    server.inst_metric[metric].last_sample_count;
768 769 770 771
    long long ops_sec;

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

772 773 774
    server.inst_metric[metric].samples[server.inst_metric[metric].idx] =
        ops_sec;
    server.inst_metric[metric].idx++;
A
antirez 已提交
775
    server.inst_metric[metric].idx %= STATS_METRIC_SAMPLES;
776 777
    server.inst_metric[metric].last_sample_time = mstime();
    server.inst_metric[metric].last_sample_count = current_reading;
778 779 780
}

/* Return the mean of all the samples. */
781
long long getInstantaneousMetric(int metric) {
782 783 784
    int j;
    long long sum = 0;

A
antirez 已提交
785
    for (j = 0; j < STATS_METRIC_SAMPLES; j++)
786
        sum += server.inst_metric[metric].samples[j];
A
antirez 已提交
787
    return sum / STATS_METRIC_SAMPLES;
788 789
}

A
antirez 已提交
790 791 792 793
/* Check for timeouts. Returns non-zero if the client was terminated.
 * The function gets the current time in milliseconds as argument since
 * it gets called multiple times in a loop, so calling gettimeofday() for
 * each iteration would be costly without any actual gain. */
794
int clientsCronHandleTimeout(client *c, mstime_t now_ms) {
A
antirez 已提交
795
    time_t now = now_ms/1000;
796 797

    if (server.maxidletime &&
A
antirez 已提交
798 799 800 801
        !(c->flags & CLIENT_SLAVE) &&    /* no timeout for slaves */
        !(c->flags & CLIENT_MASTER) &&   /* no timeout for masters */
        !(c->flags & CLIENT_BLOCKED) &&  /* no timeout for BLPOP */
        !(c->flags & CLIENT_PUBSUB) &&   /* no timeout for Pub/Sub clients */
802 803
        (now - c->lastinteraction > server.maxidletime))
    {
A
antirez 已提交
804
        serverLog(LL_VERBOSE,"Closing idle client");
805
        freeClient(c);
A
antirez 已提交
806
        return 1;
A
antirez 已提交
807
    } else if (c->flags & CLIENT_BLOCKED) {
A
antirez 已提交
808 809 810 811 812
        /* Blocked OPS timeout is handled with milliseconds resolution.
         * However note that the actual resolution is limited by
         * server.hz. */

        if (c->bpop.timeout != 0 && c->bpop.timeout < now_ms) {
813
            /* Handle blocking operation specific timeout. */
814 815
            replyToBlockedClientTimedOut(c);
            unblockClient(c);
816 817 818 819 820
        } else if (server.cluster_enabled) {
            /* Cluster: handle unblock & redirect of clients blocked
             * into keys no longer served by this server. */
            if (clusterRedirectBlockedClientIfNeeded(c))
                unblockClient(c);
821 822
        }
    }
A
antirez 已提交
823
    return 0;
824 825
}

826
/* The client query buffer is an sds.c string that can end with a lot of
A
antirez 已提交
827 828
 * free space not used, this function reclaims space if needed.
 *
G
guiquanz 已提交
829
 * The function always returns 0 as it never terminates the client. */
830
int clientsCronResizeQueryBuffer(client *c) {
831 832 833 834 835 836
    size_t querybuf_size = sdsAllocSize(c->querybuf);
    time_t idletime = server.unixtime - c->lastinteraction;

    /* There are two conditions to resize the query buffer:
     * 1) Query buffer is > BIG_ARG and too big for latest peak.
     * 2) Client is inactive and the buffer is bigger than 1k. */
A
antirez 已提交
837
    if (((querybuf_size > PROTO_MBULK_BIG_ARG) &&
838 839 840 841 842 843 844 845 846 847 848
         (querybuf_size/(c->querybuf_peak+1)) > 2) ||
         (querybuf_size > 1024 && idletime > 2))
    {
        /* Only resize the query buffer if it is actually wasting space. */
        if (sdsavail(c->querybuf) > 1024) {
            c->querybuf = sdsRemoveFreeSpace(c->querybuf);
        }
    }
    /* Reset the peak again to capture the peak memory usage in the next
     * cycle. */
    c->querybuf_peak = 0;
A
antirez 已提交
849
    return 0;
850 851
}

A
antirez 已提交
852
#define CLIENTS_CRON_MIN_ITERATIONS 5
853
void clientsCron(void) {
A
antirez 已提交
854
    /* Make sure to process at least numclients/server.hz of clients
A
antirez 已提交
855
     * per call. Since this function is called server.hz times per second
A
antirez 已提交
856 857
     * we are sure that in the worst case we process all the clients in 1
     * second. */
858
    int numclients = listLength(server.clients);
A
antirez 已提交
859 860 861 862 863 864 865 866 867
    int iterations = numclients/server.hz;
    mstime_t now = mstime();

    /* Process at least a few clients while we are at it, even if we need
     * to process less than CLIENTS_CRON_MIN_ITERATIONS to meet our contract
     * of processing each client once per second. */
    if (iterations < CLIENTS_CRON_MIN_ITERATIONS)
        iterations = (numclients < CLIENTS_CRON_MIN_ITERATIONS) ?
                     numclients : CLIENTS_CRON_MIN_ITERATIONS;
868 869

    while(listLength(server.clients) && iterations--) {
870
        client *c;
871 872 873 874 875 876 877 878
        listNode *head;

        /* Rotate the list, take the current head, process.
         * This way if the client must be removed from the list it's the
         * first element and we don't incur into O(N) computation. */
        listRotate(server.clients);
        head = listFirst(server.clients);
        c = listNodeValue(head);
A
antirez 已提交
879 880 881
        /* The following functions do different service checks on the client.
         * The protocol is that they return non-zero if the client was
         * terminated. */
A
antirez 已提交
882
        if (clientsCronHandleTimeout(c,now)) continue;
A
antirez 已提交
883
        if (clientsCronResizeQueryBuffer(c)) continue;
884 885 886
    }
}

887 888 889 890
/* This function handles 'background' operations we are required to do
 * incrementally in Redis databases, such as active key expiring, resizing,
 * rehashing. */
void databasesCron(void) {
891 892
    /* Expire keys by random sampling. Not required for slaves
     * as master will synthesize DELs for us. */
893
    if (server.active_expire_enabled && server.masterhost == NULL) {
A
antirez 已提交
894
        activeExpireCycle(ACTIVE_EXPIRE_CYCLE_SLOW);
895 896 897
    } else if (server.masterhost != NULL) {
        expireSlaveKeys();
    }
898

O
oranagra 已提交
899 900 901 902
    /* Defrag keys gradually. */
    if (server.active_defrag_enabled)
        activeDefragCycle();

903 904 905
    /* Perform hash tables rehashing if needed, but only if there are no
     * other processes saving the DB on disk. Otherwise rehashing is bad
     * as will cause a lot of copy-on-write of memory pages. */
906
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
907 908 909
        /* We use global counters so if we stop the computation at a given
         * DB we'll be able to start from the successive in the next
         * cron loop iteration. */
910 911
        static unsigned int resize_db = 0;
        static unsigned int rehash_db = 0;
A
antirez 已提交
912
        int dbs_per_call = CRON_DBS_PER_CALL;
913
        int j;
914

915 916 917
        /* Don't test more DBs than we have. */
        if (dbs_per_call > server.dbnum) dbs_per_call = server.dbnum;

918
        /* Resize */
919
        for (j = 0; j < dbs_per_call; j++) {
920 921 922 923 924 925
            tryResizeHashTables(resize_db % server.dbnum);
            resize_db++;
        }

        /* Rehash */
        if (server.activerehashing) {
926
            for (j = 0; j < dbs_per_call; j++) {
Z
zhaozhao.zz 已提交
927
                int work_done = incrementallyRehash(rehash_db);
928 929 930 931
                if (work_done) {
                    /* If the function did some work, stop here, we'll do
                     * more at the next cron loop. */
                    break;
Z
zhaozhao.zz 已提交
932 933 934 935
                } else {
                    /* If this db didn't need rehash, we'll try the next one. */
                    rehash_db++;
                    rehash_db %= server.dbnum;
936 937 938
                }
            }
        }
939 940 941
    }
}

942 943 944 945 946
/* We take a cached value of the unix time in the global state because with
 * virtual memory and aging there is to store the current time in objects at
 * every object access, and accuracy is not needed. To access a global var is
 * a lot faster than calling time(NULL) */
void updateCachedTime(void) {
947 948
    time_t unixtime = time(NULL);
    atomicSet(server.unixtime,unixtime);
949 950 951
    server.mstime = mstime();
}

952
/* This is our timer interrupt, called server.hz times per second.
953 954 955 956 957
 * Here is where we do a number of things that need to be done asynchronously.
 * For instance:
 *
 * - Active expired keys collection (it is also performed in a lazy way on
 *   lookup).
G
guiquanz 已提交
958
 * - Software watchdog.
959 960 961
 * - Update some statistic.
 * - Incremental rehashing of the DBs hash tables.
 * - Triggering BGSAVE / AOF rewrite, and handling of terminated children.
G
guiquanz 已提交
962
 * - Clients timeout of different kinds.
963 964 965
 * - Replication reconnection.
 * - Many more...
 *
966
 * Everything directly called here will be called server.hz times per second,
967 968 969 970
 * so in order to throttle execution of things we want to do less frequently
 * a macro is used: run_with_period(milliseconds) { .... }
 */

971
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
972
    int j;
A
antirez 已提交
973 974 975
    UNUSED(eventLoop);
    UNUSED(id);
    UNUSED(clientData);
976

A
antirez 已提交
977 978 979 980
    /* Software watchdog: deliver the SIGALRM that will reach the signal
     * handler if we don't return here fast enough. */
    if (server.watchdog_period) watchdogScheduleSignal(server.watchdog_period);

981 982
    /* Update the time cache. */
    updateCachedTime();
A
antirez 已提交
983

984
    run_with_period(100) {
A
antirez 已提交
985 986
        trackInstantaneousMetric(STATS_METRIC_COMMAND,server.stat_numcommands);
        trackInstantaneousMetric(STATS_METRIC_NET_INPUT,
987
                server.stat_net_input_bytes);
A
antirez 已提交
988
        trackInstantaneousMetric(STATS_METRIC_NET_OUTPUT,
989 990
                server.stat_net_output_bytes);
    }
991

A
antirez 已提交
992
    /* We have just LRU_BITS bits per object for LRU information.
993
     * So we use an (eventually wrapping) LRU clock.
994
     *
995 996 997 998 999
     * Note that even if the counter wraps it's not a big problem,
     * everything will still work but some object will appear younger
     * to Redis. However for this to happen a given object should never be
     * touched for all the time needed to the counter to wrap, which is
     * not likely.
1000 1001
     *
     * Note that you can change the resolution altering the
A
antirez 已提交
1002
     * LRU_CLOCK_RESOLUTION define. */
1003 1004
    unsigned long lruclock = getLRUClock();
    atomicSet(server.lruclock,lruclock);
1005

1006 1007 1008 1009
    /* Record the max memory used since the server was started. */
    if (zmalloc_used_memory() > server.stat_peak_memory)
        server.stat_peak_memory = zmalloc_used_memory();

A
antirez 已提交
1010 1011 1012
    /* Sample the RSS here since this is a relatively slow call. */
    server.resident_set_size = zmalloc_get_rss();

1013 1014 1015
    /* We received a SIGTERM, shutting down here in a safe way, as it is
     * not ok doing so inside the signal handler. */
    if (server.shutdown_asap) {
1016
        if (prepareForShutdown(SHUTDOWN_NOFLAGS) == C_OK) exit(0);
A
antirez 已提交
1017
        serverLog(LL_WARNING,"SIGTERM received but errors trying to shut down the server, check the logs for more information");
1018
        server.shutdown_asap = 0;
1019 1020 1021
    }

    /* Show some info about non-empty databases */
1022 1023 1024 1025 1026 1027 1028 1029
    run_with_period(5000) {
        for (j = 0; j < server.dbnum; j++) {
            long long size, used, vkeys;

            size = dictSlots(server.db[j].dict);
            used = dictSize(server.db[j].dict);
            vkeys = dictSize(server.db[j].expires);
            if (used || vkeys) {
A
antirez 已提交
1030
                serverLog(LL_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
1031 1032
                /* dictPrintStats(server.dict); */
            }
1033 1034 1035 1036
        }
    }

    /* Show information about connected clients */
1037 1038
    if (!server.sentinel_mode) {
        run_with_period(5000) {
A
antirez 已提交
1039
            serverLog(LL_VERBOSE,
1040
                "%lu clients connected (%lu slaves), %zu bytes in use",
1041 1042 1043 1044
                listLength(server.clients)-listLength(server.slaves),
                listLength(server.slaves),
                zmalloc_used_memory());
        }
1045 1046
    }

1047 1048
    /* We need to do a few operations on clients asynchronously. */
    clientsCron();
1049

1050 1051 1052
    /* Handle background operations on Redis databases. */
    databasesCron();

1053 1054
    /* Start a scheduled AOF rewrite if this was requested by the user while
     * a BGSAVE was in progress. */
A
antirez 已提交
1055
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
1056
        server.aof_rewrite_scheduled)
1057 1058 1059 1060
    {
        rewriteAppendOnlyFileBackground();
    }

A
antirez 已提交
1061
    /* Check if a background saving or AOF rewrite in progress terminated. */
1062 1063 1064
    if (server.rdb_child_pid != -1 || server.aof_child_pid != -1 ||
        ldbPendingChildren())
    {
1065 1066 1067 1068
        int statloc;
        pid_t pid;

        if ((pid = wait3(&statloc,WNOHANG,NULL)) != 0) {
1069 1070
            int exitcode = WEXITSTATUS(statloc);
            int bysignal = 0;
1071

1072 1073
            if (WIFSIGNALED(statloc)) bysignal = WTERMSIG(statloc);

A
antirez 已提交
1074 1075 1076 1077 1078 1079 1080
            if (pid == -1) {
                serverLog(LL_WARNING,"wait3() returned an error: %s. "
                    "rdb_child_pid = %d, aof_child_pid = %d",
                    strerror(errno),
                    (int) server.rdb_child_pid,
                    (int) server.aof_child_pid);
            } else if (pid == server.rdb_child_pid) {
1081
                backgroundSaveDoneHandler(exitcode,bysignal);
1082
                if (!bysignal && exitcode == 0) receiveChildInfo();
1083
            } else if (pid == server.aof_child_pid) {
1084
                backgroundRewriteDoneHandler(exitcode,bysignal);
1085
                if (!bysignal && exitcode == 0) receiveChildInfo();
1086
            } else {
1087 1088 1089 1090 1091
                if (!ldbRemoveChild(pid)) {
                    serverLog(LL_WARNING,
                        "Warning, detected child with unmatched pid: %ld",
                        (long)pid);
                }
1092 1093
            }
            updateDictResizePolicy();
1094
            closeChildInfoPipe();
1095
        }
A
antirez 已提交
1096
    } else {
1097 1098
        /* If there is not a background saving/rewrite in progress check if
         * we have to save/rewrite now */
1099 1100 1101
         for (j = 0; j < server.saveparamslen; j++) {
            struct saveparam *sp = server.saveparams+j;

1102 1103 1104
            /* Save if we reached the given amount of changes,
             * the given amount of seconds, and if the latest bgsave was
             * successful or if, in case of an error, at least
A
antirez 已提交
1105
             * CONFIG_BGSAVE_RETRY_DELAY seconds already elapsed. */
1106
            if (server.dirty >= sp->changes &&
1107 1108
                server.unixtime-server.lastsave > sp->seconds &&
                (server.unixtime-server.lastbgsave_try >
A
antirez 已提交
1109
                 CONFIG_BGSAVE_RETRY_DELAY ||
1110
                 server.lastbgsave_status == C_OK))
1111
            {
A
antirez 已提交
1112
                serverLog(LL_NOTICE,"%d changes in %d seconds. Saving...",
1113
                    sp->changes, (int)sp->seconds);
1114 1115 1116
                rdbSaveInfo rsi, *rsiptr;
                rsiptr = rdbPopulateSaveInfo(&rsi);
                rdbSaveBackground(server.rdb_filename,rsiptr);
1117 1118 1119
                break;
            }
         }
1120 1121

         /* Trigger an AOF rewrite if needed */
A
antirez 已提交
1122
         if (server.rdb_child_pid == -1 &&
A
antirez 已提交
1123
             server.aof_child_pid == -1 &&
1124 1125
             server.aof_rewrite_perc &&
             server.aof_current_size > server.aof_rewrite_min_size)
1126
         {
1127 1128 1129 1130
            long long base = server.aof_rewrite_base_size ?
                            server.aof_rewrite_base_size : 1;
            long long growth = (server.aof_current_size*100/base) - 100;
            if (growth >= server.aof_rewrite_perc) {
A
antirez 已提交
1131
                serverLog(LL_NOTICE,"Starting automatic rewriting of AOF on %lld%% growth",growth);
1132 1133 1134
                rewriteAppendOnlyFileBackground();
            }
         }
1135 1136
    }

1137

1138 1139 1140 1141 1142 1143 1144 1145 1146
    /* AOF postponed flush: Try at every cron cycle if the slow fsync
     * completed. */
    if (server.aof_flush_postponed_start) flushAppendOnlyFile(0);

    /* AOF write errors: in this case we have a buffer to flush as well and
     * clear the AOF error in case of success to make the DB writable again,
     * however to try every second is enough in case of 'hz' is set to
     * an higher frequency. */
    run_with_period(1000) {
1147
        if (server.aof_last_write_status == C_ERR)
1148
            flushAppendOnlyFile(0);
1149
    }
1150

1151 1152 1153
    /* Close clients that need to be closed asynchronous */
    freeClientsInAsyncFreeQueue();

1154
    /* Clear the paused clients flag if needed. */
A
antirez 已提交
1155
    clientsArePaused(); /* Don't check return value, just use the side effect.*/
1156

1157 1158
    /* Replication cron function -- used to reconnect to master,
     * detect transfer failures, start background RDB transfers and so forth. */
1159
    run_with_period(1000) replicationCron();
1160

A
antirez 已提交
1161
    /* Run the Redis Cluster cron. */
1162
    run_with_period(100) {
1163 1164
        if (server.cluster_enabled) clusterCron();
    }
A
antirez 已提交
1165

A
antirez 已提交
1166
    /* Run the Sentinel timer if we are in sentinel mode. */
1167 1168 1169 1170
    run_with_period(100) {
        if (server.sentinel_mode) sentinelTimer();
    }

A
antirez 已提交
1171 1172 1173 1174 1175
    /* Cleanup expired MIGRATE cached sockets. */
    run_with_period(1000) {
        migrateCloseTimedoutSockets();
    }

1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
    /* Start a scheduled BGSAVE if the corresponding flag is set. This is
     * useful when we are forced to postpone a BGSAVE because an AOF
     * rewrite is in progress.
     *
     * Note: this code must be after the replicationCron() call above so
     * make sure when refactoring this file to keep this order. This is useful
     * because we want to give priority to RDB savings for replication. */
    if (server.rdb_child_pid == -1 && server.aof_child_pid == -1 &&
        server.rdb_bgsave_scheduled &&
        (server.unixtime-server.lastbgsave_try > CONFIG_BGSAVE_RETRY_DELAY ||
         server.lastbgsave_status == C_OK))
    {
1188 1189 1190
        rdbSaveInfo rsi, *rsiptr;
        rsiptr = rdbPopulateSaveInfo(&rsi);
        if (rdbSaveBackground(server.rdb_filename,rsiptr) == C_OK)
1191 1192 1193
            server.rdb_bgsave_scheduled = 0;
    }

1194
    server.cronloops++;
1195
    return 1000/server.hz;
1196 1197 1198 1199 1200 1201
}

/* This function gets called every time Redis is entering the
 * main loop of the event driven library, that is, before to sleep
 * for ready file descriptors. */
void beforeSleep(struct aeEventLoop *eventLoop) {
A
antirez 已提交
1202
    UNUSED(eventLoop);
1203

1204
    /* Call the Redis Cluster before sleep function. Note that this function
A
antirez 已提交
1205
     * may change the state of Redis Cluster (from ok to fail or vice versa),
1206 1207 1208 1209
     * so it's a good idea to call it before serving the unblocked clients
     * later in this function. */
    if (server.cluster_enabled) clusterBeforeSleep();

1210 1211 1212 1213
    /* Run a fast expire cycle (the called function will return
     * ASAP if a fast cycle is not needed). */
    if (server.active_expire_enabled && server.masterhost == NULL)
        activeExpireCycle(ACTIVE_EXPIRE_CYCLE_FAST);
1214

1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234
    /* Send all the slaves an ACK request if at least one client blocked
     * during the previous event loop iteration. */
    if (server.get_ack_from_slaves) {
        robj *argv[3];

        argv[0] = createStringObject("REPLCONF",8);
        argv[1] = createStringObject("GETACK",6);
        argv[2] = createStringObject("*",1); /* Not used argument. */
        replicationFeedSlaves(server.slaves, server.slaveseldb, argv, 3);
        decrRefCount(argv[0]);
        decrRefCount(argv[1]);
        decrRefCount(argv[2]);
        server.get_ack_from_slaves = 0;
    }

    /* Unblock all the clients blocked for synchronous replication
     * in WAIT. */
    if (listLength(server.clients_waiting_acks))
        processClientsWaitingReplicas();

1235 1236 1237 1238
    /* Check if there are clients unblocked by modules that implement
     * blocking commands. */
    moduleHandleBlockedClients();

1239
    /* Try to process pending commands for clients that were just unblocked. */
1240 1241
    if (listLength(server.unblocked_clients))
        processUnblockedClients();
1242

1243
    /* Write the AOF buffer on disk */
1244
    flushAppendOnlyFile(0);
1245 1246 1247

    /* Handle writes with pending output buffers. */
    handleClientsWithPendingWrites();
1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260

    /* Before we are going to sleep, let the threads access the dataset by
     * releasing the GIL. Redis main thread will not touch anything at this
     * time. */
    if (moduleCount()) moduleReleaseGIL();
}

/* This function is called immadiately after the event loop multiplexing
 * API returned, and the control is going to soon return to Redis by invoking
 * the different events callbacks. */
void afterSleep(struct aeEventLoop *eventLoop) {
    UNUSED(eventLoop);
    if (moduleCount()) moduleAcquireGIL();
1261 1262 1263 1264 1265 1266 1267
}

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

void createSharedObjects(void) {
    int j;

1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
    shared.crlf = createObject(OBJ_STRING,sdsnew("\r\n"));
    shared.ok = createObject(OBJ_STRING,sdsnew("+OK\r\n"));
    shared.err = createObject(OBJ_STRING,sdsnew("-ERR\r\n"));
    shared.emptybulk = createObject(OBJ_STRING,sdsnew("$0\r\n\r\n"));
    shared.czero = createObject(OBJ_STRING,sdsnew(":0\r\n"));
    shared.cone = createObject(OBJ_STRING,sdsnew(":1\r\n"));
    shared.cnegone = createObject(OBJ_STRING,sdsnew(":-1\r\n"));
    shared.nullbulk = createObject(OBJ_STRING,sdsnew("$-1\r\n"));
    shared.nullmultibulk = createObject(OBJ_STRING,sdsnew("*-1\r\n"));
    shared.emptymultibulk = createObject(OBJ_STRING,sdsnew("*0\r\n"));
    shared.pong = createObject(OBJ_STRING,sdsnew("+PONG\r\n"));
    shared.queued = createObject(OBJ_STRING,sdsnew("+QUEUED\r\n"));
    shared.emptyscan = createObject(OBJ_STRING,sdsnew("*2\r\n$1\r\n0\r\n*0\r\n"));
    shared.wrongtypeerr = createObject(OBJ_STRING,sdsnew(
1282
        "-WRONGTYPE Operation against a key holding the wrong kind of value\r\n"));
1283
    shared.nokeyerr = createObject(OBJ_STRING,sdsnew(
1284
        "-ERR no such key\r\n"));
1285
    shared.syntaxerr = createObject(OBJ_STRING,sdsnew(
1286
        "-ERR syntax error\r\n"));
1287
    shared.sameobjecterr = createObject(OBJ_STRING,sdsnew(
1288
        "-ERR source and destination objects are the same\r\n"));
1289
    shared.outofrangeerr = createObject(OBJ_STRING,sdsnew(
1290
        "-ERR index out of range\r\n"));
1291
    shared.noscripterr = createObject(OBJ_STRING,sdsnew(
A
antirez 已提交
1292
        "-NOSCRIPT No matching script. Please use EVAL.\r\n"));
1293
    shared.loadingerr = createObject(OBJ_STRING,sdsnew(
1294
        "-LOADING Redis is loading the dataset in memory\r\n"));
1295
    shared.slowscripterr = createObject(OBJ_STRING,sdsnew(
1296
        "-BUSY Redis is busy running a script. You can only call SCRIPT KILL or SHUTDOWN NOSAVE.\r\n"));
1297
    shared.masterdownerr = createObject(OBJ_STRING,sdsnew(
1298
        "-MASTERDOWN Link with MASTER is down and slave-serve-stale-data is set to 'no'.\r\n"));
1299
    shared.bgsaveerr = createObject(OBJ_STRING,sdsnew(
1300
        "-MISCONF Redis is configured to save RDB snapshots, but it is currently not able to persist on disk. Commands that may modify the data set are disabled, because this instance is configured to report errors during writes if RDB snapshotting fails (stop-writes-on-bgsave-error option). Please check the Redis logs for details about the RDB error.\r\n"));
1301
    shared.roslaveerr = createObject(OBJ_STRING,sdsnew(
1302
        "-READONLY You can't write against a read only slave.\r\n"));
1303
    shared.noautherr = createObject(OBJ_STRING,sdsnew(
1304
        "-NOAUTH Authentication required.\r\n"));
1305
    shared.oomerr = createObject(OBJ_STRING,sdsnew(
1306
        "-OOM command not allowed when used memory > 'maxmemory'.\r\n"));
1307
    shared.execaborterr = createObject(OBJ_STRING,sdsnew(
A
antirez 已提交
1308
        "-EXECABORT Transaction discarded because of previous errors.\r\n"));
1309
    shared.noreplicaserr = createObject(OBJ_STRING,sdsnew(
1310
        "-NOREPLICAS Not enough good slaves to write.\r\n"));
1311
    shared.busykeyerr = createObject(OBJ_STRING,sdsnew(
1312
        "-BUSYKEY Target key name already exists.\r\n"));
1313 1314 1315
    shared.space = createObject(OBJ_STRING,sdsnew(" "));
    shared.colon = createObject(OBJ_STRING,sdsnew(":"));
    shared.plus = createObject(OBJ_STRING,sdsnew("+"));
1316

A
antirez 已提交
1317
    for (j = 0; j < PROTO_SHARED_SELECT_CMDS; j++) {
1318 1319 1320 1321
        char dictid_str[64];
        int dictid_len;

        dictid_len = ll2string(dictid_str,sizeof(dictid_str),j);
1322
        shared.select[j] = createObject(OBJ_STRING,
1323 1324 1325
            sdscatprintf(sdsempty(),
                "*2\r\n$6\r\nSELECT\r\n$%d\r\n%s\r\n",
                dictid_len, dictid_str));
1326
    }
1327 1328 1329 1330 1331 1332
    shared.messagebulk = createStringObject("$7\r\nmessage\r\n",13);
    shared.pmessagebulk = createStringObject("$8\r\npmessage\r\n",14);
    shared.subscribebulk = createStringObject("$9\r\nsubscribe\r\n",15);
    shared.unsubscribebulk = createStringObject("$11\r\nunsubscribe\r\n",18);
    shared.psubscribebulk = createStringObject("$10\r\npsubscribe\r\n",17);
    shared.punsubscribebulk = createStringObject("$12\r\npunsubscribe\r\n",19);
1333
    shared.del = createStringObject("DEL",3);
1334
    shared.unlink = createStringObject("UNLINK",6);
1335 1336
    shared.rpop = createStringObject("RPOP",4);
    shared.lpop = createStringObject("LPOP",4);
1337
    shared.lpush = createStringObject("LPUSH",5);
A
antirez 已提交
1338
    for (j = 0; j < OBJ_SHARED_INTEGERS; j++) {
1339 1340
        shared.integers[j] =
            makeObjectShared(createObject(OBJ_STRING,(void*)(long)j));
1341
        shared.integers[j]->encoding = OBJ_ENCODING_INT;
1342
    }
A
antirez 已提交
1343
    for (j = 0; j < OBJ_SHARED_BULKHDR_LEN; j++) {
1344
        shared.mbulkhdr[j] = createObject(OBJ_STRING,
1345
            sdscatprintf(sdsempty(),"*%d\r\n",j));
1346
        shared.bulkhdr[j] = createObject(OBJ_STRING,
1347 1348
            sdscatprintf(sdsempty(),"$%d\r\n",j));
    }
1349 1350 1351 1352
    /* The following two shared objects, minstring and maxstrings, are not
     * actually used for their value but as a special object meaning
     * respectively the minimum possible string and the maximum possible
     * string in string comparisons for the ZRANGEBYLEX command. */
1353 1354
    shared.minstring = sdsnew("minstring");
    shared.maxstring = sdsnew("maxstring");
1355 1356
}

1357
void initServerConfig(void) {
1358 1359
    int j;

1360 1361 1362 1363
    pthread_mutex_init(&server.next_client_id_mutex,NULL);
    pthread_mutex_init(&server.lruclock_mutex,NULL);
    pthread_mutex_init(&server.unixtime_mutex,NULL);

A
antirez 已提交
1364
    getRandomHexChars(server.runid,CONFIG_RUN_ID_SIZE);
1365 1366
    server.runid[CONFIG_RUN_ID_SIZE] = '\0';
    changeReplicationId();
A
antirez 已提交
1367
    clearReplicationId2();
1368
    server.configfile = NULL;
A
antirez 已提交
1369
    server.executable = NULL;
1370
    server.hz = CONFIG_DEFAULT_HZ;
1371
    server.arch_bits = (sizeof(long) == 8) ? 64 : 32;
A
antirez 已提交
1372 1373
    server.port = CONFIG_DEFAULT_SERVER_PORT;
    server.tcp_backlog = CONFIG_DEFAULT_TCP_BACKLOG;
A
antirez 已提交
1374
    server.bindaddr_count = 0;
1375
    server.unixsocket = NULL;
1376
    server.unixsocketperm = CONFIG_DEFAULT_UNIX_SOCKET_PERM;
1377
    server.ipfd_count = 0;
1378
    server.sofd = -1;
1379
    server.protected_mode = CONFIG_DEFAULT_PROTECTED_MODE;
1380 1381
    server.dbnum = CONFIG_DEFAULT_DBNUM;
    server.verbosity = CONFIG_DEFAULT_VERBOSITY;
A
antirez 已提交
1382
    server.maxidletime = CONFIG_DEFAULT_CLIENT_TIMEOUT;
1383
    server.tcpkeepalive = CONFIG_DEFAULT_TCP_KEEPALIVE;
A
antirez 已提交
1384
    server.active_expire_enabled = 1;
O
oranagra 已提交
1385 1386 1387 1388 1389 1390
    server.active_defrag_enabled = CONFIG_DEFAULT_ACTIVE_DEFRAG;
    server.active_defrag_ignore_bytes = CONFIG_DEFAULT_DEFRAG_IGNORE_BYTES;
    server.active_defrag_threshold_lower = CONFIG_DEFAULT_DEFRAG_THRESHOLD_LOWER;
    server.active_defrag_threshold_upper = CONFIG_DEFAULT_DEFRAG_THRESHOLD_UPPER;
    server.active_defrag_cycle_min = CONFIG_DEFAULT_DEFRAG_CYCLE_MIN;
    server.active_defrag_cycle_max = CONFIG_DEFAULT_DEFRAG_CYCLE_MAX;
A
antirez 已提交
1391
    server.client_max_querybuf_len = PROTO_MAX_QUERYBUF_LEN;
1392
    server.saveparams = NULL;
1393
    server.loading = 0;
1394 1395 1396
    server.logfile = zstrdup(CONFIG_DEFAULT_LOGFILE);
    server.syslog_enabled = CONFIG_DEFAULT_SYSLOG_ENABLED;
    server.syslog_ident = zstrdup(CONFIG_DEFAULT_SYSLOG_IDENT);
J
Jonah H. Harris 已提交
1397
    server.syslog_facility = LOG_LOCAL0;
1398
    server.daemonize = CONFIG_DEFAULT_DAEMONIZE;
1399
    server.supervised = 0;
A
antirez 已提交
1400 1401
    server.supervised_mode = SUPERVISED_NONE;
    server.aof_state = AOF_OFF;
1402 1403
    server.aof_fsync = CONFIG_DEFAULT_AOF_FSYNC;
    server.aof_no_fsync_on_rewrite = CONFIG_DEFAULT_AOF_NO_FSYNC_ON_REWRITE;
A
antirez 已提交
1404 1405
    server.aof_rewrite_perc = AOF_REWRITE_PERC;
    server.aof_rewrite_min_size = AOF_REWRITE_MIN_SIZE;
1406 1407
    server.aof_rewrite_base_size = 0;
    server.aof_rewrite_scheduled = 0;
A
antirez 已提交
1408
    server.aof_last_fsync = time(NULL);
1409 1410
    server.aof_rewrite_time_last = -1;
    server.aof_rewrite_time_start = -1;
1411
    server.aof_lastbgrewrite_status = C_OK;
1412
    server.aof_delayed_fsync = 0;
A
antirez 已提交
1413 1414
    server.aof_fd = -1;
    server.aof_selected_db = -1; /* Make sure the first time will not match */
1415
    server.aof_flush_postponed_start = 0;
1416 1417
    server.aof_rewrite_incremental_fsync = CONFIG_DEFAULT_AOF_REWRITE_INCREMENTAL_FSYNC;
    server.aof_load_truncated = CONFIG_DEFAULT_AOF_LOAD_TRUNCATED;
A
antirez 已提交
1418
    server.aof_use_rdb_preamble = CONFIG_DEFAULT_AOF_USE_RDB_PREAMBLE;
1419
    server.pidfile = NULL;
1420 1421
    server.rdb_filename = zstrdup(CONFIG_DEFAULT_RDB_FILENAME);
    server.aof_filename = zstrdup(CONFIG_DEFAULT_AOF_FILENAME);
1422
    server.requirepass = NULL;
1423 1424 1425 1426
    server.rdb_compression = CONFIG_DEFAULT_RDB_COMPRESSION;
    server.rdb_checksum = CONFIG_DEFAULT_RDB_CHECKSUM;
    server.stop_writes_on_bgsave_err = CONFIG_DEFAULT_STOP_WRITES_ON_BGSAVE_ERROR;
    server.activerehashing = CONFIG_DEFAULT_ACTIVE_REHASHING;
O
oranagra 已提交
1427
    server.active_defrag_running = 0;
A
antirez 已提交
1428
    server.notify_keyspace_events = 0;
1429
    server.maxclients = CONFIG_DEFAULT_MAX_CLIENTS;
A
antirez 已提交
1430 1431 1432
    server.blocked_clients = 0;
    memset(server.blocked_clients_by_type,0,
           sizeof(server.blocked_clients_by_type));
1433 1434 1435
    server.maxmemory = CONFIG_DEFAULT_MAXMEMORY;
    server.maxmemory_policy = CONFIG_DEFAULT_MAXMEMORY_POLICY;
    server.maxmemory_samples = CONFIG_DEFAULT_MAXMEMORY_SAMPLES;
1436 1437
    server.lfu_log_factor = CONFIG_DEFAULT_LFU_LOG_FACTOR;
    server.lfu_decay_time = CONFIG_DEFAULT_LFU_DECAY_TIME;
1438 1439 1440 1441 1442 1443 1444
    server.hash_max_ziplist_entries = OBJ_HASH_MAX_ZIPLIST_ENTRIES;
    server.hash_max_ziplist_value = OBJ_HASH_MAX_ZIPLIST_VALUE;
    server.list_max_ziplist_size = OBJ_LIST_MAX_ZIPLIST_SIZE;
    server.list_compress_depth = OBJ_LIST_COMPRESS_DEPTH;
    server.set_max_intset_entries = OBJ_SET_MAX_INTSET_ENTRIES;
    server.zset_max_ziplist_entries = OBJ_ZSET_MAX_ZIPLIST_ENTRIES;
    server.zset_max_ziplist_value = OBJ_ZSET_MAX_ZIPLIST_VALUE;
1445
    server.hll_sparse_max_bytes = CONFIG_DEFAULT_HLL_SPARSE_MAX_BYTES;
1446
    server.shutdown_asap = 0;
A
antirez 已提交
1447
    server.cluster_enabled = 0;
A
antirez 已提交
1448 1449 1450 1451
    server.cluster_node_timeout = CLUSTER_DEFAULT_NODE_TIMEOUT;
    server.cluster_migration_barrier = CLUSTER_DEFAULT_MIGRATION_BARRIER;
    server.cluster_slave_validity_factor = CLUSTER_DEFAULT_SLAVE_VALIDITY;
    server.cluster_require_full_coverage = CLUSTER_DEFAULT_REQUIRE_FULL_COVERAGE;
1452
    server.cluster_configfile = zstrdup(CONFIG_DEFAULT_CLUSTER_CONFIG_FILE);
1453 1454 1455
    server.cluster_announce_ip = CONFIG_DEFAULT_CLUSTER_ANNOUNCE_IP;
    server.cluster_announce_port = CONFIG_DEFAULT_CLUSTER_ANNOUNCE_PORT;
    server.cluster_announce_bus_port = CONFIG_DEFAULT_CLUSTER_ANNOUNCE_BUS_PORT;
A
antirez 已提交
1456
    server.migrate_cached_sockets = dictCreate(&migrateCacheDictType,NULL);
1457
    server.next_client_id = 1; /* Client IDs, start from 1 .*/
1458
    server.loading_process_events_interval_bytes = (1024*1024*2);
1459 1460 1461
    server.lazyfree_lazy_eviction = CONFIG_DEFAULT_LAZYFREE_LAZY_EVICTION;
    server.lazyfree_lazy_expire = CONFIG_DEFAULT_LAZYFREE_LAZY_EXPIRE;
    server.lazyfree_lazy_server_del = CONFIG_DEFAULT_LAZYFREE_LAZY_SERVER_DEL;
1462
    server.always_show_logo = CONFIG_DEFAULT_ALWAYS_SHOW_LOGO;
1463
    server.lua_time_limit = LUA_SCRIPT_TIME_LIMIT;
1464

1465 1466
    unsigned int lruclock = getLRUClock();
    atomicSet(server.lruclock,lruclock);
1467 1468 1469 1470 1471
    resetServerSaveParams();

    appendServerSaveParams(60*60,1);  /* save after 1 hour and 1 change */
    appendServerSaveParams(300,100);  /* save after 5 minutes and 100 changes */
    appendServerSaveParams(60,10000); /* save after 1 minute and 10000 changes */
1472

1473 1474 1475 1476 1477
    /* Replication related */
    server.masterauth = NULL;
    server.masterhost = NULL;
    server.masterport = 6379;
    server.master = NULL;
1478
    server.cached_master = NULL;
1479
    server.master_initial_offset = -1;
A
antirez 已提交
1480 1481
    server.repl_state = REPL_STATE_NONE;
    server.repl_syncio_timeout = CONFIG_REPL_SYNCIO_TIMEOUT;
1482 1483
    server.repl_serve_stale_data = CONFIG_DEFAULT_SLAVE_SERVE_STALE_DATA;
    server.repl_slave_ro = CONFIG_DEFAULT_SLAVE_READ_ONLY;
1484
    server.repl_slave_lazy_flush = CONFIG_DEFAULT_SLAVE_LAZY_FLUSH;
1485
    server.repl_down_since = 0; /* Never connected, repl is down since EVER. */
1486 1487 1488
    server.repl_disable_tcp_nodelay = CONFIG_DEFAULT_REPL_DISABLE_TCP_NODELAY;
    server.repl_diskless_sync = CONFIG_DEFAULT_REPL_DISKLESS_SYNC;
    server.repl_diskless_sync_delay = CONFIG_DEFAULT_REPL_DISKLESS_SYNC_DELAY;
1489 1490 1491 1492
    server.repl_ping_slave_period = CONFIG_DEFAULT_REPL_PING_SLAVE_PERIOD;
    server.repl_timeout = CONFIG_DEFAULT_REPL_TIMEOUT;
    server.repl_min_slaves_to_write = CONFIG_DEFAULT_MIN_SLAVES_TO_WRITE;
    server.repl_min_slaves_max_lag = CONFIG_DEFAULT_MIN_SLAVES_MAX_LAG;
1493
    server.slave_priority = CONFIG_DEFAULT_SLAVE_PRIORITY;
1494 1495
    server.slave_announce_ip = CONFIG_DEFAULT_SLAVE_ANNOUNCE_IP;
    server.slave_announce_port = CONFIG_DEFAULT_SLAVE_ANNOUNCE_PORT;
1496 1497 1498 1499
    server.master_repl_offset = 0;

    /* Replication partial resync backlog */
    server.repl_backlog = NULL;
1500
    server.repl_backlog_size = CONFIG_DEFAULT_REPL_BACKLOG_SIZE;
1501 1502 1503
    server.repl_backlog_histlen = 0;
    server.repl_backlog_idx = 0;
    server.repl_backlog_off = 0;
1504
    server.repl_backlog_time_limit = CONFIG_DEFAULT_REPL_BACKLOG_TIME_LIMIT;
1505
    server.repl_no_slaves_since = time(NULL);
1506

1507
    /* Client output buffer limits */
A
antirez 已提交
1508
    for (j = 0; j < CLIENT_TYPE_OBUF_COUNT; j++)
1509
        server.client_obuf_limits[j] = clientBufferLimitsDefaults[j];
1510

1511 1512 1513 1514 1515
    /* Double constants initialization */
    R_Zero = 0.0;
    R_PosInf = 1.0/R_Zero;
    R_NegInf = -1.0/R_Zero;
    R_Nan = R_Zero/R_Zero;
1516

G
guiquanz 已提交
1517
    /* Command table -- we initiialize it here as it is part of the
1518 1519 1520
     * initial configuration, since command names may be changed via
     * redis.conf using the rename-command directive. */
    server.commands = dictCreate(&commandTableDictType,NULL);
1521
    server.orig_commands = dictCreate(&commandTableDictType,NULL);
1522 1523 1524
    populateCommandTable();
    server.delCommand = lookupCommandByCString("del");
    server.multiCommand = lookupCommandByCString("multi");
1525
    server.lpushCommand = lookupCommandByCString("lpush");
1526 1527
    server.lpopCommand = lookupCommandByCString("lpop");
    server.rpopCommand = lookupCommandByCString("rpop");
1528
    server.sremCommand = lookupCommandByCString("srem");
1529
    server.execCommand = lookupCommandByCString("exec");
1530 1531
    server.expireCommand = lookupCommandByCString("expire");
    server.pexpireCommand = lookupCommandByCString("pexpire");
1532

1533
    /* Slow log */
A
antirez 已提交
1534 1535
    server.slowlog_log_slower_than = CONFIG_DEFAULT_SLOWLOG_LOG_SLOWER_THAN;
    server.slowlog_max_len = CONFIG_DEFAULT_SLOWLOG_MAX_LEN;
A
antirez 已提交
1536

1537
    /* Latency monitor */
1538
    server.latency_monitor_threshold = CONFIG_DEFAULT_LATENCY_MONITOR_THRESHOLD;
1539

A
antirez 已提交
1540
    /* Debugging */
A
antirez 已提交
1541 1542 1543 1544
    server.assert_failed = "<no assertion failed>";
    server.assert_file = "<no file>";
    server.assert_line = 0;
    server.bug_report_start = 0;
A
antirez 已提交
1545
    server.watchdog_period = 0;
1546 1547
}

A
antirez 已提交
1548 1549 1550 1551 1552
extern char **environ;

/* Restart the server, executing the same executable that started this
 * instance, with the same arguments and configuration file.
 *
1553 1554 1555
 * The function is designed to directly call execve() so that the new
 * server instance will retain the PID of the previous one.
 *
A
antirez 已提交
1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569
 * The list of flags, that may be bitwise ORed together, alter the
 * behavior of this function:
 *
 * RESTART_SERVER_NONE              No flags.
 * RESTART_SERVER_GRACEFULLY        Do a proper shutdown before restarting.
 * RESTART_SERVER_CONFIG_REWRITE    Rewrite the config file before restarting.
 *
 * On success the function does not return, because the process turns into
 * a different process. On error C_ERR is returned. */
int restartServer(int flags, mstime_t delay) {
    int j;

    /* Check if we still have accesses to the executable that started this
     * server instance. */
1570 1571 1572 1573 1574
    if (access(server.executable,X_OK) == -1) {
        serverLog(LL_WARNING,"Can't restart: this process has no "
                             "permissions to execute %s", server.executable);
        return C_ERR;
    }
A
antirez 已提交
1575 1576 1577 1578

    /* Config rewriting. */
    if (flags & RESTART_SERVER_CONFIG_REWRITE &&
        server.configfile &&
1579 1580 1581 1582 1583 1584
        rewriteConfig(server.configfile) == -1)
    {
        serverLog(LL_WARNING,"Can't restart: configuration rewrite process "
                             "failed");
        return C_ERR;
    }
A
antirez 已提交
1585 1586 1587

    /* Perform a proper shutdown. */
    if (flags & RESTART_SERVER_GRACEFULLY &&
1588 1589 1590 1591 1592
        prepareForShutdown(SHUTDOWN_NOFLAGS) != C_OK)
    {
        serverLog(LL_WARNING,"Can't restart: error preparing for shutdown");
        return C_ERR;
    }
A
antirez 已提交
1593 1594 1595

    /* Close all file descriptors, with the exception of stdin, stdout, strerr
     * which are useful if we restart a Redis server which is not daemonized. */
1596 1597 1598 1599 1600
    for (j = 3; j < (int)server.maxclients + 1024; j++) {
        /* Test the descriptor validity before closing it, otherwise
         * Valgrind issues a warning on close(). */
        if (fcntl(j,F_GETFD) != -1) close(j);
    }
A
antirez 已提交
1601 1602 1603

    /* Execute the server with the original command line. */
    if (delay) usleep(delay*1000);
1604 1605
    zfree(server.exec_argv[0]);
    server.exec_argv[0] = zstrdup(server.executable);
A
antirez 已提交
1606 1607 1608 1609 1610 1611 1612 1613
    execve(server.executable,server.exec_argv,environ);

    /* If an error occurred here, there is nothing we can do, but exit. */
    _exit(1);

    return C_ERR; /* Never reached. */
}

1614
/* This function will try to raise the max number of open files accordingly to
1615
 * the configured max number of clients. It also reserves a number of file
A
antirez 已提交
1616
 * descriptors (CONFIG_MIN_RESERVED_FDS) for extra operations of
1617
 * persistence, listening sockets, log files and so forth.
1618 1619 1620 1621 1622
 *
 * If it will not be possible to set the limit accordingly to the configured
 * max number of clients, the function will do the reverse setting
 * server.maxclients to the value that we can actually handle. */
void adjustOpenFilesLimit(void) {
A
antirez 已提交
1623
    rlim_t maxfiles = server.maxclients+CONFIG_MIN_RESERVED_FDS;
1624 1625 1626
    struct rlimit limit;

    if (getrlimit(RLIMIT_NOFILE,&limit) == -1) {
A
antirez 已提交
1627
        serverLog(LL_WARNING,"Unable to obtain the current NOFILE limit (%s), assuming 1024 and setting the max clients configuration accordingly.",
1628
            strerror(errno));
A
antirez 已提交
1629
        server.maxclients = 1024-CONFIG_MIN_RESERVED_FDS;
1630 1631 1632 1633 1634 1635
    } else {
        rlim_t oldlimit = limit.rlim_cur;

        /* Set the max number of files if the current limit is not enough
         * for our needs. */
        if (oldlimit < maxfiles) {
1636
            rlim_t bestlimit;
A
antirez 已提交
1637 1638 1639 1640
            int setrlimit_error = 0;

            /* Try to set the file limit to match 'maxfiles' or at least
             * to the higher value supported less than maxfiles. */
1641 1642
            bestlimit = maxfiles;
            while(bestlimit > oldlimit) {
1643
                rlim_t decr_step = 16;
A
antirez 已提交
1644

1645 1646
                limit.rlim_cur = bestlimit;
                limit.rlim_max = bestlimit;
1647
                if (setrlimit(RLIMIT_NOFILE,&limit) != -1) break;
A
antirez 已提交
1648 1649
                setrlimit_error = errno;

1650
                /* We failed to set file limit to 'bestlimit'. Try with a
A
antirez 已提交
1651
                 * smaller limit decrementing by a few FDs per iteration. */
1652 1653
                if (bestlimit < decr_step) break;
                bestlimit -= decr_step;
1654
            }
A
antirez 已提交
1655 1656 1657

            /* Assume that the limit we get initially is still valid if
             * our last try was even lower. */
1658
            if (bestlimit < oldlimit) bestlimit = oldlimit;
A
antirez 已提交
1659

1660
            if (bestlimit < maxfiles) {
1661
                unsigned int old_maxclients = server.maxclients;
1662 1663 1664 1665
                server.maxclients = bestlimit-CONFIG_MIN_RESERVED_FDS;
                /* maxclients is unsigned so may overflow: in order
                 * to check if maxclients is now logically less than 1
                 * we test indirectly via bestlimit. */
1666
                if (bestlimit <= CONFIG_MIN_RESERVED_FDS) {
A
antirez 已提交
1667
                    serverLog(LL_WARNING,"Your current 'ulimit -n' "
1668
                        "of %llu is not enough for the server to start. "
1669
                        "Please increase your open file limit to at least "
A
antirez 已提交
1670 1671 1672
                        "%llu. Exiting.",
                        (unsigned long long) oldlimit,
                        (unsigned long long) maxfiles);
1673 1674
                    exit(1);
                }
A
antirez 已提交
1675
                serverLog(LL_WARNING,"You requested maxclients of %d "
1676
                    "requiring at least %llu max file descriptors.",
A
antirez 已提交
1677 1678
                    old_maxclients,
                    (unsigned long long) maxfiles);
1679
                serverLog(LL_WARNING,"Server can't set maximum open files "
1680
                    "to %llu because of OS error: %s.",
A
antirez 已提交
1681
                    (unsigned long long) maxfiles, strerror(setrlimit_error));
A
antirez 已提交
1682
                serverLog(LL_WARNING,"Current maximum open files is %llu. "
1683 1684 1685
                    "maxclients has been reduced to %d to compensate for "
                    "low ulimit. "
                    "If you need higher maxclients increase 'ulimit -n'.",
1686
                    (unsigned long long) bestlimit, server.maxclients);
1687
            } else {
A
antirez 已提交
1688
                serverLog(LL_NOTICE,"Increased maximum number of open files "
1689
                    "to %llu (it was originally set to %llu).",
A
antirez 已提交
1690 1691
                    (unsigned long long) maxfiles,
                    (unsigned long long) oldlimit);
1692 1693 1694 1695 1696
            }
        }
    }
}

1697 1698 1699 1700 1701 1702 1703 1704 1705 1706
/* Check that server.tcp_backlog can be actually enforced in Linux according
 * to the value of /proc/sys/net/core/somaxconn, or warn about it. */
void checkTcpBacklogSettings(void) {
#ifdef HAVE_PROC_SOMAXCONN
    FILE *fp = fopen("/proc/sys/net/core/somaxconn","r");
    char buf[1024];
    if (!fp) return;
    if (fgets(buf,sizeof(buf),fp) != NULL) {
        int somaxconn = atoi(buf);
        if (somaxconn > 0 && somaxconn < server.tcp_backlog) {
A
antirez 已提交
1707
            serverLog(LL_WARNING,"WARNING: The TCP backlog setting of %d cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of %d.", server.tcp_backlog, somaxconn);
1708 1709 1710 1711 1712 1713
        }
    }
    fclose(fp);
#endif
}

1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724
/* Initialize a set of file descriptors to listen to the specified 'port'
 * binding the addresses specified in the Redis server configuration.
 *
 * The listening file descriptors are stored in the integer array 'fds'
 * and their number is set in '*count'.
 *
 * The addresses to bind are specified in the global server.bindaddr array
 * and their number is server.bindaddr_count. If the server configuration
 * contains no specific addresses to bind, this function will try to
 * bind * (all addresses) for both the IPv4 and IPv6 protocols.
 *
1725
 * On success the function returns C_OK.
1726
 *
1727
 * On error the function returns C_ERR. For the function to be on
1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739
 * error, at least one of the server.bindaddr addresses was
 * impossible to bind, or no bind addresses were specified in the server
 * configuration but the function is not able to bind * for at least
 * one of the IPv4 or IPv6 protocols. */
int listenToPort(int port, int *fds, int *count) {
    int j;

    /* Force binding of 0.0.0.0 if no bind address is specified, always
     * entering the loop if j == 0. */
    if (server.bindaddr_count == 0) server.bindaddr[0] = NULL;
    for (j = 0; j < server.bindaddr_count || j == 0; j++) {
        if (server.bindaddr[j] == NULL) {
1740
            int unsupported = 0;
1741 1742
            /* Bind * for both IPv6 and IPv4, we enter here only if
             * server.bindaddr_count == 0. */
1743 1744
            fds[*count] = anetTcp6Server(server.neterr,port,NULL,
                server.tcp_backlog);
A
antirez 已提交
1745 1746 1747
            if (fds[*count] != ANET_ERR) {
                anetNonBlock(NULL,fds[*count]);
                (*count)++;
A
antirez 已提交
1748
            } else if (errno == EAFNOSUPPORT) {
1749
                unsupported++;
A
antirez 已提交
1750 1751
                serverLog(LL_WARNING,"Not listening to IPv6: unsupproted");
            }
1752

A
antirez 已提交
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
            if (*count == 1 || unsupported) {
                /* Bind the IPv4 address as well. */
                fds[*count] = anetTcpServer(server.neterr,port,NULL,
                    server.tcp_backlog);
                if (fds[*count] != ANET_ERR) {
                    anetNonBlock(NULL,fds[*count]);
                    (*count)++;
                } else if (errno == EAFNOSUPPORT) {
                    unsupported++;
                    serverLog(LL_WARNING,"Not listening to IPv4: unsupproted");
                }
            }
1765
            /* Exit the loop if we were able to bind * on IPv4 and IPv6,
1766 1767
             * otherwise fds[*count] will be ANET_ERR and we'll print an
             * error and return to the caller with an error. */
1768
            if (*count + unsupported == 2) break;
1769 1770
        } else if (strchr(server.bindaddr[j],':')) {
            /* Bind IPv6 address. */
1771 1772
            fds[*count] = anetTcp6Server(server.neterr,port,server.bindaddr[j],
                server.tcp_backlog);
1773 1774
        } else {
            /* Bind IPv4 address. */
1775 1776
            fds[*count] = anetTcpServer(server.neterr,port,server.bindaddr[j],
                server.tcp_backlog);
1777 1778
        }
        if (fds[*count] == ANET_ERR) {
A
antirez 已提交
1779
            serverLog(LL_WARNING,
1780 1781
                "Creating Server TCP listening socket %s:%d: %s",
                server.bindaddr[j] ? server.bindaddr[j] : "*",
1782
                port, server.neterr);
1783
            return C_ERR;
1784
        }
A
antirez 已提交
1785
        anetNonBlock(NULL,fds[*count]);
1786 1787
        (*count)++;
    }
1788
    return C_OK;
1789 1790
}

1791 1792 1793 1794
/* Resets the stats that we expose via INFO or other means that we want
 * to reset via CONFIG RESETSTAT. The function is also used in order to
 * initialize these fields in initServer() at server startup. */
void resetServerStats(void) {
1795 1796
    int j;

1797 1798 1799 1800 1801 1802
    server.stat_numcommands = 0;
    server.stat_numconnections = 0;
    server.stat_expiredkeys = 0;
    server.stat_evictedkeys = 0;
    server.stat_keyspace_misses = 0;
    server.stat_keyspace_hits = 0;
O
oranagra 已提交
1803 1804 1805 1806
    server.stat_active_defrag_hits = 0;
    server.stat_active_defrag_misses = 0;
    server.stat_active_defrag_key_hits = 0;
    server.stat_active_defrag_key_misses = 0;
1807
    server.stat_fork_time = 0;
1808
    server.stat_fork_rate = 0;
1809 1810 1811 1812
    server.stat_rejected_conn = 0;
    server.stat_sync_full = 0;
    server.stat_sync_partial_ok = 0;
    server.stat_sync_partial_err = 0;
A
antirez 已提交
1813
    for (j = 0; j < STATS_METRIC_COUNT; j++) {
1814 1815 1816 1817 1818 1819 1820 1821
        server.inst_metric[j].idx = 0;
        server.inst_metric[j].last_sample_time = mstime();
        server.inst_metric[j].last_sample_count = 0;
        memset(server.inst_metric[j].samples,0,
            sizeof(server.inst_metric[j].samples));
    }
    server.stat_net_input_bytes = 0;
    server.stat_net_output_bytes = 0;
T
Tom Kiemes 已提交
1822
    server.aof_delayed_fsync = 0;
1823 1824
}

1825
void initServer(void) {
1826 1827 1828 1829
    int j;

    signal(SIGHUP, SIG_IGN);
    signal(SIGPIPE, SIG_IGN);
1830
    setupSignalHandlers();
1831

J
Jonah H. Harris 已提交
1832 1833 1834 1835 1836
    if (server.syslog_enabled) {
        openlog(server.syslog_ident, LOG_PID | LOG_NDELAY | LOG_NOWAIT,
            server.syslog_facility);
    }

A
antirez 已提交
1837
    server.pid = getpid();
1838
    server.current_client = NULL;
1839
    server.clients = listCreate();
1840
    server.clients_to_close = listCreate();
1841 1842
    server.slaves = listCreate();
    server.monitors = listCreate();
1843
    server.clients_pending_write = listCreate();
1844
    server.slaveseldb = -1; /* Force to emit the first SELECT command. */
1845
    server.unblocked_clients = listCreate();
1846
    server.ready_keys = listCreate();
1847 1848
    server.clients_waiting_acks = listCreate();
    server.get_ack_from_slaves = 0;
1849
    server.clients_paused = 0;
1850
    server.system_memory_size = zmalloc_get_memory_size();
1851

1852
    createSharedObjects();
1853
    adjustOpenFilesLimit();
A
antirez 已提交
1854
    server.el = aeCreateEventLoop(server.maxclients+CONFIG_FDSET_INCR);
1855 1856 1857 1858 1859 1860
    if (server.el == NULL) {
        serverLog(LL_WARNING,
            "Failed creating the event loop. Error message: '%s'",
            strerror(errno));
        exit(1);
    }
1861
    server.db = zmalloc(sizeof(redisDb)*server.dbnum);
1862

1863
    /* Open the TCP listening socket for the user commands. */
1864
    if (server.port != 0 &&
1865
        listenToPort(server.port,server.ipfd,&server.ipfd_count) == C_ERR)
1866
        exit(1);
1867 1868

    /* Open the listening Unix domain socket. */
1869 1870
    if (server.unixsocket != NULL) {
        unlink(server.unixsocket); /* don't care if this fails */
1871 1872
        server.sofd = anetUnixServer(server.neterr,server.unixsocket,
            server.unixsocketperm, server.tcp_backlog);
1873
        if (server.sofd == ANET_ERR) {
A
antirez 已提交
1874
            serverLog(LL_WARNING, "Opening Unix socket: %s", server.neterr);
1875 1876
            exit(1);
        }
A
antirez 已提交
1877
        anetNonBlock(NULL,server.sofd);
1878
    }
1879 1880 1881

    /* Abort if there are no listening sockets at all. */
    if (server.ipfd_count == 0 && server.sofd < 0) {
A
antirez 已提交
1882
        serverLog(LL_WARNING, "Configured to not listen anywhere, exiting.");
1883 1884
        exit(1);
    }
1885 1886

    /* Create the Redis databases, and initialize other internal state. */
1887 1888 1889 1890
    for (j = 0; j < server.dbnum; j++) {
        server.db[j].dict = dictCreate(&dbDictType,NULL);
        server.db[j].expires = dictCreate(&keyptrDictType,NULL);
        server.db[j].blocking_keys = dictCreate(&keylistDictType,NULL);
1891
        server.db[j].ready_keys = dictCreate(&objectKeyPointerValueDictType,NULL);
1892 1893
        server.db[j].watched_keys = dictCreate(&keylistDictType,NULL);
        server.db[j].id = j;
1894
        server.db[j].avg_ttl = 0;
1895
    }
1896
    evictionPoolAlloc(); /* Initialize the LRU keys pool. */
1897 1898 1899 1900 1901
    server.pubsub_channels = dictCreate(&keylistDictType,NULL);
    server.pubsub_patterns = listCreate();
    listSetFreeMethod(server.pubsub_patterns,freePubsubPattern);
    listSetMatchMethod(server.pubsub_patterns,listMatchPubsubPattern);
    server.cronloops = 0;
A
antirez 已提交
1902
    server.rdb_child_pid = -1;
A
antirez 已提交
1903
    server.aof_child_pid = -1;
A
antirez 已提交
1904
    server.rdb_child_type = RDB_CHILD_TYPE_NONE;
1905
    server.rdb_bgsave_scheduled = 0;
1906 1907 1908
    server.child_info_pipe[0] = -1;
    server.child_info_pipe[1] = -1;
    server.child_info_data.magic = 0;
1909
    aofRewriteBufferReset();
A
antirez 已提交
1910
    server.aof_buf = sdsempty();
1911 1912
    server.lastsave = time(NULL); /* At startup we consider the DB saved. */
    server.lastbgsave_try = 0;    /* At startup we never tried to BGSAVE. */
1913 1914
    server.rdb_save_time_last = -1;
    server.rdb_save_time_start = -1;
1915
    server.dirty = 0;
1916 1917
    resetServerStats();
    /* A few stats we don't want to reset: server startup time, and peak mem. */
1918
    server.stat_starttime = time(NULL);
1919
    server.stat_peak_memory = 0;
1920 1921
    server.stat_rdb_cow_bytes = 0;
    server.stat_aof_cow_bytes = 0;
A
antirez 已提交
1922
    server.resident_set_size = 0;
1923 1924
    server.lastbgsave_status = C_OK;
    server.aof_last_write_status = C_OK;
1925
    server.aof_last_write_errno = 0;
1926
    server.repl_good_slaves_count = 0;
1927
    updateCachedTime();
1928

1929 1930 1931
    /* Create the timer callback, this is our way to process many background
     * operations incrementally, like clients timeout, eviction of unaccessed
     * expired keys and so forth. */
1932
    if (aeCreateTimeEvent(server.el, 1, serverCron, NULL, NULL) == AE_ERR) {
1933
        serverPanic("Can't create event loop timers.");
stamhe's avatar
stamhe 已提交
1934 1935
        exit(1);
    }
1936 1937 1938 1939 1940 1941 1942

    /* Create an event handler for accepting new connections in TCP and Unix
     * domain sockets. */
    for (j = 0; j < server.ipfd_count; j++) {
        if (aeCreateFileEvent(server.el, server.ipfd[j], AE_READABLE,
            acceptTcpHandler,NULL) == AE_ERR)
            {
A
antirez 已提交
1943
                serverPanic(
1944 1945 1946
                    "Unrecoverable error creating server.ipfd file event.");
            }
    }
1947
    if (server.sofd > 0 && aeCreateFileEvent(server.el,server.sofd,AE_READABLE,
A
antirez 已提交
1948
        acceptUnixHandler,NULL) == AE_ERR) serverPanic("Unrecoverable error creating server.sofd file event.");
1949

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959

    /* Register a readable event for the pipe used to awake the event loop
     * when a blocked client in a module needs attention. */
    if (aeCreateFileEvent(server.el, server.module_blocked_pipe[0], AE_READABLE,
        moduleBlockedClientPipeReadable,NULL) == AE_ERR) {
            serverPanic(
                "Error registering the readable event for the module "
                "blocked clients subsystem.");
    }

1960
    /* Open the AOF file if needed. */
A
antirez 已提交
1961
    if (server.aof_state == AOF_ON) {
A
antirez 已提交
1962
        server.aof_fd = open(server.aof_filename,
1963
                               O_WRONLY|O_APPEND|O_CREAT,0644);
A
antirez 已提交
1964
        if (server.aof_fd == -1) {
A
antirez 已提交
1965
            serverLog(LL_WARNING, "Can't open the append-only file: %s",
1966 1967 1968 1969 1970
                strerror(errno));
            exit(1);
        }
    }

1971 1972
    /* 32 bit instances are limited to 4GB of address space, so if there is
     * no explicit limit in the user provided configuration we set a limit
1973 1974
     * at 3 GB using maxmemory with 'noeviction' policy'. This avoids
     * useless crashes of the Redis instance for out of memory. */
1975
    if (server.arch_bits == 32 && server.maxmemory == 0) {
A
antirez 已提交
1976
        serverLog(LL_WARNING,"Warning: 32 bit instance detected but no memory limit set. Setting 3 GB maxmemory limit with 'noeviction' policy now.");
1977
        server.maxmemory = 3072LL*(1024*1024); /* 3 GB */
A
antirez 已提交
1978
        server.maxmemory_policy = MAXMEMORY_NO_EVICTION;
1979 1980
    }

A
antirez 已提交
1981
    if (server.cluster_enabled) clusterInit();
1982
    replicationScriptCacheInit();
1983
    scriptingInit(1);
1984
    slowlogInit();
1985
    latencyMonitorInit();
1986
    bioInit();
1987
    server.initial_memory_usage = zmalloc_used_memory();
1988 1989
}

1990 1991 1992 1993
/* Populates the Redis Command Table starting from the hard coded list
 * we have on top of redis.c file. */
void populateCommandTable(void) {
    int j;
1994
    int numcommands = sizeof(redisCommandTable)/sizeof(struct redisCommand);
1995 1996

    for (j = 0; j < numcommands; j++) {
1997
        struct redisCommand *c = redisCommandTable+j;
1998
        char *f = c->sflags;
1999
        int retval1, retval2;
2000

2001 2002
        while(*f != '\0') {
            switch(*f) {
A
antirez 已提交
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
            case 'w': c->flags |= CMD_WRITE; break;
            case 'r': c->flags |= CMD_READONLY; break;
            case 'm': c->flags |= CMD_DENYOOM; break;
            case 'a': c->flags |= CMD_ADMIN; break;
            case 'p': c->flags |= CMD_PUBSUB; break;
            case 's': c->flags |= CMD_NOSCRIPT; break;
            case 'R': c->flags |= CMD_RANDOM; break;
            case 'S': c->flags |= CMD_SORT_FOR_SCRIPT; break;
            case 'l': c->flags |= CMD_LOADING; break;
            case 't': c->flags |= CMD_STALE; break;
            case 'M': c->flags |= CMD_SKIP_MONITOR; break;
            case 'k': c->flags |= CMD_ASKING; break;
            case 'F': c->flags |= CMD_FAST; break;
            default: serverPanic("Unsupported command flag"); break;
2017 2018 2019 2020
            }
            f++;
        }

2021 2022 2023 2024
        retval1 = dictAdd(server.commands, sdsnew(c->name), c);
        /* Populate an additional dictionary that will be unaffected
         * by rename-command statements in redis.conf. */
        retval2 = dictAdd(server.orig_commands, sdsnew(c->name), c);
A
antirez 已提交
2025
        serverAssert(retval1 == DICT_OK && retval2 == DICT_OK);
2026
    }
2027 2028
}

2029
void resetCommandTableStats(void) {
2030 2031 2032
    struct redisCommand *c;
    dictEntry *de;
    dictIterator *di;
2033

2034 2035 2036
    di = dictGetSafeIterator(server.commands);
    while((de = dictNext(di)) != NULL) {
        c = (struct redisCommand *) dictGetVal(de);
2037 2038 2039
        c->microseconds = 0;
        c->calls = 0;
    }
2040 2041
    dictReleaseIterator(di);

2042 2043
}

2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080
/* ========================== Redis OP Array API ============================ */

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

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

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

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

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

2081 2082
/* ====================== Commands lookup and execution ===================== */

2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093
struct redisCommand *lookupCommand(sds name) {
    return dictFetchValue(server.commands, name);
}

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

    cmd = dictFetchValue(server.commands, name);
    sdsfree(name);
    return cmd;
2094 2095
}

2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109
/* Lookup the command in the current table, if not found also check in
 * the original table containing the original command names unaffected by
 * redis.conf rename-command statement.
 *
 * This is used by functions rewriting the argument vector such as
 * rewriteClientCommandVector() in order to set client->cmd pointer
 * correctly even if the command was renamed. */
struct redisCommand *lookupCommandOrOriginal(sds name) {
    struct redisCommand *cmd = dictFetchValue(server.commands, name);

    if (!cmd) cmd = dictFetchValue(server.orig_commands,name);
    return cmd;
}

A
antirez 已提交
2110
/* Propagate the specified command (in the context of the specified database id)
2111
 * to AOF and Slaves.
A
antirez 已提交
2112 2113
 *
 * flags are an xor between:
A
antirez 已提交
2114 2115 2116
 * + PROPAGATE_NONE (no propagation of command at all)
 * + PROPAGATE_AOF (propagate into the AOF file if is enabled)
 * + PROPAGATE_REPL (propagate into the replication link)
2117 2118 2119
 *
 * This should not be used inside commands implementation. Use instead
 * alsoPropagate(), preventCommandPropagation(), forceCommandPropagation().
A
antirez 已提交
2120 2121 2122 2123
 */
void propagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
               int flags)
{
A
antirez 已提交
2124
    if (server.aof_state != AOF_OFF && flags & PROPAGATE_AOF)
A
antirez 已提交
2125
        feedAppendOnlyFile(cmd,dbid,argv,argc);
A
antirez 已提交
2126
    if (flags & PROPAGATE_REPL)
A
antirez 已提交
2127 2128 2129
        replicationFeedSlaves(server.slaves,dbid,argv,argc);
}

2130
/* Used inside commands to schedule the propagation of additional commands
2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
 * after the current command is propagated to AOF / Replication.
 *
 * 'cmd' must be a pointer to the Redis command to replicate, dbid is the
 * database ID the command should be propagated into.
 * Arguments of the command to propagte are passed as an array of redis
 * objects pointers of len 'argc', using the 'argv' vector.
 *
 * The function does not take a reference to the passed 'argv' vector,
 * so it is up to the caller to release the passed argv (but it is usually
 * stack allocated).  The function autoamtically increments ref count of
 * passed objects, so the caller does not need to. */
2142 2143 2144
void alsoPropagate(struct redisCommand *cmd, int dbid, robj **argv, int argc,
                   int target)
{
2145
    robj **argvcopy;
2146 2147
    int j;

2148 2149 2150
    if (server.loading) return; /* No propagation during loading. */

    argvcopy = zmalloc(sizeof(robj*)*argc);
2151 2152 2153 2154 2155
    for (j = 0; j < argc; j++) {
        argvcopy[j] = argv[j];
        incrRefCount(argv[j]);
    }
    redisOpArrayAppend(&server.also_propagate,cmd,dbid,argvcopy,argc,target);
2156 2157
}

A
antirez 已提交
2158
/* It is possible to call the function forceCommandPropagation() inside a
J
Juarez Bochi 已提交
2159
 * Redis command implementation in order to to force the propagation of a
A
antirez 已提交
2160
 * specific command execution into AOF / Replication. */
2161
void forceCommandPropagation(client *c, int flags) {
A
antirez 已提交
2162 2163
    if (flags & PROPAGATE_REPL) c->flags |= CLIENT_FORCE_REPL;
    if (flags & PROPAGATE_AOF) c->flags |= CLIENT_FORCE_AOF;
A
antirez 已提交
2164 2165
}

2166 2167 2168
/* Avoid that the executed command is propagated at all. This way we
 * are free to just propagate what we want using the alsoPropagate()
 * API. */
2169
void preventCommandPropagation(client *c) {
A
antirez 已提交
2170
    c->flags |= CLIENT_PREVENT_PROP;
2171 2172
}

2173 2174 2175 2176 2177 2178 2179 2180 2181 2182
/* AOF specific version of preventCommandPropagation(). */
void preventCommandAOF(client *c) {
    c->flags |= CLIENT_PREVENT_AOF_PROP;
}

/* Replication specific version of preventCommandPropagation(). */
void preventCommandReplication(client *c) {
    c->flags |= CLIENT_PREVENT_REPL_PROP;
}

A
antirez 已提交
2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219
/* Call() is the core of Redis execution of a command.
 *
 * The following flags can be passed:
 * CMD_CALL_NONE        No flags.
 * CMD_CALL_SLOWLOG     Check command speed and log in the slow log if needed.
 * CMD_CALL_STATS       Populate command stats.
 * CMD_CALL_PROPAGATE_AOF   Append command to AOF if it modified the dataset
 *                          or if the client flags are forcing propagation.
 * CMD_CALL_PROPAGATE_REPL  Send command to salves if it modified the dataset
 *                          or if the client flags are forcing propagation.
 * CMD_CALL_PROPAGATE   Alias for PROPAGATE_AOF|PROPAGATE_REPL.
 * CMD_CALL_FULL        Alias for SLOWLOG|STATS|PROPAGATE.
 *
 * The exact propagation behavior depends on the client flags.
 * Specifically:
 *
 * 1. If the client flags CLIENT_FORCE_AOF or CLIENT_FORCE_REPL are set
 *    and assuming the corresponding CMD_CALL_PROPAGATE_AOF/REPL is set
 *    in the call flags, then the command is propagated even if the
 *    dataset was not affected by the command.
 * 2. If the client flags CLIENT_PREVENT_REPL_PROP or CLIENT_PREVENT_AOF_PROP
 *    are set, the propagation into AOF or to slaves is not performed even
 *    if the command modified the dataset.
 *
 * Note that regardless of the client flags, if CMD_CALL_PROPAGATE_AOF
 * or CMD_CALL_PROPAGATE_REPL are not set, then respectively AOF or
 * slaves propagation will never occur.
 *
 * Client flags are modified by the implementation of a given command
 * using the following API:
 *
 * forceCommandPropagation(client *c, int flags);
 * preventCommandPropagation(client *c);
 * preventCommandAOF(client *c);
 * preventCommandReplication(client *c);
 *
 */
2220
void call(client *c, int flags) {
2221
    long long dirty, start, duration;
A
antirez 已提交
2222
    int client_old_flags = c->flags;
2223

A
antirez 已提交
2224
    /* Sent the command to clients in MONITOR mode, only if the commands are
G
guiquanz 已提交
2225
     * not generated from reading an AOF. */
2226 2227
    if (listLength(server.monitors) &&
        !server.loading &&
A
antirez 已提交
2228
        !(c->cmd->flags & (CMD_SKIP_MONITOR|CMD_ADMIN)))
2229
    {
2230
        replicationFeedMonitors(c,server.monitors,c->db->id,c->argv,c->argc);
2231
    }
A
antirez 已提交
2232

2233 2234 2235
    /* Initialization: clear the flags that must be set by the command on
     * demand, and initialize the array for additional commands propagation. */
    c->flags &= ~(CLIENT_FORCE_AOF|CLIENT_FORCE_REPL|CLIENT_PREVENT_PROP);
A
antirez 已提交
2236
    redisOpArray prev_also_propagate = server.also_propagate;
2237
    redisOpArrayInit(&server.also_propagate);
2238 2239

    /* Call the command. */
2240
    dirty = server.dirty;
2241
    start = ustime();
2242
    c->cmd->proc(c);
2243
    duration = ustime()-start;
2244
    dirty = server.dirty-dirty;
A
antirez 已提交
2245
    if (dirty < 0) dirty = 0;
2246 2247 2248

    /* When EVAL is called loading the AOF we don't want commands called
     * from Lua to go into the slowlog or to populate statistics. */
A
antirez 已提交
2249 2250
    if (server.loading && c->flags & CLIENT_LUA)
        flags &= ~(CMD_CALL_SLOWLOG | CMD_CALL_STATS);
2251

A
antirez 已提交
2252 2253 2254
    /* If the caller is Lua, we want to force the EVAL caller to propagate
     * the script if the command flag or client flag are forcing the
     * propagation. */
A
antirez 已提交
2255 2256 2257 2258 2259
    if (c->flags & CLIENT_LUA && server.lua_caller) {
        if (c->flags & CLIENT_FORCE_REPL)
            server.lua_caller->flags |= CLIENT_FORCE_REPL;
        if (c->flags & CLIENT_FORCE_AOF)
            server.lua_caller->flags |= CLIENT_FORCE_AOF;
A
antirez 已提交
2260 2261
    }

A
antirez 已提交
2262 2263
    /* Log the command into the Slow log if needed, and populate the
     * per-command statistics that we show in INFO commandstats. */
A
antirez 已提交
2264 2265
    if (flags & CMD_CALL_SLOWLOG && c->cmd->proc != execCommand) {
        char *latency_event = (c->cmd->flags & CMD_FAST) ?
2266
                              "fast-command" : "command";
2267
        latencyAddSampleIfNeeded(latency_event,duration/1000);
2268
        slowlogPushEntryIfNeeded(c,c->argv,c->argc,duration);
2269
    }
A
antirez 已提交
2270
    if (flags & CMD_CALL_STATS) {
2271 2272
        c->lastcmd->microseconds += duration;
        c->lastcmd->calls++;
2273
    }
A
antirez 已提交
2274 2275

    /* Propagate the command into the AOF and replication link */
A
antirez 已提交
2276 2277 2278 2279
    if (flags & CMD_CALL_PROPAGATE &&
        (c->flags & CLIENT_PREVENT_PROP) != CLIENT_PREVENT_PROP)
    {
        int propagate_flags = PROPAGATE_NONE;
A
antirez 已提交
2280

2281 2282
        /* Check if the command operated changes in the data set. If so
         * set for replication / AOF propagation. */
A
antirez 已提交
2283
        if (dirty) propagate_flags |= (PROPAGATE_AOF|PROPAGATE_REPL);
2284

A
antirez 已提交
2285
        /* If the client forced AOF / replication of the command, set
2286
         * the flags regardless of the command effects on the data set. */
A
antirez 已提交
2287 2288
        if (c->flags & CLIENT_FORCE_REPL) propagate_flags |= PROPAGATE_REPL;
        if (c->flags & CLIENT_FORCE_AOF) propagate_flags |= PROPAGATE_AOF;
2289

A
antirez 已提交
2290 2291 2292 2293 2294 2295 2296 2297 2298
        /* However prevent AOF / replication propagation if the command
         * implementatino called preventCommandPropagation() or similar,
         * or if we don't have the call() flags to do so. */
        if (c->flags & CLIENT_PREVENT_REPL_PROP ||
            !(flags & CMD_CALL_PROPAGATE_REPL))
                propagate_flags &= ~PROPAGATE_REPL;
        if (c->flags & CLIENT_PREVENT_AOF_PROP ||
            !(flags & CMD_CALL_PROPAGATE_AOF))
                propagate_flags &= ~PROPAGATE_AOF;
2299 2300

        /* Call propagate() only if at least one of AOF / replication
2301 2302 2303
         * propagation is needed. Note that modules commands handle replication
         * in an explicit way, so we never replicate them automatically. */
        if (propagate_flags != PROPAGATE_NONE && !(c->cmd->flags & CMD_MODULE))
A
antirez 已提交
2304
            propagate(c->cmd,c->db->id,c->argv,c->argc,propagate_flags);
2305
    }
2306

2307
    /* Restore the old replication flags, since call() can be executed
A
antirez 已提交
2308
     * recursively. */
A
antirez 已提交
2309
    c->flags &= ~(CLIENT_FORCE_AOF|CLIENT_FORCE_REPL|CLIENT_PREVENT_PROP);
2310
    c->flags |= client_old_flags &
A
antirez 已提交
2311
        (CLIENT_FORCE_AOF|CLIENT_FORCE_REPL|CLIENT_PREVENT_PROP);
A
antirez 已提交
2312

2313
    /* Handle the alsoPropagate() API to handle commands that want to propagate
2314
     * multiple separated commands. Note that alsoPropagate() is not affected
A
antirez 已提交
2315
     * by CLIENT_PREVENT_PROP flag. */
2316
    if (server.also_propagate.numops) {
2317
        int j;
2318
        redisOp *rop;
2319

A
antirez 已提交
2320
        if (flags & CMD_CALL_PROPAGATE) {
2321 2322
            for (j = 0; j < server.also_propagate.numops; j++) {
                rop = &server.also_propagate.ops[j];
A
antirez 已提交
2323 2324 2325 2326 2327 2328
                int target = rop->target;
                /* Whatever the command wish is, we honor the call() flags. */
                if (!(flags&CMD_CALL_PROPAGATE_AOF)) target &= ~PROPAGATE_AOF;
                if (!(flags&CMD_CALL_PROPAGATE_REPL)) target &= ~PROPAGATE_REPL;
                if (target)
                    propagate(rop->cmd,rop->dbid,rop->argv,rop->argc,target);
2329
            }
2330 2331
        }
        redisOpArrayFree(&server.also_propagate);
2332
    }
A
antirez 已提交
2333
    server.also_propagate = prev_also_propagate;
2334 2335 2336 2337
    server.stat_numcommands++;
}

/* If this function gets called we already read a whole
2338
 * command, arguments are in the client argv/argc fields.
2339 2340 2341
 * processCommand() execute the command or prepare the
 * server for a bulk read from the client.
 *
2342
 * If C_OK is returned the client is still alive and valid and
G
guiquanz 已提交
2343
 * other operations can be performed by the caller. Otherwise
2344
 * if C_ERR is returned the client was destroyed (i.e. after QUIT). */
2345
int processCommand(client *c) {
P
Pieter Noordhuis 已提交
2346 2347 2348 2349
    /* The QUIT command is handled separately. Normal command procs will
     * go through checking for replication and QUIT will cause trouble
     * when FORCE_REPLICATION is enabled and would be implemented in
     * a regular command proc. */
2350
    if (!strcasecmp(c->argv[0]->ptr,"quit")) {
P
Pieter Noordhuis 已提交
2351
        addReply(c,shared.ok);
A
antirez 已提交
2352
        c->flags |= CLIENT_CLOSE_AFTER_REPLY;
2353
        return C_ERR;
2354 2355 2356
    }

    /* Now lookup the command and check ASAP about trivial error conditions
2357
     * such as wrong arity, bad command name and so forth. */
2358
    c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
2359
    if (!c->cmd) {
A
antirez 已提交
2360
        flagTransaction(c);
2361 2362
        addReplyErrorFormat(c,"unknown command '%s'",
            (char*)c->argv[0]->ptr);
2363
        return C_OK;
2364 2365
    } else if ((c->cmd->arity > 0 && c->cmd->arity != c->argc) ||
               (c->argc < -c->cmd->arity)) {
A
antirez 已提交
2366
        flagTransaction(c);
2367
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
2368
            c->cmd->name);
2369
        return C_OK;
2370 2371 2372
    }

    /* Check if the user is authenticated */
2373 2374
    if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand)
    {
A
antirez 已提交
2375
        flagTransaction(c);
2376
        addReply(c,shared.noautherr);
2377
        return C_OK;
2378 2379
    }

2380 2381 2382 2383
    /* If cluster is enabled perform the cluster redirection here.
     * However we don't perform the redirection if:
     * 1) The sender of this command is our master.
     * 2) The command has no key arguments. */
A
antirez 已提交
2384
    if (server.cluster_enabled &&
A
antirez 已提交
2385 2386 2387
        !(c->flags & CLIENT_MASTER) &&
        !(c->flags & CLIENT_LUA &&
          server.lua_caller->flags & CLIENT_MASTER) &&
2388 2389
        !(c->cmd->getkeys_proc == NULL && c->cmd->firstkey == 0 &&
          c->cmd->proc != execCommand))
2390
    {
A
antirez 已提交
2391
        int hashslot;
2392 2393 2394 2395 2396 2397 2398 2399
        int error_code;
        clusterNode *n = getNodeByQuery(c,c->cmd,c->argv,c->argc,
                                        &hashslot,&error_code);
        if (n == NULL || n != server.cluster->myself) {
            if (c->cmd->proc == execCommand) {
                discardTransaction(c);
            } else {
                flagTransaction(c);
A
antirez 已提交
2400
            }
2401 2402
            clusterRedirectClient(c,n,hashslot,error_code);
            return C_OK;
A
antirez 已提交
2403 2404 2405
        }
    }

2406 2407 2408 2409 2410
    /* Handle the maxmemory directive.
     *
     * First we try to free some memory if possible (if there are volatile
     * keys in the dataset). If there are not the only thing we can do
     * is returning an error. */
2411 2412
    if (server.maxmemory) {
        int retval = freeMemoryIfNeeded();
2413 2414 2415 2416 2417 2418
        /* freeMemoryIfNeeded may flush slave output buffers. This may result
         * into a slave, that may be the active client, to be freed. */
        if (server.current_client == NULL) return C_ERR;

        /* It was impossible to free enough memory, and the command the client
         * is trying to execute is denied during OOM conditions? Error. */
A
antirez 已提交
2419
        if ((c->cmd->flags & CMD_DENYOOM) && retval == C_ERR) {
A
antirez 已提交
2420
            flagTransaction(c);
2421
            addReply(c, shared.oomerr);
2422
            return C_OK;
2423
        }
2424 2425
    }

2426 2427
    /* Don't accept write commands if there are problems persisting on disk
     * and if this is a master instance. */
2428 2429
    if (((server.stop_writes_on_bgsave_err &&
          server.saveparamslen > 0 &&
2430 2431
          server.lastbgsave_status == C_ERR) ||
          server.aof_last_write_status == C_ERR) &&
2432
        server.masterhost == NULL &&
A
antirez 已提交
2433
        (c->cmd->flags & CMD_WRITE ||
2434
         c->cmd->proc == pingCommand))
2435
    {
A
antirez 已提交
2436
        flagTransaction(c);
2437
        if (server.aof_last_write_status == C_OK)
2438 2439 2440 2441 2442 2443
            addReply(c, shared.bgsaveerr);
        else
            addReplySds(c,
                sdscatprintf(sdsempty(),
                "-MISCONF Errors writing to the AOF file: %s\r\n",
                strerror(server.aof_last_write_errno)));
2444
        return C_OK;
2445 2446
    }

2447
    /* Don't accept write commands if there are not enough good slaves and
2448
     * user configured the min-slaves-to-write option. */
2449 2450
    if (server.masterhost == NULL &&
        server.repl_min_slaves_to_write &&
2451
        server.repl_min_slaves_max_lag &&
A
antirez 已提交
2452
        c->cmd->flags & CMD_WRITE &&
2453 2454 2455 2456
        server.repl_good_slaves_count < server.repl_min_slaves_to_write)
    {
        flagTransaction(c);
        addReply(c, shared.noreplicaserr);
2457
        return C_OK;
2458 2459
    }

2460
    /* Don't accept write commands if this is a read only slave. But
2461 2462
     * accept write commands if this is our master. */
    if (server.masterhost && server.repl_slave_ro &&
A
antirez 已提交
2463 2464
        !(c->flags & CLIENT_MASTER) &&
        c->cmd->flags & CMD_WRITE)
2465 2466
    {
        addReply(c, shared.roslaveerr);
2467
        return C_OK;
2468 2469
    }

2470
    /* Only allow SUBSCRIBE and UNSUBSCRIBE in the context of Pub/Sub */
A
antirez 已提交
2471
    if (c->flags & CLIENT_PUBSUB &&
2472
        c->cmd->proc != pingCommand &&
2473 2474 2475 2476
        c->cmd->proc != subscribeCommand &&
        c->cmd->proc != unsubscribeCommand &&
        c->cmd->proc != psubscribeCommand &&
        c->cmd->proc != punsubscribeCommand) {
2477
        addReplyError(c,"only (P)SUBSCRIBE / (P)UNSUBSCRIBE / PING / QUIT allowed in this context");
2478
        return C_OK;
2479 2480
    }

W
WuYunlong 已提交
2481 2482 2483
    /* Only allow commands with flag "t", such as INFO, SLAVEOF and so on,
     * when slave-serve-stale-data is no and we are a slave with a broken
     * link with master. */
A
antirez 已提交
2484
    if (server.masterhost && server.repl_state != REPL_STATE_CONNECTED &&
2485
        server.repl_serve_stale_data == 0 &&
A
antirez 已提交
2486
        !(c->cmd->flags & CMD_STALE))
2487
    {
A
antirez 已提交
2488
        flagTransaction(c);
2489
        addReply(c, shared.masterdownerr);
2490
        return C_OK;
2491 2492
    }

2493
    /* Loading DB? Return an error if the command has not the
A
antirez 已提交
2494 2495
     * CMD_LOADING flag. */
    if (server.loading && !(c->cmd->flags & CMD_LOADING)) {
2496
        addReply(c, shared.loadingerr);
2497
        return C_OK;
2498 2499
    }

A
antirez 已提交
2500
    /* Lua script too slow? Only allow a limited number of commands. */
2501
    if (server.lua_timedout &&
2502
          c->cmd->proc != authCommand &&
A
antirez 已提交
2503
          c->cmd->proc != replconfCommand &&
2504
        !(c->cmd->proc == shutdownCommand &&
2505 2506 2507 2508 2509 2510
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'n') &&
        !(c->cmd->proc == scriptCommand &&
          c->argc == 2 &&
          tolower(((char*)c->argv[1]->ptr)[0]) == 'k'))
    {
A
antirez 已提交
2511
        flagTransaction(c);
2512
        addReply(c, shared.slowscripterr);
2513
        return C_OK;
2514 2515
    }

2516
    /* Exec the command */
A
antirez 已提交
2517
    if (c->flags & CLIENT_MULTI &&
2518 2519
        c->cmd->proc != execCommand && c->cmd->proc != discardCommand &&
        c->cmd->proc != multiCommand && c->cmd->proc != watchCommand)
2520
    {
2521
        queueMultiCommand(c);
2522 2523
        addReply(c,shared.queued);
    } else {
A
antirez 已提交
2524
        call(c,CMD_CALL_FULL);
2525
        c->woff = server.master_repl_offset;
2526
        if (listLength(server.ready_keys))
2527
            handleClientsBlockedOnKeys();
2528
    }
2529
    return C_OK;
2530 2531 2532 2533
}

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

2534 2535 2536 2537 2538 2539 2540 2541 2542 2543
/* Close listening sockets. Also unlink the unix domain socket if
 * unlink_unix_socket is non-zero. */
void closeListeningSockets(int unlink_unix_socket) {
    int j;

    for (j = 0; j < server.ipfd_count; j++) close(server.ipfd[j]);
    if (server.sofd != -1) close(server.sofd);
    if (server.cluster_enabled)
        for (j = 0; j < server.cfd_count; j++) close(server.cfd[j]);
    if (unlink_unix_socket && server.unixsocket) {
A
antirez 已提交
2544
        serverLog(LL_NOTICE,"Removing the unix socket file.");
2545 2546 2547 2548
        unlink(server.unixsocket); /* don't care if this fails */
    }
}

2549
int prepareForShutdown(int flags) {
A
antirez 已提交
2550 2551
    int save = flags & SHUTDOWN_SAVE;
    int nosave = flags & SHUTDOWN_NOSAVE;
2552

A
antirez 已提交
2553
    serverLog(LL_WARNING,"User requested shutdown...");
2554

2555 2556 2557
    /* Kill all the Lua debugger forked sessions. */
    ldbKillForkedSessions();

2558 2559 2560
    /* Kill the saving child if there is a background saving in progress.
       We want to avoid race conditions, for instance our saving child may
       overwrite the synchronous saving did by SHUTDOWN. */
A
antirez 已提交
2561
    if (server.rdb_child_pid != -1) {
A
antirez 已提交
2562
        serverLog(LL_WARNING,"There is a child saving an .rdb. Killing it!");
2563
        kill(server.rdb_child_pid,SIGUSR1);
A
antirez 已提交
2564
        rdbRemoveTempFile(server.rdb_child_pid);
2565
    }
2566

A
antirez 已提交
2567
    if (server.aof_state != AOF_OFF) {
2568 2569
        /* Kill the AOF saving child as the AOF we already have may be longer
         * but contains the full dataset anyway. */
A
antirez 已提交
2570
        if (server.aof_child_pid != -1) {
2571 2572
            /* If we have AOF enabled but haven't written the AOF yet, don't
             * shutdown or else the dataset will be lost. */
A
antirez 已提交
2573 2574
            if (server.aof_state == AOF_WAIT_REWRITE) {
                serverLog(LL_WARNING, "Writing initial AOF, can't exit.");
2575
                return C_ERR;
2576
            }
A
antirez 已提交
2577
            serverLog(LL_WARNING,
2578
                "There is a child rewriting the AOF. Killing it!");
2579
            kill(server.aof_child_pid,SIGUSR1);
2580
        }
2581
        /* Append only file: flush buffers and fsync() the AOF at exit */
A
antirez 已提交
2582
        serverLog(LL_NOTICE,"Calling fsync() on the AOF file.");
2583
        flushAppendOnlyFile(1);
A
antirez 已提交
2584
        aof_fsync(server.aof_fd);
2585
    }
2586 2587

    /* Create a new RDB file before exiting. */
2588
    if ((server.saveparamslen > 0 && !nosave) || save) {
A
antirez 已提交
2589
        serverLog(LL_NOTICE,"Saving the final RDB snapshot before exiting.");
2590
        /* Snapshotting. Perform a SYNC SAVE and exit */
2591 2592 2593
        rdbSaveInfo rsi, *rsiptr;
        rsiptr = rdbPopulateSaveInfo(&rsi);
        if (rdbSave(server.rdb_filename,rsiptr) != C_OK) {
2594 2595 2596 2597 2598
            /* Ooops.. error saving! The best we can do is to continue
             * operating. Note that if there was a background saving process,
             * in the next cron() Redis will be notified that the background
             * saving aborted, handling special stuff like slaves pending for
             * synchronization... */
A
antirez 已提交
2599
            serverLog(LL_WARNING,"Error trying to save the DB, can't exit.");
2600
            return C_ERR;
2601 2602
        }
    }
2603 2604

    /* Remove the pid file if possible and needed. */
R
rebx 已提交
2605
    if (server.daemonize || server.pidfile) {
A
antirez 已提交
2606
        serverLog(LL_NOTICE,"Removing the pid file.");
2607 2608
        unlink(server.pidfile);
    }
2609 2610 2611 2612 2613

    /* Best effort flush of slave output buffers, so that we hopefully
     * send them pending writes. */
    flushSlavesOutputBuffers();

2614
    /* Close the listening sockets. Apparently this allows faster restarts. */
2615
    closeListeningSockets(1);
A
antirez 已提交
2616
    serverLog(LL_WARNING,"%s is now ready to exit, bye bye...",
2617
        server.sentinel_mode ? "Sentinel" : "Redis");
2618
    return C_OK;
2619 2620 2621 2622
}

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

2623
/* Return zero if strings are the same, non-zero if they are not.
2624 2625 2626 2627 2628 2629 2630 2631 2632
 * The comparison is performed in a way that prevents an attacker to obtain
 * information about the nature of the strings just monitoring the execution
 * time of the function.
 *
 * Note that limiting the comparison length to strings up to 512 bytes we
 * can avoid leaking any information about the password length and any
 * possible branch misprediction related leak.
 */
int time_independent_strcmp(char *a, char *b) {
A
antirez 已提交
2633
    char bufa[CONFIG_AUTHPASS_MAX_LEN], bufb[CONFIG_AUTHPASS_MAX_LEN];
2634 2635 2636 2637
    /* The above two strlen perform len(a) + len(b) operations where either
     * a or b are fixed (our password) length, and the difference is only
     * relative to the length of the user provided string, so no information
     * leak is possible in the following two lines of code. */
2638 2639 2640
    unsigned int alen = strlen(a);
    unsigned int blen = strlen(b);
    unsigned int j;
2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664
    int diff = 0;

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

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

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

2665
void authCommand(client *c) {
2666 2667
    if (!server.requirepass) {
        addReplyError(c,"Client sent AUTH, but no password is set");
2668
    } else if (!time_independent_strcmp(c->argv[1]->ptr, server.requirepass)) {
2669 2670 2671 2672
      c->authenticated = 1;
      addReply(c,shared.ok);
    } else {
      c->authenticated = 0;
2673
      addReplyError(c,"invalid password");
2674 2675 2676
    }
}

2677 2678
/* The PING command. It works in a different way if the client is in
 * in Pub/Sub mode. */
2679
void pingCommand(client *c) {
2680 2681
    /* The command takes zero or one arguments. */
    if (c->argc > 2) {
2682 2683
        addReplyErrorFormat(c,"wrong number of arguments for '%s' command",
            c->cmd->name);
2684 2685 2686
        return;
    }

A
antirez 已提交
2687
    if (c->flags & CLIENT_PUBSUB) {
2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699
        addReply(c,shared.mbulkhdr[2]);
        addReplyBulkCBuffer(c,"pong",4);
        if (c->argc == 1)
            addReplyBulkCBuffer(c,"",0);
        else
            addReplyBulk(c,c->argv[1]);
    } else {
        if (c->argc == 1)
            addReply(c,shared.pong);
        else
            addReplyBulk(c,c->argv[1]);
    }
2700 2701
}

2702
void echoCommand(client *c) {
2703 2704 2705
    addReplyBulk(c,c->argv[1]);
}

2706
void timeCommand(client *c) {
A
antirez 已提交
2707 2708
    struct timeval tv;

G
guiquanz 已提交
2709
    /* gettimeofday() can only fail if &tv is a bad address so we
A
antirez 已提交
2710 2711 2712 2713 2714 2715 2716
     * don't check for errors. */
    gettimeofday(&tv,NULL);
    addReplyMultiBulkLen(c,2);
    addReplyBulkLongLong(c,tv.tv_sec);
    addReplyBulkLongLong(c,tv.tv_usec);
}

A
antirez 已提交
2717
/* Helper function for addReplyCommand() to output flags. */
2718
int addReplyCommandFlag(client *c, struct redisCommand *cmd, int f, char *reply) {
M
Matt Stancliff 已提交
2719 2720 2721 2722 2723 2724
    if (cmd->flags & f) {
        addReplyStatus(c, reply);
        return 1;
    }
    return 0;
}
2725

A
antirez 已提交
2726
/* Output the representation of a Redis command. Used by the COMMAND command. */
2727
void addReplyCommand(client *c, struct redisCommand *cmd) {
M
Matt Stancliff 已提交
2728 2729 2730 2731 2732 2733 2734 2735 2736 2737
    if (!cmd) {
        addReply(c, shared.nullbulk);
    } else {
        /* We are adding: command name, arg count, flags, first, last, offset */
        addReplyMultiBulkLen(c, 6);
        addReplyBulkCString(c, cmd->name);
        addReplyLongLong(c, cmd->arity);

        int flagcount = 0;
        void *flaglen = addDeferredMultiBulkLength(c);
A
antirez 已提交
2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
        flagcount += addReplyCommandFlag(c,cmd,CMD_WRITE, "write");
        flagcount += addReplyCommandFlag(c,cmd,CMD_READONLY, "readonly");
        flagcount += addReplyCommandFlag(c,cmd,CMD_DENYOOM, "denyoom");
        flagcount += addReplyCommandFlag(c,cmd,CMD_ADMIN, "admin");
        flagcount += addReplyCommandFlag(c,cmd,CMD_PUBSUB, "pubsub");
        flagcount += addReplyCommandFlag(c,cmd,CMD_NOSCRIPT, "noscript");
        flagcount += addReplyCommandFlag(c,cmd,CMD_RANDOM, "random");
        flagcount += addReplyCommandFlag(c,cmd,CMD_SORT_FOR_SCRIPT,"sort_for_script");
        flagcount += addReplyCommandFlag(c,cmd,CMD_LOADING, "loading");
        flagcount += addReplyCommandFlag(c,cmd,CMD_STALE, "stale");
        flagcount += addReplyCommandFlag(c,cmd,CMD_SKIP_MONITOR, "skip_monitor");
        flagcount += addReplyCommandFlag(c,cmd,CMD_ASKING, "asking");
        flagcount += addReplyCommandFlag(c,cmd,CMD_FAST, "fast");
2751 2752 2753
        if ((cmd->getkeys_proc && !(cmd->flags & CMD_MODULE)) ||
            cmd->flags & CMD_MODULE_GETKEYS)
        {
M
Matt Stancliff 已提交
2754 2755 2756 2757 2758 2759 2760 2761 2762 2763
            addReplyStatus(c, "movablekeys");
            flagcount += 1;
        }
        setDeferredMultiBulkLength(c, flaglen, flagcount);

        addReplyLongLong(c, cmd->firstkey);
        addReplyLongLong(c, cmd->lastkey);
        addReplyLongLong(c, cmd->keystep);
    }
}
2764

A
antirez 已提交
2765
/* COMMAND <subcommand> <args> */
2766
void commandCommand(client *c) {
M
Matt Stancliff 已提交
2767 2768 2769
    dictIterator *di;
    dictEntry *de;

I
Itamar Haber 已提交
2770 2771
    if (c->argc == 2 && !strcasecmp(c->argv[1]->ptr,"help")) {
        const char *help[] = {
2772 2773 2774 2775 2776
"(no subcommand) -- Return details about all Redis commands.",
"count -- Return the total number of commands in this Redis server.",
"getkeys <full-command> -- Return the keys from a full Redis command.",
"info [command-name ...] -- Return details about multiple Redis commands.",
NULL
I
Itamar Haber 已提交
2777 2778 2779
        };
        addReplyHelp(c, help);
    } else if (c->argc == 1) {
A
antirez 已提交
2780 2781 2782 2783 2784 2785 2786
        addReplyMultiBulkLen(c, dictSize(server.commands));
        di = dictGetIterator(server.commands);
        while ((de = dictNext(di)) != NULL) {
            addReplyCommand(c, dictGetVal(de));
        }
        dictReleaseIterator(di);
    } else if (!strcasecmp(c->argv[1]->ptr, "info")) {
M
Matt Stancliff 已提交
2787 2788 2789
        int i;
        addReplyMultiBulkLen(c, c->argc-2);
        for (i = 2; i < c->argc; i++) {
2790
            addReplyCommand(c, dictFetchValue(server.commands, c->argv[i]->ptr));
M
Matt Stancliff 已提交
2791
        }
A
antirez 已提交
2792 2793
    } else if (!strcasecmp(c->argv[1]->ptr, "count") && c->argc == 2) {
        addReplyLongLong(c, dictSize(server.commands));
2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811
    } else if (!strcasecmp(c->argv[1]->ptr,"getkeys") && c->argc >= 3) {
        struct redisCommand *cmd = lookupCommand(c->argv[2]->ptr);
        int *keys, numkeys, j;

        if (!cmd) {
            addReplyErrorFormat(c,"Invalid command specified");
            return;
        } else if ((cmd->arity > 0 && cmd->arity != c->argc-2) ||
                   ((c->argc-2) < -cmd->arity))
        {
            addReplyError(c,"Invalid number of arguments specified for command");
            return;
        }

        keys = getKeysFromCommand(cmd,c->argv+2,c->argc-2,&numkeys);
        addReplyMultiBulkLen(c,numkeys);
        for (j = 0; j < numkeys; j++) addReplyBulk(c,c->argv[keys[j]+2]);
        getKeysFreeResult(keys);
A
antirez 已提交
2812
    } else {
2813
        addReplyErrorFormat(c, "Unknown subcommand or wrong number of arguments for '%s'. Try COMMAND HELP", (char*)c->argv[1]->ptr);
M
Matt Stancliff 已提交
2814 2815 2816
    }
}

2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834
/* Convert an amount of bytes into a human readable string in the form
 * of 100B, 2G, 100M, 4K, and so forth. */
void bytesToHuman(char *s, unsigned long long n) {
    double d;

    if (n < 1024) {
        /* Bytes */
        sprintf(s,"%lluB",n);
        return;
    } else if (n < (1024*1024)) {
        d = (double)n/(1024);
        sprintf(s,"%.2fK",d);
    } else if (n < (1024LL*1024*1024)) {
        d = (double)n/(1024*1024);
        sprintf(s,"%.2fM",d);
    } else if (n < (1024LL*1024*1024*1024)) {
        d = (double)n/(1024LL*1024*1024);
        sprintf(s,"%.2fG",d);
2835 2836 2837 2838 2839 2840 2841 2842 2843
    } else if (n < (1024LL*1024*1024*1024*1024)) {
        d = (double)n/(1024LL*1024*1024*1024);
        sprintf(s,"%.2fT",d);
    } else if (n < (1024LL*1024*1024*1024*1024*1024)) {
        d = (double)n/(1024LL*1024*1024*1024*1024);
        sprintf(s,"%.2fP",d);
    } else {
        /* Let's hope we never need this */
        sprintf(s,"%lluB",n);
2844 2845 2846 2847 2848 2849
    }
}

/* Create the string returned by the INFO command. This is decoupled
 * by the INFO command itself as we need to report the same information
 * on memory corruption problems. */
2850 2851
sds genRedisInfoString(char *section) {
    sds info = sdsempty();
2852
    time_t uptime = server.unixtime-server.stat_starttime;
2853
    int j;
2854
    struct rusage self_ru, c_ru;
2855
    unsigned long lol, bib;
2856 2857
    int allsections = 0, defsections = 0;
    int sections = 0;
2858

2859 2860 2861
    if (section == NULL) section = "default";
    allsections = strcasecmp(section,"all") == 0;
    defsections = strcasecmp(section,"default") == 0;
2862 2863 2864

    getrusage(RUSAGE_SELF, &self_ru);
    getrusage(RUSAGE_CHILDREN, &c_ru);
2865
    getClientsMaxBuffers(&lol,&bib);
2866 2867 2868

    /* Server */
    if (allsections || defsections || !strcasecmp(section,"server")) {
2869 2870
        static int call_uname = 1;
        static struct utsname name;
2871
        char *mode;
2872

2873 2874 2875
        if (server.cluster_enabled) mode = "cluster";
        else if (server.sentinel_mode) mode = "sentinel";
        else mode = "standalone";
2876

2877
        if (sections++) info = sdscat(info,"\r\n");
2878 2879 2880 2881 2882 2883 2884

        if (call_uname) {
            /* Uname can be slow and is always the same output. Cache it. */
            uname(&name);
            call_uname = 0;
        }

2885 2886
        unsigned int lruclock;
        atomicGet(server.lruclock,lruclock);
2887
        info = sdscatprintf(info,
2888 2889 2890 2891
            "# Server\r\n"
            "redis_version:%s\r\n"
            "redis_git_sha1:%s\r\n"
            "redis_git_dirty:%d\r\n"
2892
            "redis_build_id:%llx\r\n"
2893
            "redis_mode:%s\r\n"
2894
            "os:%s %s %s\r\n"
2895
            "arch_bits:%d\r\n"
2896
            "multiplexing_api:%s\r\n"
2897
            "atomicvar_api:%s\r\n"
A
antirez 已提交
2898
            "gcc_version:%d.%d.%d\r\n"
2899
            "process_id:%ld\r\n"
A
antirez 已提交
2900
            "run_id:%s\r\n"
2901
            "tcp_port:%d\r\n"
Y
YAMAMOTO Takashi 已提交
2902 2903
            "uptime_in_seconds:%jd\r\n"
            "uptime_in_days:%jd\r\n"
2904
            "hz:%d\r\n"
2905
            "lru_clock:%ld\r\n"
A
antirez 已提交
2906
            "executable:%s\r\n"
2907
            "config_file:%s\r\n",
2908 2909 2910
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
2911
            (unsigned long long) redisBuildId(),
2912
            mode,
2913
            name.sysname, name.release, name.machine,
2914
            server.arch_bits,
2915
            aeGetApiName(),
2916
            REDIS_ATOMIC_API,
A
antirez 已提交
2917 2918 2919 2920 2921
#ifdef __GNUC__
            __GNUC__,__GNUC_MINOR__,__GNUC_PATCHLEVEL__,
#else
            0,0,0,
#endif
2922
            (long) getpid(),
A
antirez 已提交
2923
            server.runid,
2924
            server.port,
Y
YAMAMOTO Takashi 已提交
2925 2926
            (intmax_t)uptime,
            (intmax_t)(uptime/(3600*24)),
2927
            server.hz,
2928
            (unsigned long) lruclock,
A
antirez 已提交
2929
            server.executable ? server.executable : "",
2930
            server.configfile ? server.configfile : "");
2931 2932 2933 2934 2935 2936 2937
    }

    /* Clients */
    if (allsections || defsections || !strcasecmp(section,"clients")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Clients\r\n"
2938
            "connected_clients:%lu\r\n"
2939 2940 2941 2942 2943
            "client_longest_output_list:%lu\r\n"
            "client_biggest_input_buf:%lu\r\n"
            "blocked_clients:%d\r\n",
            listLength(server.clients)-listLength(server.slaves),
            lol, bib,
A
antirez 已提交
2944
            server.blocked_clients);
2945 2946 2947 2948
    }

    /* Memory */
    if (allsections || defsections || !strcasecmp(section,"memory")) {
2949 2950
        char hmem[64];
        char peak_hmem[64];
2951
        char total_system_hmem[64];
2952 2953
        char used_memory_lua_hmem[64];
        char used_memory_rss_hmem[64];
2954
        char maxmemory_hmem[64];
2955
        size_t zmalloc_used = zmalloc_used_memory();
2956
        size_t total_system_mem = server.system_memory_size;
T
therealbill 已提交
2957
        const char *evict_policy = evictPolicyToString();
2958
        long long memory_lua = (long long)lua_gc(server.lua,LUA_GCCOUNT,0)*1024;
2959
        struct redisMemOverhead *mh = getMemoryOverheadData();
2960

2961 2962 2963 2964 2965
        /* Peak memory is updated from time to time by serverCron() so it
         * may happen that the instantaneous value is slightly bigger than
         * the peak value. This may confuse users, so we update the peak
         * if found smaller than the current memory usage. */
        if (zmalloc_used > server.stat_peak_memory)
2966 2967 2968
            server.stat_peak_memory = zmalloc_used;

        bytesToHuman(hmem,zmalloc_used);
2969
        bytesToHuman(peak_hmem,server.stat_peak_memory);
2970
        bytesToHuman(total_system_hmem,total_system_mem);
2971 2972
        bytesToHuman(used_memory_lua_hmem,memory_lua);
        bytesToHuman(used_memory_rss_hmem,server.resident_set_size);
2973
        bytesToHuman(maxmemory_hmem,server.maxmemory);
2974

2975 2976 2977 2978 2979 2980
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Memory\r\n"
            "used_memory:%zu\r\n"
            "used_memory_human:%s\r\n"
            "used_memory_rss:%zu\r\n"
2981
            "used_memory_rss_human:%s\r\n"
2982 2983
            "used_memory_peak:%zu\r\n"
            "used_memory_peak_human:%s\r\n"
2984
            "used_memory_peak_perc:%.2f%%\r\n"
2985 2986 2987 2988
            "used_memory_overhead:%zu\r\n"
            "used_memory_startup:%zu\r\n"
            "used_memory_dataset:%zu\r\n"
            "used_memory_dataset_perc:%.2f%%\r\n"
2989 2990
            "total_system_memory:%lu\r\n"
            "total_system_memory_human:%s\r\n"
A
antirez 已提交
2991
            "used_memory_lua:%lld\r\n"
2992
            "used_memory_lua_human:%s\r\n"
2993 2994 2995
            "maxmemory:%lld\r\n"
            "maxmemory_human:%s\r\n"
            "maxmemory_policy:%s\r\n"
2996
            "mem_fragmentation_ratio:%.2f\r\n"
2997
            "mem_allocator:%s\r\n"
O
oranagra 已提交
2998
            "active_defrag_running:%d\r\n"
2999
            "lazyfree_pending_objects:%zu\r\n",
3000
            zmalloc_used,
3001
            hmem,
A
antirez 已提交
3002
            server.resident_set_size,
3003
            used_memory_rss_hmem,
3004 3005
            server.stat_peak_memory,
            peak_hmem,
3006
            mh->peak_perc,
3007 3008 3009
            mh->overhead_total,
            mh->startup_allocated,
            mh->dataset,
3010
            mh->dataset_perc,
3011 3012
            (unsigned long)total_system_mem,
            total_system_hmem,
3013 3014
            memory_lua,
            used_memory_lua_hmem,
3015 3016 3017
            server.maxmemory,
            maxmemory_hmem,
            evict_policy,
A
antirez 已提交
3018
            mh->fragmentation,
3019
            ZMALLOC_LIB,
O
oranagra 已提交
3020
            server.active_defrag_running,
3021
            lazyfreeGetPendingObjectsCount()
3022 3023
        );
        freeMemoryOverheadData(mh);
3024 3025
    }

3026 3027 3028
    /* Persistence */
    if (allsections || defsections || !strcasecmp(section,"persistence")) {
        if (sections++) info = sdscat(info,"\r\n");
3029
        info = sdscatprintf(info,
3030 3031
            "# Persistence\r\n"
            "loading:%d\r\n"
3032 3033
            "rdb_changes_since_last_save:%lld\r\n"
            "rdb_bgsave_in_progress:%d\r\n"
Y
YAMAMOTO Takashi 已提交
3034
            "rdb_last_save_time:%jd\r\n"
3035
            "rdb_last_bgsave_status:%s\r\n"
Y
YAMAMOTO Takashi 已提交
3036 3037
            "rdb_last_bgsave_time_sec:%jd\r\n"
            "rdb_current_bgsave_time_sec:%jd\r\n"
3038
            "rdb_last_cow_size:%zu\r\n"
3039
            "aof_enabled:%d\r\n"
3040 3041
            "aof_rewrite_in_progress:%d\r\n"
            "aof_rewrite_scheduled:%d\r\n"
Y
YAMAMOTO Takashi 已提交
3042 3043
            "aof_last_rewrite_time_sec:%jd\r\n"
            "aof_current_rewrite_time_sec:%jd\r\n"
3044
            "aof_last_bgrewrite_status:%s\r\n"
3045 3046
            "aof_last_write_status:%s\r\n"
            "aof_last_cow_size:%zu\r\n",
3047 3048
            server.loading,
            server.dirty,
A
antirez 已提交
3049
            server.rdb_child_pid != -1,
Y
YAMAMOTO Takashi 已提交
3050
            (intmax_t)server.lastsave,
3051
            (server.lastbgsave_status == C_OK) ? "ok" : "err",
Y
YAMAMOTO Takashi 已提交
3052 3053 3054
            (intmax_t)server.rdb_save_time_last,
            (intmax_t)((server.rdb_child_pid == -1) ?
                -1 : time(NULL)-server.rdb_save_time_start),
3055
            server.stat_rdb_cow_bytes,
A
antirez 已提交
3056
            server.aof_state != AOF_OFF,
3057
            server.aof_child_pid != -1,
3058
            server.aof_rewrite_scheduled,
Y
YAMAMOTO Takashi 已提交
3059 3060 3061
            (intmax_t)server.aof_rewrite_time_last,
            (intmax_t)((server.aof_child_pid == -1) ?
                -1 : time(NULL)-server.aof_rewrite_time_start),
3062
            (server.aof_lastbgrewrite_status == C_OK) ? "ok" : "err",
3063 3064
            (server.aof_last_write_status == C_OK) ? "ok" : "err",
            server.stat_aof_cow_bytes);
3065

A
antirez 已提交
3066
        if (server.aof_state != AOF_OFF) {
3067 3068 3069
            info = sdscatprintf(info,
                "aof_current_size:%lld\r\n"
                "aof_base_size:%lld\r\n"
3070 3071
                "aof_pending_rewrite:%d\r\n"
                "aof_buffer_length:%zu\r\n"
3072
                "aof_rewrite_buffer_length:%lu\r\n"
3073 3074
                "aof_pending_bio_fsync:%llu\r\n"
                "aof_delayed_fsync:%lu\r\n",
3075 3076 3077
                (long long) server.aof_current_size,
                (long long) server.aof_rewrite_base_size,
                server.aof_rewrite_scheduled,
A
antirez 已提交
3078
                sdslen(server.aof_buf),
3079
                aofRewriteBufferSize(),
A
antirez 已提交
3080
                bioPendingJobsOfType(BIO_AOF_FSYNC),
3081
                server.aof_delayed_fsync);
3082 3083
        }

3084 3085 3086 3087 3088 3089 3090
        if (server.loading) {
            double perc;
            time_t eta, elapsed;
            off_t remaining_bytes = server.loading_total_bytes-
                                    server.loading_loaded_bytes;

            perc = ((double)server.loading_loaded_bytes /
A
antirez 已提交
3091
                   (server.loading_total_bytes+1)) * 100;
3092

A
antirez 已提交
3093
            elapsed = time(NULL)-server.loading_start_time;
3094 3095 3096 3097
            if (elapsed == 0) {
                eta = 1; /* A fake 1 second figure if we don't have
                            enough info */
            } else {
A
antirez 已提交
3098
                eta = (elapsed*remaining_bytes)/(server.loading_loaded_bytes+1);
3099 3100 3101
            }

            info = sdscatprintf(info,
Y
YAMAMOTO Takashi 已提交
3102
                "loading_start_time:%jd\r\n"
3103 3104 3105
                "loading_total_bytes:%llu\r\n"
                "loading_loaded_bytes:%llu\r\n"
                "loading_loaded_perc:%.2f\r\n"
Y
YAMAMOTO Takashi 已提交
3106 3107
                "loading_eta_seconds:%jd\r\n",
                (intmax_t) server.loading_start_time,
3108 3109 3110
                (unsigned long long) server.loading_total_bytes,
                (unsigned long long) server.loading_loaded_bytes,
                perc,
Y
YAMAMOTO Takashi 已提交
3111
                (intmax_t)eta
3112 3113
            );
        }
3114
    }
3115 3116 3117 3118

    /* Stats */
    if (allsections || defsections || !strcasecmp(section,"stats")) {
        if (sections++) info = sdscat(info,"\r\n");
3119
        info = sdscatprintf(info,
3120 3121 3122
            "# Stats\r\n"
            "total_connections_received:%lld\r\n"
            "total_commands_processed:%lld\r\n"
3123
            "instantaneous_ops_per_sec:%lld\r\n"
3124 3125 3126 3127
            "total_net_input_bytes:%lld\r\n"
            "total_net_output_bytes:%lld\r\n"
            "instantaneous_input_kbps:%.2f\r\n"
            "instantaneous_output_kbps:%.2f\r\n"
3128
            "rejected_connections:%lld\r\n"
3129 3130 3131
            "sync_full:%lld\r\n"
            "sync_partial_ok:%lld\r\n"
            "sync_partial_err:%lld\r\n"
3132 3133 3134 3135 3136
            "expired_keys:%lld\r\n"
            "evicted_keys:%lld\r\n"
            "keyspace_hits:%lld\r\n"
            "keyspace_misses:%lld\r\n"
            "pubsub_channels:%ld\r\n"
3137
            "pubsub_patterns:%lu\r\n"
3138
            "latest_fork_usec:%lld\r\n"
3139
            "migrate_cached_sockets:%ld\r\n"
O
oranagra 已提交
3140 3141 3142 3143 3144
            "slave_expires_tracked_keys:%zu\r\n"
            "active_defrag_hits:%lld\r\n"
            "active_defrag_misses:%lld\r\n"
            "active_defrag_key_hits:%lld\r\n"
            "active_defrag_key_misses:%lld\r\n",
3145 3146
            server.stat_numconnections,
            server.stat_numcommands,
A
antirez 已提交
3147
            getInstantaneousMetric(STATS_METRIC_COMMAND),
3148 3149
            server.stat_net_input_bytes,
            server.stat_net_output_bytes,
A
antirez 已提交
3150 3151
            (float)getInstantaneousMetric(STATS_METRIC_NET_INPUT)/1024,
            (float)getInstantaneousMetric(STATS_METRIC_NET_OUTPUT)/1024,
3152
            server.stat_rejected_conn,
3153 3154 3155
            server.stat_sync_full,
            server.stat_sync_partial_ok,
            server.stat_sync_partial_err,
3156 3157 3158 3159 3160
            server.stat_expiredkeys,
            server.stat_evictedkeys,
            server.stat_keyspace_hits,
            server.stat_keyspace_misses,
            dictSize(server.pubsub_channels),
3161
            listLength(server.pubsub_patterns),
3162
            server.stat_fork_time,
3163
            dictSize(server.migrate_cached_sockets),
O
oranagra 已提交
3164 3165 3166 3167 3168
            getSlaveKeyWithExpireCount(),
            server.stat_active_defrag_hits,
            server.stat_active_defrag_misses,
            server.stat_active_defrag_key_hits,
            server.stat_active_defrag_key_misses);
3169
    }
A
antirez 已提交
3170

3171 3172 3173 3174 3175 3176 3177 3178
    /* Replication */
    if (allsections || defsections || !strcasecmp(section,"replication")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
            "# Replication\r\n"
            "role:%s\r\n",
            server.masterhost == NULL ? "master" : "slave");
        if (server.masterhost) {
3179 3180 3181 3182 3183 3184 3185
            long long slave_repl_offset = 1;

            if (server.master)
                slave_repl_offset = server.master->reploff;
            else if (server.cached_master)
                slave_repl_offset = server.cached_master->reploff;

3186 3187 3188 3189 3190 3191
            info = sdscatprintf(info,
                "master_host:%s\r\n"
                "master_port:%d\r\n"
                "master_link_status:%s\r\n"
                "master_last_io_seconds_ago:%d\r\n"
                "master_sync_in_progress:%d\r\n"
3192
                "slave_repl_offset:%lld\r\n"
3193 3194
                ,server.masterhost,
                server.masterport,
A
antirez 已提交
3195
                (server.repl_state == REPL_STATE_CONNECTED) ?
3196 3197
                    "up" : "down",
                server.master ?
3198
                ((int)(server.unixtime-server.master->lastinteraction)) : -1,
A
antirez 已提交
3199
                server.repl_state == REPL_STATE_TRANSFER,
3200
                slave_repl_offset
3201 3202
            );

A
antirez 已提交
3203
            if (server.repl_state == REPL_STATE_TRANSFER) {
3204
                info = sdscatprintf(info,
3205
                    "master_sync_left_bytes:%lld\r\n"
3206
                    "master_sync_last_io_seconds_ago:%d\r\n"
3207 3208
                    , (long long)
                        (server.repl_transfer_size - server.repl_transfer_read),
3209
                    (int)(server.unixtime-server.repl_transfer_lastio)
3210 3211
                );
            }
3212

A
antirez 已提交
3213
            if (server.repl_state != REPL_STATE_CONNECTED) {
3214
                info = sdscatprintf(info,
Y
YAMAMOTO Takashi 已提交
3215 3216
                    "master_link_down_since_seconds:%jd\r\n",
                    (intmax_t)server.unixtime-server.repl_down_since);
3217
            }
3218
            info = sdscatprintf(info,
3219 3220 3221 3222
                "slave_priority:%d\r\n"
                "slave_read_only:%d\r\n",
                server.slave_priority,
                server.repl_slave_ro);
A
antirez 已提交
3223
        }
3224

3225
        info = sdscatprintf(info,
3226
            "connected_slaves:%lu\r\n",
3227
            listLength(server.slaves));
3228 3229 3230 3231 3232 3233 3234 3235 3236 3237

        /* If min-slaves-to-write is active, write the number of slaves
         * currently considered 'good'. */
        if (server.repl_min_slaves_to_write &&
            server.repl_min_slaves_max_lag) {
            info = sdscatprintf(info,
                "min_slaves_good_slaves:%d\r\n",
                server.repl_good_slaves_count);
        }

3238 3239 3240 3241 3242 3243 3244
        if (listLength(server.slaves)) {
            int slaveid = 0;
            listNode *ln;
            listIter li;

            listRewind(server.slaves,&li);
            while((ln = listNext(&li))) {
3245
                client *slave = listNodeValue(ln);
3246
                char *state = NULL;
3247
                char ip[NET_IP_STR_LEN], *slaveip = slave->slave_ip;
3248
                int port;
3249
                long lag = 0;
3250

3251 3252 3253 3254 3255
                if (slaveip[0] == '\0') {
                    if (anetPeerToString(slave->fd,ip,sizeof(ip),&port) == -1)
                        continue;
                    slaveip = ip;
                }
3256
                switch(slave->replstate) {
A
antirez 已提交
3257 3258
                case SLAVE_STATE_WAIT_BGSAVE_START:
                case SLAVE_STATE_WAIT_BGSAVE_END:
3259 3260
                    state = "wait_bgsave";
                    break;
A
antirez 已提交
3261
                case SLAVE_STATE_SEND_BULK:
3262 3263
                    state = "send_bulk";
                    break;
A
antirez 已提交
3264
                case SLAVE_STATE_ONLINE:
3265 3266 3267 3268
                    state = "online";
                    break;
                }
                if (state == NULL) continue;
A
antirez 已提交
3269
                if (slave->replstate == SLAVE_STATE_ONLINE)
3270 3271 3272 3273
                    lag = time(NULL) - slave->repl_ack_time;

                info = sdscatprintf(info,
                    "slave%d:ip=%s,port=%d,state=%s,"
3274
                    "offset=%lld,lag=%ld\r\n",
3275
                    slaveid,slaveip,slave->slave_listening_port,state,
3276
                    slave->repl_ack_off, lag);
3277 3278 3279
                slaveid++;
            }
        }
3280
        info = sdscatprintf(info,
3281 3282
            "master_replid:%s\r\n"
            "master_replid2:%s\r\n"
3283
            "master_repl_offset:%lld\r\n"
3284
            "second_repl_offset:%lld\r\n"
3285 3286 3287 3288
            "repl_backlog_active:%d\r\n"
            "repl_backlog_size:%lld\r\n"
            "repl_backlog_first_byte_offset:%lld\r\n"
            "repl_backlog_histlen:%lld\r\n",
3289 3290
            server.replid,
            server.replid2,
3291
            server.master_repl_offset,
3292
            server.second_replid_offset,
3293 3294 3295 3296
            server.repl_backlog != NULL,
            server.repl_backlog_size,
            server.repl_backlog_off,
            server.repl_backlog_histlen);
A
antirez 已提交
3297 3298
    }

3299 3300
    /* CPU */
    if (allsections || defsections || !strcasecmp(section,"cpu")) {
3301 3302
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
3303
        "# CPU\r\n"
3304 3305
        "used_cpu_sys:%.2f\r\n"
        "used_cpu_user:%.2f\r\n"
3306 3307
        "used_cpu_sys_children:%.2f\r\n"
        "used_cpu_user_children:%.2f\r\n",
3308
        (float)self_ru.ru_stime.tv_sec+(float)self_ru.ru_stime.tv_usec/1000000,
3309 3310 3311
        (float)self_ru.ru_utime.tv_sec+(float)self_ru.ru_utime.tv_usec/1000000,
        (float)c_ru.ru_stime.tv_sec+(float)c_ru.ru_stime.tv_usec/1000000,
        (float)c_ru.ru_utime.tv_sec+(float)c_ru.ru_utime.tv_usec/1000000);
3312
    }
3313

3314
    /* Command statistics */
3315 3316 3317 3318
    if (allsections || !strcasecmp(section,"commandstats")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Commandstats\r\n");

3319 3320 3321 3322 3323 3324
        struct redisCommand *c;
        dictEntry *de;
        dictIterator *di;
        di = dictGetSafeIterator(server.commands);
        while((de = dictNext(di)) != NULL) {
            c = (struct redisCommand *) dictGetVal(de);
3325 3326 3327 3328 3329
            if (!c->calls) continue;
            info = sdscatprintf(info,
                "cmdstat_%s:calls=%lld,usec=%lld,usec_per_call=%.2f\r\n",
                c->name, c->calls, c->microseconds,
                (c->calls == 0) ? 0 : ((float)c->microseconds/c->calls));
3330
        }
3331
        dictReleaseIterator(di);
A
antirez 已提交
3332 3333
    }

3334
    /* Cluster */
3335 3336 3337 3338 3339 3340 3341 3342
    if (allsections || defsections || !strcasecmp(section,"cluster")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info,
        "# Cluster\r\n"
        "cluster_enabled:%d\r\n",
        server.cluster_enabled);
    }

3343 3344 3345 3346 3347 3348
    /* Key space */
    if (allsections || defsections || !strcasecmp(section,"keyspace")) {
        if (sections++) info = sdscat(info,"\r\n");
        info = sdscatprintf(info, "# Keyspace\r\n");
        for (j = 0; j < server.dbnum; j++) {
            long long keys, vkeys;
3349

3350 3351 3352
            keys = dictSize(server.db[j].dict);
            vkeys = dictSize(server.db[j].expires);
            if (keys || vkeys) {
3353 3354 3355
                info = sdscatprintf(info,
                    "db%d:keys=%lld,expires=%lld,avg_ttl=%lld\r\n",
                    j, keys, vkeys, server.db[j].avg_ttl);
3356
            }
3357 3358 3359 3360 3361
        }
    }
    return info;
}

3362
void infoCommand(client *c) {
3363 3364 3365 3366 3367 3368
    char *section = c->argc == 2 ? c->argv[1]->ptr : "default";

    if (c->argc > 2) {
        addReply(c,shared.syntaxerr);
        return;
    }
M
Matt Stancliff 已提交
3369
    addReplyBulkSds(c, genRedisInfoString(section));
3370 3371
}

3372
void monitorCommand(client *c) {
3373
    /* ignore MONITOR if already slave or in monitor mode */
A
antirez 已提交
3374
    if (c->flags & CLIENT_SLAVE) return;
3375

A
antirez 已提交
3376
    c->flags |= (CLIENT_SLAVE|CLIENT_MONITOR);
3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397
    listAddNodeTail(server.monitors,c);
    addReply(c,shared.ok);
}

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

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

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

    return atoi(buf);
}

3398
void linuxMemoryWarnings(void) {
3399
    if (linuxOvercommitMemoryValue() == 0) {
A
antirez 已提交
3400
        serverLog(LL_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.");
3401
    }
3402
    if (THPIsEnabled()) {
A
antirez 已提交
3403
        serverLog(LL_WARNING,"WARNING you have Transparent Huge Pages (THP) support enabled in your kernel. This will create latency and memory usage issues with Redis. To fix this issue run the command 'echo never > /sys/kernel/mm/transparent_hugepage/enabled' as root, and add it to your /etc/rc.local in order to retain the setting after a reboot. Redis must be restarted after THP is disabled.");
3404
    }
3405 3406 3407
}
#endif /* __linux__ */

3408
void createPidFile(void) {
3409 3410
    /* If pidfile requested, but no pidfile defined, use
     * default pidfile path */
3411
    if (!server.pidfile) server.pidfile = zstrdup(CONFIG_DEFAULT_PID_FILE);
3412

3413 3414 3415
    /* Try to write the pid file in a best-effort way. */
    FILE *fp = fopen(server.pidfile,"w");
    if (fp) {
3416
        fprintf(fp,"%d\n",(int)getpid());
3417 3418 3419 3420
        fclose(fp);
    }
}

3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437
void daemonize(void) {
    int fd;

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

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

3438
void version(void) {
3439
    printf("Redis server v=%s sha=%s:%d malloc=%s bits=%d build=%llx\n",
A
antirez 已提交
3440 3441 3442 3443
        REDIS_VERSION,
        redisGitSHA1(),
        atoi(redisGitDirty()) > 0,
        ZMALLOC_LIB,
3444
        sizeof(long) == 4 ? 32 : 64,
3445
        (unsigned long long) redisBuildId());
3446 3447 3448
    exit(0);
}

3449
void usage(void) {
3450
    fprintf(stderr,"Usage: ./redis-server [/path/to/redis.conf] [options]\n");
3451
    fprintf(stderr,"       ./redis-server - (read config from stdin)\n");
3452
    fprintf(stderr,"       ./redis-server -v or --version\n");
3453 3454
    fprintf(stderr,"       ./redis-server -h or --help\n");
    fprintf(stderr,"       ./redis-server --test-memory <megabytes>\n\n");
3455 3456 3457 3458 3459
    fprintf(stderr,"Examples:\n");
    fprintf(stderr,"       ./redis-server (run the server with default conf)\n");
    fprintf(stderr,"       ./redis-server /etc/redis/6379.conf\n");
    fprintf(stderr,"       ./redis-server --port 7777\n");
    fprintf(stderr,"       ./redis-server --port 7777 --slaveof 127.0.0.1 8888\n");
3460 3461 3462
    fprintf(stderr,"       ./redis-server /etc/myredis.conf --loglevel verbose\n\n");
    fprintf(stderr,"Sentinel mode:\n");
    fprintf(stderr,"       ./redis-server /etc/sentinel.conf --sentinel\n");
3463 3464 3465
    exit(1);
}

A
antirez 已提交
3466 3467 3468
void redisAsciiArt(void) {
#include "asciilogo.h"
    char *buf = zmalloc(1024*16);
3469
    char *mode;
3470 3471 3472

    if (server.cluster_enabled) mode = "cluster";
    else if (server.sentinel_mode) mode = "sentinel";
3473
    else mode = "standalone";
A
antirez 已提交
3474

3475 3476 3477 3478 3479 3480 3481 3482 3483
    /* Show the ASCII logo if: log file is stdout AND stdout is a
     * tty AND syslog logging is disabled. Also show logo if the user
     * forced us to do so via redis.conf. */
    int show_logo = ((!server.syslog_enabled &&
                      server.logfile[0] == '\0' &&
                      isatty(fileno(stdout))) ||
                     server.always_show_logo);

    if (!show_logo) {
A
antirez 已提交
3484
        serverLog(LL_NOTICE,
3485 3486
            "Running mode=%s, port=%d.",
            mode, server.port
3487 3488 3489 3490 3491 3492 3493 3494 3495 3496
        );
    } else {
        snprintf(buf,1024*16,ascii_logo,
            REDIS_VERSION,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
            (sizeof(long) == 8) ? "64" : "32",
            mode, server.port,
            (long) getpid()
        );
A
antirez 已提交
3497
        serverLogRaw(LL_NOTICE|LL_RAW,buf);
3498
    }
A
antirez 已提交
3499 3500 3501
    zfree(buf);
}

3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515
static void sigShutdownHandler(int sig) {
    char *msg;

    switch (sig) {
    case SIGINT:
        msg = "Received SIGINT scheduling shutdown...";
        break;
    case SIGTERM:
        msg = "Received SIGTERM scheduling shutdown...";
        break;
    default:
        msg = "Received shutdown signal, scheduling shutdown...";
    };

3516 3517 3518 3519 3520
    /* SIGINT is often delivered via Ctrl+C in an interactive session.
     * If we receive the signal the second time, we interpret this as
     * the user really wanting to quit ASAP without waiting to persist
     * on disk. */
    if (server.shutdown_asap && sig == SIGINT) {
A
antirez 已提交
3521
        serverLogFromHandler(LL_WARNING, "You insist... exiting now.");
3522 3523 3524 3525 3526 3527
        rdbRemoveTempFile(getpid());
        exit(1); /* Exit with an error since this was not a clean shutdown. */
    } else if (server.loading) {
        exit(0);
    }

A
antirez 已提交
3528
    serverLogFromHandler(LL_WARNING, msg);
3529 3530 3531
    server.shutdown_asap = 1;
}

3532
void setupSignalHandlers(void) {
3533 3534
    struct sigaction act;

3535 3536 3537
    /* When the SA_SIGINFO flag is set in sa_flags then sa_sigaction is used.
     * Otherwise, sa_handler is used. */
    sigemptyset(&act.sa_mask);
3538
    act.sa_flags = 0;
3539
    act.sa_handler = sigShutdownHandler;
3540
    sigaction(SIGTERM, &act, NULL);
3541
    sigaction(SIGINT, &act, NULL);
3542

3543 3544
#ifdef HAVE_BACKTRACE
    sigemptyset(&act.sa_mask);
3545
    act.sa_flags = SA_NODEFER | SA_RESETHAND | SA_SIGINFO;
3546 3547 3548 3549 3550 3551 3552
    act.sa_sigaction = sigsegvHandler;
    sigaction(SIGSEGV, &act, NULL);
    sigaction(SIGBUS, &act, NULL);
    sigaction(SIGFPE, &act, NULL);
    sigaction(SIGILL, &act, NULL);
#endif
    return;
3553 3554
}

3555 3556
void memtest(size_t megabytes, int passes);

3557
/* Returns 1 if there is --sentinel among the arguments or if
3558
 * argv[0] contains "redis-sentinel". */
3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570
int checkForSentinelMode(int argc, char **argv) {
    int j;

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

/* Function called at startup to load RDB or AOF file in memory. */
void loadDataFromDisk(void) {
    long long start = ustime();
A
antirez 已提交
3571
    if (server.aof_state == AOF_ON) {
3572
        if (loadAppendOnlyFile(server.aof_filename) == C_OK)
A
antirez 已提交
3573
            serverLog(LL_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
3574
    } else {
3575 3576
        rdbSaveInfo rsi = RDB_SAVE_INFO_INIT;
        if (rdbLoad(server.rdb_filename,&rsi) == C_OK) {
A
antirez 已提交
3577
            serverLog(LL_NOTICE,"DB loaded from disk: %.3f seconds",
3578
                (float)(ustime()-start)/1000000);
3579 3580

            /* Restore the replication ID / offset from the RDB file. */
3581 3582 3583 3584
            if (server.masterhost &&
                rsi.repl_id_is_set &&
                rsi.repl_offset != -1 &&
                /* Note that older implementations may save a repl_stream_db
3585 3586
                 * of -1 inside the RDB file in a wrong way, see more information
                 * in function rdbPopulateSaveInfo. */
3587 3588
                rsi.repl_stream_db != -1)
            {
3589 3590 3591 3592 3593
                memcpy(server.replid,rsi.repl_id,sizeof(server.replid));
                server.master_repl_offset = rsi.repl_offset;
                /* If we are a slave, create a cached master from this
                 * information, in order to allow partial resynchronizations
                 * with masters. */
3594 3595
                replicationCacheMasterUsingMyself();
                selectDb(server.cached_master,rsi.repl_stream_db);
3596
            }
3597
        } else if (errno != ENOENT) {
A
antirez 已提交
3598
            serverLog(LL_WARNING,"Fatal error loading the DB: %s. Exiting.",strerror(errno));
3599 3600 3601 3602 3603
            exit(1);
        }
    }
}

A
antirez 已提交
3604
void redisOutOfMemoryHandler(size_t allocation_size) {
A
antirez 已提交
3605
    serverLog(LL_WARNING,"Out Of Memory allocating %zu bytes!",
A
antirez 已提交
3606
        allocation_size);
A
antirez 已提交
3607
    serverPanic("Redis aborting for OUT OF MEMORY");
A
antirez 已提交
3608 3609
}

3610
void redisSetProcTitle(char *title) {
3611
#ifdef USE_SETPROCTITLE
3612 3613 3614 3615 3616
    char *server_mode = "";
    if (server.cluster_enabled) server_mode = " [cluster]";
    else if (server.sentinel_mode) server_mode = " [sentinel]";

    setproctitle("%s %s:%d%s",
3617
        title,
A
antirez 已提交
3618
        server.bindaddr_count ? server.bindaddr[0] : "*",
3619 3620
        server.port,
        server_mode);
3621
#else
A
antirez 已提交
3622
    UNUSED(title);
3623
#endif
3624 3625
}

3626 3627 3628
/*
 * Check whether systemd or upstart have been used to start redis.
 */
3629 3630

int redisSupervisedUpstart(void) {
3631
    const char *upstart_job = getenv("UPSTART_JOB");
3632 3633

    if (!upstart_job) {
A
antirez 已提交
3634
        serverLog(LL_WARNING,
3635 3636 3637 3638
                "upstart supervision requested, but UPSTART_JOB not found");
        return 0;
    }

I
Itamar Haber 已提交
3639
    serverLog(LL_NOTICE, "supervised by upstart, will stop to signal readiness");
3640 3641 3642 3643 3644 3645
    raise(SIGSTOP);
    unsetenv("UPSTART_JOB");
    return 1;
}

int redisSupervisedSystemd(void) {
3646 3647 3648 3649 3650
    const char *notify_socket = getenv("NOTIFY_SOCKET");
    int fd = 1;
    struct sockaddr_un su;
    struct iovec iov;
    struct msghdr hdr;
3651
    int sendto_flags = 0;
3652

3653
    if (!notify_socket) {
A
antirez 已提交
3654
        serverLog(LL_WARNING,
3655
                "systemd supervision requested, but NOTIFY_SOCKET not found");
3656 3657 3658
        return 0;
    }

3659
    if ((strchr("@/", notify_socket[0])) == NULL || strlen(notify_socket) < 2) {
3660
        return 0;
3661
    }
3662

I
Itamar Haber 已提交
3663
    serverLog(LL_NOTICE, "supervised by systemd, will signal readiness");
3664
    if ((fd = socket(AF_UNIX, SOCK_DGRAM, 0)) == -1) {
A
antirez 已提交
3665
        serverLog(LL_WARNING,
3666
                "Can't connect to systemd socket %s", notify_socket);
3667 3668 3669
        return 0;
    }

3670
    memset(&su, 0, sizeof(su));
3671 3672 3673 3674 3675 3676 3677
    su.sun_family = AF_UNIX;
    strncpy (su.sun_path, notify_socket, sizeof(su.sun_path) -1);
    su.sun_path[sizeof(su.sun_path) - 1] = '\0';

    if (notify_socket[0] == '@')
        su.sun_path[0] = '\0';

3678
    memset(&iov, 0, sizeof(iov));
3679 3680 3681
    iov.iov_base = "READY=1";
    iov.iov_len = strlen("READY=1");

3682
    memset(&hdr, 0, sizeof(hdr));
3683 3684 3685 3686 3687 3688 3689
    hdr.msg_name = &su;
    hdr.msg_namelen = offsetof(struct sockaddr_un, sun_path) +
        strlen(notify_socket);
    hdr.msg_iov = &iov;
    hdr.msg_iovlen = 1;

    unsetenv("NOTIFY_SOCKET");
3690 3691 3692 3693
#ifdef HAVE_MSG_NOSIGNAL
    sendto_flags |= MSG_NOSIGNAL;
#endif
    if (sendmsg(fd, &hdr, sendto_flags) < 0) {
A
antirez 已提交
3694
        serverLog(LL_WARNING, "Can't send notification to systemd");
3695 3696 3697 3698 3699 3700 3701
        close(fd);
        return 0;
    }
    close(fd);
    return 1;
}

3702
int redisIsSupervised(int mode) {
A
antirez 已提交
3703
    if (mode == SUPERVISED_AUTODETECT) {
3704 3705 3706 3707 3708 3709 3710 3711
        const char *upstart_job = getenv("UPSTART_JOB");
        const char *notify_socket = getenv("NOTIFY_SOCKET");

        if (upstart_job) {
            redisSupervisedUpstart();
        } else if (notify_socket) {
            redisSupervisedSystemd();
        }
A
antirez 已提交
3712
    } else if (mode == SUPERVISED_UPSTART) {
3713
        return redisSupervisedUpstart();
A
antirez 已提交
3714
    } else if (mode == SUPERVISED_SYSTEMD) {
3715 3716 3717 3718 3719 3720 3721
        return redisSupervisedSystemd();
    }

    return 0;
}


3722
int main(int argc, char **argv) {
3723
    struct timeval tv;
A
antirez 已提交
3724
    int j;
3725

3726 3727 3728 3729
#ifdef REDIS_TEST
    if (argc == 3 && !strcasecmp(argv[1], "test")) {
        if (!strcasecmp(argv[2], "ziplist")) {
            return ziplistTest(argc, argv);
M
Matt Stancliff 已提交
3730 3731
        } else if (!strcasecmp(argv[2], "quicklist")) {
            quicklistTest(argc, argv);
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751
        } else if (!strcasecmp(argv[2], "intset")) {
            return intsetTest(argc, argv);
        } else if (!strcasecmp(argv[2], "zipmap")) {
            return zipmapTest(argc, argv);
        } else if (!strcasecmp(argv[2], "sha1test")) {
            return sha1Test(argc, argv);
        } else if (!strcasecmp(argv[2], "util")) {
            return utilTest(argc, argv);
        } else if (!strcasecmp(argv[2], "sds")) {
            return sdsTest(argc, argv);
        } else if (!strcasecmp(argv[2], "endianconv")) {
            return endianconvTest(argc, argv);
        } else if (!strcasecmp(argv[2], "crc64")) {
            return crc64Test(argc, argv);
        }

        return -1; /* test not found */
    }
#endif

A
antirez 已提交
3752
    /* We need to initialize our libraries, and the server configuration. */
3753 3754 3755
#ifdef INIT_SETPROCTITLE_REPLACEMENT
    spt_init(argc, argv);
#endif
3756
    setlocale(LC_COLLATE,"");
A
antirez 已提交
3757
    zmalloc_set_oom_handler(redisOutOfMemoryHandler);
3758 3759
    srand(time(NULL)^getpid());
    gettimeofday(&tv,NULL);
3760 3761 3762
    char hashseed[16];
    getRandomHexChars(hashseed,sizeof(hashseed));
    dictSetHashFunctionSeed((uint8_t*)hashseed);
3763
    server.sentinel_mode = checkForSentinelMode(argc,argv);
3764
    initServerConfig();
A
antirez 已提交
3765
    moduleInitModulesSystem();
3766

A
antirez 已提交
3767 3768 3769 3770 3771 3772 3773
    /* Store the executable path and arguments in a safe place in order
     * to be able to restart the server later. */
    server.executable = getAbsolutePath(argv[0]);
    server.exec_argv = zmalloc(sizeof(char*)*(argc+1));
    server.exec_argv[argc] = NULL;
    for (j = 0; j < argc; j++) server.exec_argv[j] = zstrdup(argv[j]);

3774 3775 3776 3777 3778 3779 3780 3781
    /* We need to init sentinel right now as parsing the configuration file
     * in sentinel mode will have the effect of populating the sentinel
     * data structures with master nodes to monitor. */
    if (server.sentinel_mode) {
        initSentinelConfig();
        initSentinel();
    }

3782
    /* Check if we need to start in redis-check-rdb/aof mode. We just execute
3783 3784 3785
     * the program main. However the program is part of the Redis executable
     * so that we can easily execute an RDB check on loading errors. */
    if (strstr(argv[0],"redis-check-rdb") != NULL)
3786 3787 3788
        redis_check_rdb_main(argc,argv,NULL);
    else if (strstr(argv[0],"redis-check-aof") != NULL)
        redis_check_aof_main(argc,argv);
3789

3790
    if (argc >= 2) {
A
antirez 已提交
3791
        j = 1; /* First option to parse in argv[] */
3792 3793 3794 3795
        sds options = sdsempty();
        char *configfile = NULL;

        /* Handle special options --help and --version */
3796 3797
        if (strcmp(argv[1], "-v") == 0 ||
            strcmp(argv[1], "--version") == 0) version();
3798 3799
        if (strcmp(argv[1], "--help") == 0 ||
            strcmp(argv[1], "-h") == 0) usage();
3800 3801
        if (strcmp(argv[1], "--test-memory") == 0) {
            if (argc == 3) {
3802
                memtest(atoi(argv[2]),50);
3803 3804 3805 3806 3807 3808 3809 3810
                exit(0);
            } else {
                fprintf(stderr,"Please specify the amount of memory to test in megabytes.\n");
                fprintf(stderr,"Example: ./redis-server --test-memory 4096\n\n");
                exit(1);
            }
        }

3811
        /* First argument is the config file name? */
A
antirez 已提交
3812 3813 3814 3815 3816 3817 3818 3819 3820 3821
        if (argv[j][0] != '-' || argv[j][1] != '-') {
            configfile = argv[j];
            server.configfile = getAbsolutePath(configfile);
            /* Replace the config file in server.exec_argv with
             * its absoulte path. */
            zfree(server.exec_argv[j]);
            server.exec_argv[j] = zstrdup(server.configfile);
            j++;
        }

3822 3823 3824 3825 3826 3827 3828
        /* All the other options are parsed and conceptually appended to the
         * configuration file. For instance --port 6380 will generate the
         * string "port 6380\n" to be parsed after the actual file name
         * is parsed, if any. */
        while(j != argc) {
            if (argv[j][0] == '-' && argv[j][1] == '-') {
                /* Option name */
3829 3830 3831 3832 3833
                if (!strcmp(argv[j], "--check-rdb")) {
                    /* Argument has no options, need to skip for parsing. */
                    j++;
                    continue;
                }
3834 3835 3836 3837 3838 3839 3840 3841 3842 3843
                if (sdslen(options)) options = sdscat(options,"\n");
                options = sdscat(options,argv[j]+2);
                options = sdscat(options," ");
            } else {
                /* Option argument */
                options = sdscatrepr(options,argv[j],strlen(argv[j]));
                options = sdscat(options," ");
            }
            j++;
        }
3844
        if (server.sentinel_mode && configfile && *configfile == '-') {
A
antirez 已提交
3845
            serverLog(LL_WARNING,
3846
                "Sentinel config from STDIN not allowed.");
A
antirez 已提交
3847
            serverLog(LL_WARNING,
3848 3849 3850
                "Sentinel needs config file on disk to save state.  Exiting...");
            exit(1);
        }
3851
        resetServerSaveParams();
3852 3853
        loadServerConfig(configfile,options);
        sdsfree(options);
3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865
    }

    serverLog(LL_WARNING, "oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo");
    serverLog(LL_WARNING,
        "Redis version=%s, bits=%d, commit=%s, modified=%d, pid=%d, just started",
            REDIS_VERSION,
            (sizeof(long) == 8) ? 64 : 32,
            redisGitSHA1(),
            strtol(redisGitDirty(),NULL,10) > 0,
            (int)getpid());

    if (argc == 1) {
A
antirez 已提交
3866
        serverLog(LL_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");
3867 3868
    } else {
        serverLog(LL_WARNING, "Configuration loaded");
3869
    }
3870

3871 3872
    server.supervised = redisIsSupervised(server.supervised_mode);
    int background = server.daemonize && !server.supervised;
R
rebx 已提交
3873
    if (background) daemonize();
3874

3875
    initServer();
R
rebx 已提交
3876
    if (background || server.pidfile) createPidFile();
3877
    redisSetProcTitle(argv[0]);
3878
    redisAsciiArt();
3879
    checkTcpBacklogSettings();
3880 3881

    if (!server.sentinel_mode) {
A
antirez 已提交
3882
        /* Things not needed when running in Sentinel mode. */
3883
        serverLog(LL_WARNING,"Server initialized");
3884
    #ifdef __linux__
3885
        linuxMemoryWarnings();
3886
    #endif
A
antirez 已提交
3887
        moduleLoadFromQueue();
3888
        loadDataFromDisk();
3889
        if (server.cluster_enabled) {
3890
            if (verifyClusterConfigWithData() == C_ERR) {
A
antirez 已提交
3891
                serverLog(LL_WARNING,
3892 3893 3894 3895 3896
                    "You can't have keys in a DB different than DB 0 when in "
                    "Cluster mode. Exiting.");
                exit(1);
            }
        }
3897
        if (server.ipfd_count > 0)
3898
            serverLog(LL_NOTICE,"Ready to accept connections");
3899
        if (server.sofd > 0)
A
antirez 已提交
3900
            serverLog(LL_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
3901
    } else {
A
antirez 已提交
3902
        sentinelIsRunning();
3903
    }
3904

3905 3906
    /* Warning the user about suspicious maxmemory setting. */
    if (server.maxmemory > 0 && server.maxmemory < 1024*1024) {
A
antirez 已提交
3907
        serverLog(LL_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);
3908 3909
    }

3910
    aeSetBeforeSleepProc(server.el,beforeSleep);
3911
    aeSetAfterSleepProc(server.el,afterSleep);
3912 3913 3914 3915 3916
    aeMain(server.el);
    aeDeleteEventLoop(server.el);
    return 0;
}

3917
/* The End */