* 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.
+ * 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.
*/
struct redisCommand redisCommandTable[] = {
{"get",getCommand,2,"r",0,NULL,1,1,1,0,0},
{"exec",execCommand,1,"s",0,NULL,0,0,0,0,0},
{"discard",discardCommand,1,"rs",0,NULL,0,0,0,0,0},
{"sync",syncCommand,1,"ars",0,NULL,0,0,0,0,0},
+ {"replconf",replconfCommand,-1,"ars",0,NULL,0,0,0,0,0},
{"flushdb",flushdbCommand,1,"w",0,NULL,0,0,0,0,0},
{"flushall",flushallCommand,1,"w",0,NULL,0,0,0,0,0},
{"sort",sortCommand,-2,"wmS",0,NULL,1,1,1,0,0},
- {"info",infoCommand,-1,"r",0,NULL,0,0,0,0,0},
+ {"info",infoCommand,-1,"rlt",0,NULL,0,0,0,0,0},
{"monitor",monitorCommand,1,"ars",0,NULL,0,0,0,0,0},
{"ttl",ttlCommand,2,"r",0,NULL,1,1,1,0,0},
{"pttl",pttlCommand,2,"r",0,NULL,1,1,1,0,0},
{"persist",persistCommand,2,"w",0,NULL,1,1,1,0,0},
- {"slaveof",slaveofCommand,3,"as",0,NULL,0,0,0,0,0},
+ {"slaveof",slaveofCommand,3,"ast",0,NULL,0,0,0,0,0},
{"debug",debugCommand,-2,"as",0,NULL,0,0,0,0,0},
{"config",configCommand,-2,"ar",0,NULL,0,0,0,0,0},
- {"subscribe",subscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
- {"unsubscribe",unsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
- {"psubscribe",psubscribeCommand,-2,"rps",0,NULL,0,0,0,0,0},
- {"punsubscribe",punsubscribeCommand,-1,"rps",0,NULL,0,0,0,0,0},
- {"publish",publishCommand,3,"pf",0,NULL,0,0,0,0,0},
+ {"subscribe",subscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
+ {"unsubscribe",unsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
+ {"psubscribe",psubscribeCommand,-2,"rpslt",0,NULL,0,0,0,0,0},
+ {"punsubscribe",punsubscribeCommand,-1,"rpslt",0,NULL,0,0,0,0,0},
+ {"publish",publishCommand,3,"pflt",0,NULL,0,0,0,0,0},
{"watch",watchCommand,-2,"rs",0,noPreloadGetKeys,1,-1,1,0,0},
{"unwatch",unwatchCommand,1,"rs",0,NULL,0,0,0,0,0},
{"cluster",clusterCommand,-2,"ar",0,NULL,0,0,0,0,0},
{"evalsha",evalShaCommand,-3,"s",0,zunionInterGetKeys,0,0,0,0,0},
{"slowlog",slowlogCommand,-2,"r",0,NULL,0,0,0,0,0},
{"script",scriptCommand,-2,"ras",0,NULL,0,0,0,0,0},
- {"time",timeCommand,1,"rR",0,NULL,0,0,0,0,0}
+ {"time",timeCommand,1,"rR",0,NULL,0,0,0,0,0},
+ {"bitop",bitopCommand,-4,"wm",0,NULL,2,-1,1,0,0},
+ {"bitcount",bitcountCommand,-2,"r",0,NULL,1,1,1,0,0}
};
/*============================ Utility functions ============================ */
int j;
for (j = 0; j < server.dbnum; j++) {
+ /* Keys dictionary */
if (dictIsRehashing(server.db[j].dict)) {
dictRehashMilliseconds(server.db[j].dict,1);
break; /* already used our millisecond for this loop... */
}
+ /* Expires */
+ if (dictIsRehashing(server.db[j].expires)) {
+ dictRehashMilliseconds(server.db[j].expires,1);
+ break; /* already used our millisecond for this loop... */
+ }
}
}
* it will get more aggressive to avoid that too much memory is used by
* keys that can be removed from the keyspace. */
void activeExpireCycle(void) {
- int j;
+ int j, iteration = 0;
+ long long start = ustime(), timelimit;
+
+ /* We can use at max REDIS_EXPIRELOOKUPS_TIME_PERC percentage of CPU time
+ * per iteration. Since this function gets called with a frequency of
+ * REDIS_HZ times per second, the following is the max amount of
+ * microseconds we can spend in this function. */
+ timelimit = 1000000*REDIS_EXPIRELOOKUPS_TIME_PERC/REDIS_HZ/100;
+ if (timelimit <= 0) timelimit = 1;
for (j = 0; j < server.dbnum; j++) {
int expired;
/* Continue to expire if at the end of the cycle more than 25%
* of the keys were expired. */
do {
- long num = dictSize(db->expires);
+ unsigned long num = dictSize(db->expires);
+ unsigned long slots = dictSlots(db->expires);
long long now = mstime();
+ /* When there are less than 1% filled slots getting random
+ * keys is expensive, so stop here waiting for better times...
+ * The dictionary will be resized asap. */
+ if (num && slots > DICT_HT_INITIAL_SIZE &&
+ (num*100/slots < 1)) break;
+
+ /* The main collection cycle. Sample random keys among keys
+ * with an expire set, checking for expired ones. */
expired = 0;
if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
num = REDIS_EXPIRELOOKUPS_PER_CRON;
server.stat_expiredkeys++;
}
}
+ /* We can't block forever here even if there are many keys to
+ * expire. So after a given amount of milliseconds return to the
+ * caller waiting for the other active expire cycle. */
+ iteration++;
+ if ((iteration & 0xf) == 0 && /* check once every 16 cycles. */
+ (ustime()-start) > timelimit) return;
} while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
}
}
}
void clientsCron(void) {
- /* Make sure to process at least 1/100 of clients per call.
- * Since this function is called 10 times per second we are sure that
+ /* Make sure to process at least 1/(REDIS_HZ*10) of clients per call.
+ * Since this function is called REDIS_HZ times per second we are sure that
* in the worst case we process all the clients in 10 seconds.
* In normal conditions (a reasonable number of clients) we process
* all the clients in a shorter time. */
int numclients = listLength(server.clients);
- int iterations = numclients/100;
+ int iterations = numclients/(REDIS_HZ*10);
if (iterations < 50)
iterations = (numclients < 50) ? numclients : 50;
}
}
+/* This is our timer interrupt, called REDIS_HZ times per second.
+ * Here is where we do a number of things that need to be done asynchronously.
+ * For instance:
+ *
+ * - Active expired keys collection (it is also performed in a lazy way on
+ * lookup).
+ * - Software watchdong.
+ * - Update some statistic.
+ * - Incremental rehashing of the DBs hash tables.
+ * - Triggering BGSAVE / AOF rewrite, and handling of terminated children.
+ * - Clients timeout of differnet kinds.
+ * - Replication reconnection.
+ * - Many more...
+ *
+ * Everything directly called here will be called REDIS_HZ times per second,
+ * so in order to throttle execution of things we want to do less frequently
+ * a macro is used: run_with_period(milliseconds) { .... }
+ */
+
int serverCron(struct aeEventLoop *eventLoop, long long id, void *clientData) {
- int j, loops = server.cronloops;
+ int j;
REDIS_NOTUSED(eventLoop);
REDIS_NOTUSED(id);
REDIS_NOTUSED(clientData);
* To access a global var is faster than calling time(NULL) */
server.unixtime = time(NULL);
- trackOperationsPerSecond();
+ run_with_period(100) trackOperationsPerSecond();
/* We have just 22 bits per object for LRU information.
* So we use an (eventually wrapping) LRU clock with 10 seconds resolution.
}
/* Show some info about non-empty databases */
- for (j = 0; j < server.dbnum; j++) {
- long long size, used, vkeys;
-
- size = dictSlots(server.db[j].dict);
- used = dictSize(server.db[j].dict);
- vkeys = dictSize(server.db[j].expires);
- if (!(loops % 50) && (used || vkeys)) {
- redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
- /* dictPrintStats(server.dict); */
+ run_with_period(5000) {
+ for (j = 0; j < server.dbnum; j++) {
+ long long size, used, vkeys;
+
+ size = dictSlots(server.db[j].dict);
+ used = dictSize(server.db[j].dict);
+ vkeys = dictSize(server.db[j].expires);
+ if (used || vkeys) {
+ redisLog(REDIS_VERBOSE,"DB %d: %lld keys (%lld volatile) in %lld slots HT.",j,used,vkeys,size);
+ /* dictPrintStats(server.dict); */
+ }
}
}
* a lot of memory movements in the parent will cause a lot of pages
* copied. */
if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) {
- if (!(loops % 10)) tryResizeHashTables();
+ tryResizeHashTables();
if (server.activerehashing) incrementallyRehash();
}
/* Show information about connected clients */
- if (!(loops % 50)) {
- redisLog(REDIS_VERBOSE,"%d clients connected (%d slaves), %zu bytes in use",
- listLength(server.clients)-listLength(server.slaves),
- listLength(server.slaves),
- zmalloc_used_memory());
+ if (!server.sentinel_mode) {
+ run_with_period(5000) {
+ redisLog(REDIS_VERBOSE,
+ "%d clients connected (%d slaves), %zu bytes in use",
+ listLength(server.clients)-listLength(server.slaves),
+ listLength(server.slaves),
+ zmalloc_used_memory());
+ }
}
/* We need to do a few operations on clients asynchronously. */
/* Replication cron function -- used to reconnect to master and
* to detect transfer failures. */
- if (!(loops % 10)) replicationCron();
+ run_with_period(1000) replicationCron();
/* Run other sub-systems specific cron jobs */
- if (server.cluster_enabled && !(loops % 10)) clusterCron();
+ run_with_period(1000) {
+ if (server.cluster_enabled) clusterCron();
+ }
+
+ /* Run the sentinel timer if we are in sentinel mode. */
+ run_with_period(100) {
+ if (server.sentinel_mode) sentinelTimer();
+ }
server.cronloops++;
- return 100;
+ return 1000/REDIS_HZ;
}
/* This function gets called every time Redis is entering the
server.aof_rewrite_base_size = 0;
server.aof_rewrite_scheduled = 0;
server.aof_last_fsync = time(NULL);
+ server.aof_rewrite_time_last = -1;
+ server.aof_rewrite_time_start = -1;
server.aof_delayed_fsync = 0;
server.aof_fd = -1;
server.aof_selected_db = -1; /* Make sure the first time will not match */
server.cronloops = 0;
server.rdb_child_pid = -1;
server.aof_child_pid = -1;
- server.aof_rewrite_buf = sdsempty();
+ aofRewriteBufferReset();
server.aof_buf = sdsempty();
server.lastsave = time(NULL);
+ server.rdb_save_time_last = -1;
+ server.rdb_save_time_start = -1;
server.dirty = 0;
server.stat_numcommands = 0;
server.stat_numconnections = 0;
case 's': c->flags |= REDIS_CMD_NOSCRIPT; break;
case 'R': c->flags |= REDIS_CMD_RANDOM; break;
case 'S': c->flags |= REDIS_CMD_SORT_FOR_SCRIPT; break;
+ case 'l': c->flags |= REDIS_CMD_LOADING; break;
+ case 't': c->flags |= REDIS_CMD_STALE; break;
default: redisPanic("Unsupported command flag"); break;
}
f++;
return REDIS_OK;
}
- /* Don't accept wirte commands if this is a read only slave. But
+ /* Don't accept write commands if this is a read only slave. But
* accept write commands if this is our master. */
if (server.masterhost && server.repl_slave_ro &&
!(c->flags & REDIS_MASTER) &&
* we are a slave with a broken link with master. */
if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED &&
server.repl_serve_stale_data == 0 &&
- c->cmd->proc != infoCommand && c->cmd->proc != slaveofCommand)
+ !(c->cmd->flags & REDIS_CMD_STALE))
{
addReply(c, shared.masterdownerr);
return REDIS_OK;
}
- /* Loading DB? Return an error if the command is not INFO */
- if (server.loading && c->cmd->proc != infoCommand) {
+ /* Loading DB? Return an error if the command has not the
+ * REDIS_CMD_LOADING flag. */
+ if (server.loading && !(c->cmd->flags & REDIS_CMD_LOADING)) {
addReply(c, shared.loadingerr);
return REDIS_OK;
}
- /* Lua script too slow? Only allow SHUTDOWN NOSAVE and SCRIPT KILL. */
+ /* Lua script too slow? Only allow commands with REDIS_CMD_STALE flag. */
if (server.lua_timedout &&
!(c->cmd->proc == shutdownCommand &&
c->argc == 2 &&
/*================================== Commands =============================== */
+/* Return zero if strings are the same, non-zero if they are not.
+ * The comparison is performed in a way that prevents an attacker to obtain
+ * information about the nature of the strings just monitoring the execution
+ * time of the function.
+ *
+ * Note that limiting the comparison length to strings up to 512 bytes we
+ * can avoid leaking any information about the password length and any
+ * possible branch misprediction related leak.
+ */
+int time_independent_strcmp(char *a, char *b) {
+ char bufa[REDIS_AUTHPASS_MAX_LEN], bufb[REDIS_AUTHPASS_MAX_LEN];
+ /* The above two strlen perform len(a) + len(b) operations where either
+ * a or b are fixed (our password) length, and the difference is only
+ * relative to the length of the user provided string, so no information
+ * leak is possible in the following two lines of code. */
+ int alen = strlen(a);
+ int blen = strlen(b);
+ int j;
+ int diff = 0;
+
+ /* We can't compare strings longer than our static buffers.
+ * Note that this will never pass the first test in practical circumstances
+ * so there is no info leak. */
+ if (alen > sizeof(bufa) || blen > sizeof(bufb)) return 1;
+
+ memset(bufa,0,sizeof(bufa)); /* Constant time. */
+ memset(bufb,0,sizeof(bufb)); /* Constant time. */
+ /* Again the time of the following two copies is proportional to
+ * len(a) + len(b) so no info is leaked. */
+ memcpy(bufa,a,alen);
+ memcpy(bufb,b,blen);
+
+ /* Always compare all the chars in the two buffers without
+ * conditional expressions. */
+ for (j = 0; j < sizeof(bufa); j++) {
+ diff |= (bufa[j] ^ bufb[j]);
+ }
+ /* Length must be equal as well. */
+ diff |= alen ^ blen;
+ return diff; /* If zero strings are the same. */
+}
+
void authCommand(redisClient *c) {
if (!server.requirepass) {
addReplyError(c,"Client sent AUTH, but no password is set");
- } else if (!strcmp(c->argv[1]->ptr, server.requirepass)) {
+ } else if (!time_independent_strcmp(c->argv[1]->ptr, server.requirepass)) {
c->authenticated = 1;
addReply(c,shared.ok);
} else {
info = sdscatprintf(info,
"# Persistence\r\n"
"loading:%d\r\n"
+ "rdb_changes_since_last_save:%lld\r\n"
+ "rdb_bgsave_in_progress:%d\r\n"
+ "rdb_last_save_time:%ld\r\n"
+ "rdb_last_bgsave_status:%s\r\n"
+ "rdb_last_bgsave_time_sec:%ld\r\n"
+ "rdb_current_bgsave_time_sec:%ld\r\n"
"aof_enabled:%d\r\n"
- "changes_since_last_save:%lld\r\n"
- "bgsave_in_progress:%d\r\n"
- "last_save_time:%ld\r\n"
- "last_bgsave_status:%s\r\n"
- "bgrewriteaof_in_progress:%d\r\n"
- "bgrewriteaof_scheduled:%d\r\n",
+ "aof_rewrite_in_progress:%d\r\n"
+ "aof_rewrite_scheduled:%d\r\n"
+ "aof_last_rewrite_time_sec:%ld\r\n"
+ "aof_current_rewrite_time_sec:%ld\r\n",
server.loading,
- server.aof_state != REDIS_AOF_OFF,
server.dirty,
server.rdb_child_pid != -1,
server.lastsave,
server.lastbgsave_status == REDIS_OK ? "ok" : "err",
+ server.rdb_save_time_last,
+ (server.rdb_child_pid == -1) ?
+ -1 : time(NULL)-server.rdb_save_time_start,
+ server.aof_state != REDIS_AOF_OFF,
server.aof_child_pid != -1,
- server.aof_rewrite_scheduled);
+ server.aof_rewrite_scheduled,
+ server.aof_rewrite_time_last,
+ (server.aof_child_pid == -1) ?
+ -1 : time(NULL)-server.aof_rewrite_time_start);
if (server.aof_state != REDIS_AOF_OFF) {
info = sdscatprintf(info,
"aof_base_size:%lld\r\n"
"aof_pending_rewrite:%d\r\n"
"aof_buffer_length:%zu\r\n"
+ "aof_rewrite_buffer_length:%zu\r\n"
"aof_pending_bio_fsync:%llu\r\n"
"aof_delayed_fsync:%lu\r\n",
(long long) server.aof_current_size,
(long long) server.aof_rewrite_base_size,
server.aof_rewrite_scheduled,
sdslen(server.aof_buf),
+ aofRewriteBufferSize(),
bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC),
server.aof_delayed_fsync);
}
}
if (state == NULL) continue;
info = sdscatprintf(info,"slave%d:%s,%d,%s\r\n",
- slaveid,ip,port,state);
+ slaveid,ip,slave->slave_listening_port,state);
slaveid++;
}
}
}
if (server.aof_state != REDIS_AOF_OFF) {
mem_used -= sdslen(server.aof_buf);
- mem_used -= sdslen(server.aof_rewrite_buf);
+ mem_used -= aofRewriteBufferSize();
}
/* Check if we are over the memory limit. */
fprintf(stderr," ./redis-server /etc/redis/6379.conf\n");
fprintf(stderr," ./redis-server --port 7777\n");
fprintf(stderr," ./redis-server --port 7777 --slaveof 127.0.0.1 8888\n");
- fprintf(stderr," ./redis-server /etc/myredis.conf --loglevel verbose\n");
+ 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");
exit(1);
}
void redisAsciiArt(void) {
#include "asciilogo.h"
char *buf = zmalloc(1024*16);
+ char *mode = "stand alone";
+
+ if (server.cluster_enabled) mode = "cluster";
+ else if (server.sentinel_mode) mode = "sentinel";
snprintf(buf,1024*16,ascii_logo,
REDIS_VERSION,
redisGitSHA1(),
strtol(redisGitDirty(),NULL,10) > 0,
(sizeof(long) == 8) ? "64" : "32",
- server.cluster_enabled ? "cluster" : "stand alone",
- server.port,
+ mode, server.port,
(long) getpid()
);
redisLogRaw(REDIS_NOTICE|REDIS_LOG_RAW,buf);
void memtest(size_t megabytes, int passes);
+/* Returns 1 if there is --sentinel among the arguments or if
+ * argv[0] is exactly "redis-sentinel". */
+int checkForSentinelMode(int argc, char **argv) {
+ int j;
+
+ if (strstr(argv[0],"redis-sentinel") != NULL) return 1;
+ for (j = 1; j < argc; j++)
+ if (!strcmp(argv[j],"--sentinel")) return 1;
+ return 0;
+}
+
+/* Function called at startup to load RDB or AOF file in memory. */
+void loadDataFromDisk(void) {
+ long long start = ustime();
+ if (server.aof_state == REDIS_AOF_ON) {
+ if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
+ redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
+ } else {
+ if (rdbLoad(server.rdb_filename) == REDIS_OK) {
+ redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",
+ (float)(ustime()-start)/1000000);
+ } else if (errno != ENOENT) {
+ redisLog(REDIS_WARNING,"Fatal error loading the DB. Exiting.");
+ exit(1);
+ }
+ }
+}
+
int main(int argc, char **argv) {
- long long start;
struct timeval tv;
/* We need to initialize our libraries, and the server configuration. */
srand(time(NULL)^getpid());
gettimeofday(&tv,NULL);
dictSetHashFunctionSeed(tv.tv_sec^tv.tv_usec^getpid());
+ server.sentinel_mode = checkForSentinelMode(argc,argv);
initServerConfig();
+ /* 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();
+ }
+
if (argc >= 2) {
int j = 1; /* First option to parse in argv[] */
sds options = sdsempty();
initServer();
if (server.daemonize) createPidFile();
redisAsciiArt();
- redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
-#ifdef __linux__
- linuxOvercommitMemoryWarning();
-#endif
- start = ustime();
- if (server.aof_state == REDIS_AOF_ON) {
- if (loadAppendOnlyFile(server.aof_filename) == REDIS_OK)
- redisLog(REDIS_NOTICE,"DB loaded from append only file: %.3f seconds",(float)(ustime()-start)/1000000);
- } else {
- if (rdbLoad(server.rdb_filename) == REDIS_OK) {
- redisLog(REDIS_NOTICE,"DB loaded from disk: %.3f seconds",
- (float)(ustime()-start)/1000000);
- } else if (errno != ENOENT) {
- redisLog(REDIS_WARNING,"Fatal error loading the DB. Exiting.");
- exit(1);
- }
+
+ if (!server.sentinel_mode) {
+ /* Things only needed when not runnign in Sentinel mode. */
+ redisLog(REDIS_WARNING,"Server started, Redis version " REDIS_VERSION);
+ #ifdef __linux__
+ linuxOvercommitMemoryWarning();
+ #endif
+ loadDataFromDisk();
+ if (server.ipfd > 0)
+ redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
+ if (server.sofd > 0)
+ redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
}
- if (server.ipfd > 0)
- redisLog(REDIS_NOTICE,"The server is now ready to accept connections on port %d", server.port);
- if (server.sofd > 0)
- redisLog(REDIS_NOTICE,"The server is now ready to accept connections at %s", server.unixsocket);
+
aeSetBeforeSleepProc(server.el,beforeSleep);
aeMain(server.el);
aeDeleteEventLoop(server.el);
projects / redis.git / blobdiff