replication.c 28.4 KB
Newer Older
1 2 3 4 5 6 7
#include "redis.h"

#include <sys/time.h>
#include <unistd.h>
#include <fcntl.h>
#include <sys/stat.h>

8 9
/* ---------------------------------- MASTER -------------------------------- */

10 11 12
void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) {
    listNode *ln;
    listIter li;
13
    int j;
14 15 16 17 18 19 20 21

    listRewind(slaves,&li);
    while((ln = listNext(&li))) {
        redisClient *slave = ln->value;

        /* Don't feed slaves that are still waiting for BGSAVE to start */
        if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) continue;

22 23 24
        /* Feed slaves that are waiting for the initial SYNC (so these commands
         * are queued in the output buffer until the intial SYNC completes),
         * or are already in sync with the master. */
25 26 27
        if (slave->slaveseldb != dictid) {
            robj *selectcmd;

28 29
            if (dictid >= 0 && dictid < REDIS_SHARED_SELECT_CMDS) {
                selectcmd = shared.select[dictid];
A
antirez 已提交
30
                incrRefCount(selectcmd);
31
            } else {
32 33 34 35
                selectcmd = createObject(REDIS_STRING,
                    sdscatprintf(sdsempty(),"select %d\r\n",dictid));
            }
            addReply(slave,selectcmd);
36
            decrRefCount(selectcmd);
37 38
            slave->slaveseldb = dictid;
        }
39 40
        addReplyMultiBulkLen(slave,argc);
        for (j = 0; j < argc; j++) addReplyBulk(slave,argv[j]);
41 42 43
    }
}

44
void replicationFeedMonitors(redisClient *c, list *monitors, int dictid, robj **argv, int argc) {
45 46
    listNode *ln;
    listIter li;
47
    int j, port;
48 49
    sds cmdrepr = sdsnew("+");
    robj *cmdobj;
50
    char ip[32];
51 52 53
    struct timeval tv;

    gettimeofday(&tv,NULL);
54
    cmdrepr = sdscatprintf(cmdrepr,"%ld.%06ld ",(long)tv.tv_sec,(long)tv.tv_usec);
55 56 57 58 59 60
    if (c->flags & REDIS_LUA_CLIENT) {
        cmdrepr = sdscatprintf(cmdrepr,"[%d lua] ", dictid);
    } else {
        anetPeerToString(c->fd,ip,&port);
        cmdrepr = sdscatprintf(cmdrepr,"[%d %s:%d] ", dictid,ip,port);
    }
61 62 63

    for (j = 0; j < argc; j++) {
        if (argv[j]->encoding == REDIS_ENCODING_INT) {
64
            cmdrepr = sdscatprintf(cmdrepr, "\"%ld\"", (long)argv[j]->ptr);
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
        } else {
            cmdrepr = sdscatrepr(cmdrepr,(char*)argv[j]->ptr,
                        sdslen(argv[j]->ptr));
        }
        if (j != argc-1)
            cmdrepr = sdscatlen(cmdrepr," ",1);
    }
    cmdrepr = sdscatlen(cmdrepr,"\r\n",2);
    cmdobj = createObject(REDIS_STRING,cmdrepr);

    listRewind(monitors,&li);
    while((ln = listNext(&li))) {
        redisClient *monitor = ln->value;
        addReply(monitor,cmdobj);
    }
    decrRefCount(cmdobj);
}

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

87 88
    /* Refuse SYNC requests if we are a slave but the link with our master
     * is not ok... */
A
antirez 已提交
89
    if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED) {
90
        addReplyError(c,"Can't SYNC while not connected with my master");
91 92 93
        return;
    }

94 95 96 97 98
    /* SYNC can't be issued when the server has pending data to send to
     * the client about already issued commands. We need a fresh reply
     * buffer registering the differences between the BGSAVE and the current
     * dataset, so that we can copy to other slaves if needed. */
    if (listLength(c->reply) != 0) {
99
        addReplyError(c,"SYNC is invalid with pending input");
100 101 102 103 104 105
        return;
    }

    redisLog(REDIS_NOTICE,"Slave ask for synchronization");
    /* Here we need to check if there is a background saving operation
     * in progress, or if it is required to start one */
A
antirez 已提交
106
    if (server.rdb_child_pid != -1) {
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
        /* Ok a background save is in progress. Let's check if it is a good
         * one for replication, i.e. if there is another slave that is
         * registering differences since the server forked to save */
        redisClient *slave;
        listNode *ln;
        listIter li;

        listRewind(server.slaves,&li);
        while((ln = listNext(&li))) {
            slave = ln->value;
            if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_END) break;
        }
        if (ln) {
            /* Perfect, the server is already registering differences for
             * another slave. Set the right state, and copy the buffer. */
122
            copyClientOutputBuffer(c,slave);
123 124 125 126 127 128 129 130 131 132 133
            c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
            redisLog(REDIS_NOTICE,"Waiting for end of BGSAVE for SYNC");
        } else {
            /* No way, we need to wait for the next BGSAVE in order to
             * register differences */
            c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
            redisLog(REDIS_NOTICE,"Waiting for next BGSAVE for SYNC");
        }
    } else {
        /* Ok we don't have a BGSAVE in progress, let's start one */
        redisLog(REDIS_NOTICE,"Starting BGSAVE for SYNC");
A
antirez 已提交
134
        if (rdbSaveBackground(server.rdb_filename) != REDIS_OK) {
135
            redisLog(REDIS_NOTICE,"Replication failed, can't BGSAVE");
136
            addReplyError(c,"Unable to perform background save");
137 138 139 140 141 142 143 144 145 146 147
            return;
        }
        c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
    }
    c->repldbfd = -1;
    c->flags |= REDIS_SLAVE;
    c->slaveseldb = 0;
    listAddNodeTail(server.slaves,c);
    return;
}

A
antirez 已提交
148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
/* REPLCONF <option> <value> <option> <value> ...
 * This command is used by a slave in order to configure the replication
 * process before starting it with the SYNC command.
 *
 * Currently the only use of this command is to communicate to the master
 * what is the listening port of the Slave redis instance, so that the
 * master can accurately list slaves and their listening ports in
 * the INFO output.
 *
 * In the future the same command can be used in order to configure
 * the replication to initiate an incremental replication instead of a
 * full resync. */
void replconfCommand(redisClient *c) {
    int j;

    if ((c->argc % 2) == 0) {
        /* Number of arguments must be odd to make sure that every
         * option has a corresponding value. */
        addReply(c,shared.syntaxerr);
        return;
    }

    /* Process every option-value pair. */
    for (j = 1; j < c->argc; j+=2) {
        if (!strcasecmp(c->argv[j]->ptr,"listening-port")) {
            long port;

            if ((getLongFromObjectOrReply(c,c->argv[j+1],
                    &port,NULL) != REDIS_OK))
                return;
            c->slave_listening_port = port;
        } else {
            addReplyErrorFormat(c,"Unrecognized REPLCONF option: %s",
                (char*)c->argv[j]->ptr);
            return;
        }
    }
    addReply(c,shared.ok);
}

188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
void sendBulkToSlave(aeEventLoop *el, int fd, void *privdata, int mask) {
    redisClient *slave = privdata;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);
    char buf[REDIS_IOBUF_LEN];
    ssize_t nwritten, buflen;

    if (slave->repldboff == 0) {
        /* Write the bulk write count before to transfer the DB. In theory here
         * we don't know how much room there is in the output buffer of the
         * socket, but in pratice SO_SNDLOWAT (the minimum count for output
         * operations) will never be smaller than the few bytes we need. */
        sds bulkcount;

        bulkcount = sdscatprintf(sdsempty(),"$%lld\r\n",(unsigned long long)
            slave->repldbsize);
        if (write(fd,bulkcount,sdslen(bulkcount)) != (signed)sdslen(bulkcount))
        {
            sdsfree(bulkcount);
            freeClient(slave);
            return;
        }
        sdsfree(bulkcount);
    }
    lseek(slave->repldbfd,slave->repldboff,SEEK_SET);
    buflen = read(slave->repldbfd,buf,REDIS_IOBUF_LEN);
    if (buflen <= 0) {
        redisLog(REDIS_WARNING,"Read error sending DB to slave: %s",
            (buflen == 0) ? "premature EOF" : strerror(errno));
        freeClient(slave);
        return;
    }
    if ((nwritten = write(fd,buf,buflen)) == -1) {
        redisLog(REDIS_VERBOSE,"Write error sending DB to slave: %s",
            strerror(errno));
        freeClient(slave);
        return;
    }
    slave->repldboff += nwritten;
    if (slave->repldboff == slave->repldbsize) {
        close(slave->repldbfd);
        slave->repldbfd = -1;
        aeDeleteFileEvent(server.el,slave->fd,AE_WRITABLE);
        slave->replstate = REDIS_REPL_ONLINE;
        if (aeCreateFileEvent(server.el, slave->fd, AE_WRITABLE,
            sendReplyToClient, slave) == AE_ERR) {
            freeClient(slave);
            return;
        }
        redisLog(REDIS_NOTICE,"Synchronization with slave succeeded");
    }
}

/* This function is called at the end of every backgrond saving.
 * The argument bgsaveerr is REDIS_OK if the background saving succeeded
 * otherwise REDIS_ERR is passed to the function.
 *
 * The goal of this function is to handle slaves waiting for a successful
 * background saving in order to perform non-blocking synchronization. */
void updateSlavesWaitingBgsave(int bgsaveerr) {
    listNode *ln;
    int startbgsave = 0;
    listIter li;

    listRewind(server.slaves,&li);
    while((ln = listNext(&li))) {
        redisClient *slave = ln->value;

        if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) {
            startbgsave = 1;
            slave->replstate = REDIS_REPL_WAIT_BGSAVE_END;
        } else if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_END) {
            struct redis_stat buf;

            if (bgsaveerr != REDIS_OK) {
                freeClient(slave);
                redisLog(REDIS_WARNING,"SYNC failed. BGSAVE child returned an error");
                continue;
            }
A
antirez 已提交
267
            if ((slave->repldbfd = open(server.rdb_filename,O_RDONLY)) == -1 ||
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283
                redis_fstat(slave->repldbfd,&buf) == -1) {
                freeClient(slave);
                redisLog(REDIS_WARNING,"SYNC failed. Can't open/stat DB after BGSAVE: %s", strerror(errno));
                continue;
            }
            slave->repldboff = 0;
            slave->repldbsize = buf.st_size;
            slave->replstate = REDIS_REPL_SEND_BULK;
            aeDeleteFileEvent(server.el,slave->fd,AE_WRITABLE);
            if (aeCreateFileEvent(server.el, slave->fd, AE_WRITABLE, sendBulkToSlave, slave) == AE_ERR) {
                freeClient(slave);
                continue;
            }
        }
    }
    if (startbgsave) {
A
antirez 已提交
284
        if (rdbSaveBackground(server.rdb_filename) != REDIS_OK) {
285 286 287 288 289 290 291 292 293 294 295 296 297 298
            listIter li;

            listRewind(server.slaves,&li);
            redisLog(REDIS_WARNING,"SYNC failed. BGSAVE failed");
            while((ln = listNext(&li))) {
                redisClient *slave = ln->value;

                if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START)
                    freeClient(slave);
            }
        }
    }
}

299 300 301 302
/* ----------------------------------- SLAVE -------------------------------- */

/* Abort the async download of the bulk dataset while SYNC-ing with master */
void replicationAbortSyncTransfer(void) {
A
antirez 已提交
303
    redisAssert(server.repl_state == REDIS_REPL_TRANSFER);
304 305 306 307 308 309

    aeDeleteFileEvent(server.el,server.repl_transfer_s,AE_READABLE);
    close(server.repl_transfer_s);
    close(server.repl_transfer_fd);
    unlink(server.repl_transfer_tmpfile);
    zfree(server.repl_transfer_tmpfile);
A
antirez 已提交
310
    server.repl_state = REDIS_REPL_CONNECT;
311 312 313
}

/* Asynchronously read the SYNC payload we receive from a master */
314
#define REPL_MAX_WRITTEN_BEFORE_FSYNC (1024*1024*8) /* 8 MB */
315
void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) {
316
    char buf[4096];
317
    ssize_t nread, readlen;
318
    off_t left;
319 320 321
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(privdata);
    REDIS_NOTUSED(mask);
322

323
    /* If repl_transfer_size == -1 we still have to read the bulk length
324
     * from the master reply. */
325
    if (server.repl_transfer_size == -1) {
326
        if (syncReadLine(fd,buf,1024,server.repl_syncio_timeout*1000) == -1) {
327 328 329
            redisLog(REDIS_WARNING,
                "I/O error reading bulk count from MASTER: %s",
                strerror(errno));
330
            goto error;
331
        }
332

333 334 335 336
        if (buf[0] == '-') {
            redisLog(REDIS_WARNING,
                "MASTER aborted replication with an error: %s",
                buf+1);
337
            goto error;
338 339 340 341
        } else if (buf[0] == '\0') {
            /* At this stage just a newline works as a PING in order to take
             * the connection live. So we refresh our last interaction
             * timestamp. */
342
            server.repl_transfer_lastio = server.unixtime;
343
            return;
344 345
        } else if (buf[0] != '$') {
            redisLog(REDIS_WARNING,"Bad protocol from MASTER, the first byte is not '$', are you sure the host and port are right?");
346
            goto error;
347
        }
348
        server.repl_transfer_size = strtol(buf+1,NULL,10);
349 350
        redisLog(REDIS_NOTICE,
            "MASTER <-> SLAVE sync: receiving %ld bytes from master",
351
            server.repl_transfer_size);
352 353 354 355
        return;
    }

    /* Read bulk data */
356 357
    left = server.repl_transfer_size - server.repl_transfer_read;
    readlen = (left < (signed)sizeof(buf)) ? left : (signed)sizeof(buf);
358 359 360 361 362 363 364
    nread = read(fd,buf,readlen);
    if (nread <= 0) {
        redisLog(REDIS_WARNING,"I/O error trying to sync with MASTER: %s",
            (nread == -1) ? strerror(errno) : "connection lost");
        replicationAbortSyncTransfer();
        return;
    }
365
    server.repl_transfer_lastio = server.unixtime;
366
    if (write(server.repl_transfer_fd,buf,nread) != nread) {
367
        redisLog(REDIS_WARNING,"Write error or short write writing to the DB dump file needed for MASTER <-> SLAVE synchronization: %s", strerror(errno));
368
        goto error;
369
    }
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
    server.repl_transfer_read += nread;

    /* Sync data on disk from time to time, otherwise at the end of the transfer
     * we may suffer a big delay as the memory buffers are copied into the
     * actual disk. */
    if (server.repl_transfer_read >=
        server.repl_transfer_last_fsync_off + REPL_MAX_WRITTEN_BEFORE_FSYNC)
    {
        off_t sync_size = server.repl_transfer_read -
                          server.repl_transfer_last_fsync_off;
        rdb_fsync_range(server.repl_transfer_fd,
            server.repl_transfer_last_fsync_off, sync_size);
        server.repl_transfer_last_fsync_off += sync_size;
    }

385
    /* Check if the transfer is now complete */
386
    if (server.repl_transfer_read == server.repl_transfer_size) {
A
antirez 已提交
387
        if (rename(server.repl_transfer_tmpfile,server.rdb_filename) == -1) {
388 389 390 391
            redisLog(REDIS_WARNING,"Failed trying to rename the temp DB into dump.rdb in MASTER <-> SLAVE synchronization: %s", strerror(errno));
            replicationAbortSyncTransfer();
            return;
        }
A
antirez 已提交
392
        redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Loading DB in memory");
393
        emptyDb();
394 395 396 397 398
        /* Before loading the DB into memory we need to delete the readable
         * handler, otherwise it will get called recursively since
         * rdbLoad() will call the event loop to process events from time to
         * time for non blocking loading. */
        aeDeleteFileEvent(server.el,server.repl_transfer_s,AE_READABLE);
A
antirez 已提交
399
        if (rdbLoad(server.rdb_filename) != REDIS_OK) {
400 401 402 403 404 405 406 407 408 409
            redisLog(REDIS_WARNING,"Failed trying to load the MASTER synchronization DB from disk");
            replicationAbortSyncTransfer();
            return;
        }
        /* Final setup of the connected slave <- master link */
        zfree(server.repl_transfer_tmpfile);
        close(server.repl_transfer_fd);
        server.master = createClient(server.repl_transfer_s);
        server.master->flags |= REDIS_MASTER;
        server.master->authenticated = 1;
A
antirez 已提交
410
        server.repl_state = REDIS_REPL_CONNECTED;
411
        redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Finished with success");
412 413 414
        /* Restart the AOF subsystem now that we finished the sync. This
         * will trigger an AOF rewrite, and when done will start appending
         * to the new file. */
415
        if (server.aof_state != REDIS_AOF_OFF) {
416 417 418 419 420 421 422 423 424 425 426 427
            int retry = 10;

            stopAppendOnly();
            while (retry-- && startAppendOnly() == REDIS_ERR) {
                redisLog(REDIS_WARNING,"Failed enabling the AOF after successful master synchrnization! Trying it again in one second.");
                sleep(1);
            }
            if (!retry) {
                redisLog(REDIS_WARNING,"FATAL: this slave instance finished the synchronization with its master, but the AOF can't be turned on. Exiting now.");
                exit(1);
            }
        }
428
    }
429 430 431 432 433 434

    return;

error:
    replicationAbortSyncTransfer();
    return;
435 436
}

A
antirez 已提交
437
/* Send a synchronous command to the master. Used to send AUTH and
A
antirez 已提交
438
 * REPLCONF commands before starting the replication with SYNC.
A
antirez 已提交
439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482
 *
 * On success NULL is returned.
 * On error an sds string describing the error is returned.
 */
char *sendSynchronousCommand(int fd, ...) {
    va_list ap;
    sds cmd = sdsempty();
    char *arg, buf[256];

    /* Create the command to send to the master, we use simple inline
     * protocol for simplicity as currently we only send simple strings. */
    va_start(ap,fd);
    while(1) {
        arg = va_arg(ap, char*);
        if (arg == NULL) break;

        if (sdslen(cmd) != 0) cmd = sdscatlen(cmd," ",1);
        cmd = sdscat(cmd,arg);
    }
    cmd = sdscatlen(cmd,"\r\n",2);

    /* Transfer command to the server. */
    if (syncWrite(fd,cmd,sdslen(cmd),server.repl_syncio_timeout*1000) == -1) {
        sdsfree(cmd);
        return sdscatprintf(sdsempty(),"Writing to master: %s",
                strerror(errno));
    }
    sdsfree(cmd);

    /* Read the reply from the server. */
    if (syncReadLine(fd,buf,sizeof(buf),server.repl_syncio_timeout*1000) == -1)
    {
        return sdscatprintf(sdsempty(),"Reading from master: %s",
                strerror(errno));
    }

    /* Check for errors from the server. */
    if (buf[0] != '+') {
        return sdscatprintf(sdsempty(),"Error from master: %s", buf);
    }

    return NULL; /* No errors. */
}

483
void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
A
antirez 已提交
484
    char tmpfile[256], *err;
485
    int dfd, maxtries = 5;
486 487
    int sockerr = 0;
    socklen_t errlen = sizeof(sockerr);
488 489 490
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(privdata);
    REDIS_NOTUSED(mask);
491

492 493
    /* If this event fired after the user turned the instance into a master
     * with SLAVEOF NO ONE we must just return ASAP. */
A
antirez 已提交
494
    if (server.repl_state == REDIS_REPL_NONE) {
495 496 497 498
        close(fd);
        return;
    }

499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
    /* Check for errors in the socket. */
    if (getsockopt(fd, SOL_SOCKET, SO_ERROR, &sockerr, &errlen) == -1)
        sockerr = errno;
    if (sockerr) {
        aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
        redisLog(REDIS_WARNING,"Error condition on socket for SYNC: %s",
            strerror(sockerr));
        goto error;
    }

    /* If we were connecting, it's time to send a non blocking PING, we want to
     * make sure the master is able to reply before going into the actual
     * replication process where we have long timeouts in the order of
     * seconds (in the meantime the slave would block). */
    if (server.repl_state == REDIS_REPL_CONNECTING) {
        redisLog(REDIS_NOTICE,"Non blocking connect for SYNC fired the event.");
        /* Delete the writable event so that the readable event remains
         * registered and we can wait for the PONG reply. */
        aeDeleteFileEvent(server.el,fd,AE_WRITABLE);
        server.repl_state = REDIS_REPL_RECEIVE_PONG;
        /* Send the PING, don't check for errors at all, we have the timeout
         * that will take care about this. */
        syncWrite(fd,"PING\r\n",6,100);
        return;
    }

    /* Receive the PONG command. */
    if (server.repl_state == REDIS_REPL_RECEIVE_PONG) {
        char buf[1024];

        /* Delete the readable event, we no longer need it now that there is
         * the PING reply to read. */
        aeDeleteFileEvent(server.el,fd,AE_READABLE);

        /* Read the reply with explicit timeout. */
        buf[0] = '\0';
        if (syncReadLine(fd,buf,sizeof(buf),
            server.repl_syncio_timeout*1000) == -1)
        {
            redisLog(REDIS_WARNING,
                "I/O error reading PING reply from master: %s",
                strerror(errno));
            goto error;
        }

        /* We don't care about the reply, it can be +PONG or an error since
         * the server requires AUTH. As long as it replies correctly, it's
         * fine from our point of view. */
        if (buf[0] != '-' && buf[0] != '+') {
            redisLog(REDIS_WARNING,"Unexpected reply to PING from master.");
            goto error;
        } else {
            redisLog(REDIS_NOTICE,
                "Master replied to PING, replication can continue...");
        }
    }
555 556 557

    /* AUTH with the master if required. */
    if(server.masterauth) {
A
antirez 已提交
558 559 560 561
        err = sendSynchronousCommand(fd,"AUTH",server.masterauth,NULL);
        if (err) {
            redisLog(REDIS_WARNING,"Unable to AUTH to MASTER: %s",err);
            sdsfree(err);
562 563
            goto error;
        }
A
antirez 已提交
564 565 566 567 568 569 570 571 572 573 574 575 576 577
    }

    /* Set the slave port, so that Master's INFO command can list the
     * slave listening port correctly. */
    {
        sds port = sdsfromlonglong(server.port);
        err = sendSynchronousCommand(fd,"REPLCONF","listening-port",port,
                                         NULL);
        sdsfree(port);
        /* Ignore the error if any, not all the Redis versions support
         * REPLCONF listening-port. */
        if (err) {
            redisLog(REDIS_NOTICE,"(non critical): Master does not understand REPLCONF listening-port: %s", err);
            sdsfree(err);
578 579 580 581
        }
    }

    /* Issue the SYNC command */
582
    if (syncWrite(fd,"SYNC\r\n",6,server.repl_syncio_timeout*1000) == -1) {
583 584
        redisLog(REDIS_WARNING,"I/O error writing to MASTER: %s",
            strerror(errno));
585
        goto error;
586
    }
587 588

    /* Prepare a suitable temp file for bulk transfer */
589 590
    while(maxtries--) {
        snprintf(tmpfile,256,
591
            "temp-%d.%ld.rdb",(int)server.unixtime,(long int)getpid());
592 593 594 595 596 597
        dfd = open(tmpfile,O_CREAT|O_WRONLY|O_EXCL,0644);
        if (dfd != -1) break;
        sleep(1);
    }
    if (dfd == -1) {
        redisLog(REDIS_WARNING,"Opening the temp file needed for MASTER <-> SLAVE synchronization: %s",strerror(errno));
598
        goto error;
599 600
    }

601
    /* Setup the non blocking download of the bulk file. */
602
    if (aeCreateFileEvent(server.el,fd, AE_READABLE,readSyncBulkPayload,NULL)
603
            == AE_ERR)
604 605
    {
        redisLog(REDIS_WARNING,"Can't create readable event for SYNC");
606
        goto error;
607
    }
608

A
antirez 已提交
609
    server.repl_state = REDIS_REPL_TRANSFER;
610 611 612
    server.repl_transfer_size = -1;
    server.repl_transfer_read = 0;
    server.repl_transfer_last_fsync_off = 0;
613
    server.repl_transfer_fd = dfd;
614
    server.repl_transfer_lastio = server.unixtime;
615
    server.repl_transfer_tmpfile = zstrdup(tmpfile);
616 617 618 619
    return;

error:
    close(fd);
620 621
    server.repl_transfer_s = -1;
    server.repl_state = REDIS_REPL_CONNECT;
622 623 624 625 626 627 628 629 630 631 632 633 634
    return;
}

int connectWithMaster(void) {
    int fd;

    fd = anetTcpNonBlockConnect(NULL,server.masterhost,server.masterport);
    if (fd == -1) {
        redisLog(REDIS_WARNING,"Unable to connect to MASTER: %s",
            strerror(errno));
        return REDIS_ERR;
    }

635
    if (aeCreateFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE,syncWithMaster,NULL) ==
636
            AE_ERR)
637 638 639 640 641 642
    {
        close(fd);
        redisLog(REDIS_WARNING,"Can't create readable event for SYNC");
        return REDIS_ERR;
    }

643
    server.repl_transfer_lastio = server.unixtime;
644
    server.repl_transfer_s = fd;
A
antirez 已提交
645
    server.repl_state = REDIS_REPL_CONNECTING;
646 647 648
    return REDIS_OK;
}

649 650 651 652 653
/* This function can be called when a non blocking connection is currently
 * in progress to undo it. */
void undoConnectWithMaster(void) {
    int fd = server.repl_transfer_s;

654 655
    redisAssert(server.repl_state == REDIS_REPL_CONNECTING ||
                server.repl_state == REDIS_REPL_RECEIVE_PONG);
656 657 658
    aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
    close(fd);
    server.repl_transfer_s = -1;
A
antirez 已提交
659
    server.repl_state = REDIS_REPL_CONNECT;
660 661
}

662 663 664 665 666 667 668
void slaveofCommand(redisClient *c) {
    if (!strcasecmp(c->argv[1]->ptr,"no") &&
        !strcasecmp(c->argv[2]->ptr,"one")) {
        if (server.masterhost) {
            sdsfree(server.masterhost);
            server.masterhost = NULL;
            if (server.master) freeClient(server.master);
A
antirez 已提交
669
            if (server.repl_state == REDIS_REPL_TRANSFER)
670
                replicationAbortSyncTransfer();
671 672
            else if (server.repl_state == REDIS_REPL_CONNECTING ||
                     server.repl_state == REDIS_REPL_RECEIVE_PONG)
673
                undoConnectWithMaster();
A
antirez 已提交
674
            server.repl_state = REDIS_REPL_NONE;
675 676 677
            redisLog(REDIS_NOTICE,"MASTER MODE enabled (user request)");
        }
    } else {
678 679 680 681 682 683 684 685 686 687 688 689 690 691
        long port;

        if ((getLongFromObjectOrReply(c, c->argv[2], &port, NULL) != REDIS_OK))
            return;

        /* Check if we are already attached to the specified slave */
        if (server.masterhost && !strcasecmp(server.masterhost,c->argv[1]->ptr)
            && server.masterport == port) {
            redisLog(REDIS_NOTICE,"SLAVE OF would result into synchronization with the master we are already connected with. No operation performed.");
            addReplySds(c,sdsnew("+OK Already connected to specified master\r\n"));
            return;
        }
        /* There was no previous master or the user specified a different one,
         * we can continue. */
692 693
        sdsfree(server.masterhost);
        server.masterhost = sdsdup(c->argv[1]->ptr);
694
        server.masterport = port;
695
        if (server.master) freeClient(server.master);
696
        disconnectSlaves(); /* Force our slaves to resync with us as well. */
A
antirez 已提交
697
        if (server.repl_state == REDIS_REPL_TRANSFER)
698
            replicationAbortSyncTransfer();
A
antirez 已提交
699
        server.repl_state = REDIS_REPL_CONNECT;
700 701 702 703 704
        redisLog(REDIS_NOTICE,"SLAVE OF %s:%d enabled (user request)",
            server.masterhost, server.masterport);
    }
    addReply(c,shared.ok);
}
705 706 707 708

/* --------------------------- REPLICATION CRON  ---------------------------- */

void replicationCron(void) {
709
    /* Non blocking connection timeout? */
710 711 712
    if (server.masterhost &&
        (server.repl_state == REDIS_REPL_CONNECTING ||
         server.repl_state == REDIS_REPL_RECEIVE_PONG) &&
713 714 715 716 717 718
        (time(NULL)-server.repl_transfer_lastio) > server.repl_timeout)
    {
        redisLog(REDIS_WARNING,"Timeout connecting to the MASTER...");
        undoConnectWithMaster();
    }

719
    /* Bulk transfer I/O timeout? */
A
antirez 已提交
720
    if (server.masterhost && server.repl_state == REDIS_REPL_TRANSFER &&
A
7c6da73  
antirez 已提交
721
        (time(NULL)-server.repl_transfer_lastio) > server.repl_timeout)
722 723 724 725 726
    {
        redisLog(REDIS_WARNING,"Timeout receiving bulk data from MASTER...");
        replicationAbortSyncTransfer();
    }

727
    /* Timed out master when we are an already connected slave? */
A
antirez 已提交
728
    if (server.masterhost && server.repl_state == REDIS_REPL_CONNECTED &&
A
7c6da73  
antirez 已提交
729
        (time(NULL)-server.master->lastinteraction) > server.repl_timeout)
730 731 732 733 734
    {
        redisLog(REDIS_WARNING,"MASTER time out: no data nor PING received...");
        freeClient(server.master);
    }

735
    /* Check if we should connect to a MASTER */
A
antirez 已提交
736
    if (server.repl_state == REDIS_REPL_CONNECT) {
737
        redisLog(REDIS_NOTICE,"Connecting to MASTER...");
738 739
        if (connectWithMaster() == REDIS_OK) {
            redisLog(REDIS_NOTICE,"MASTER <-> SLAVE sync started");
740 741
        }
    }
742 743 744 745 746
    
    /* If we have attached slaves, PING them from time to time.
     * So slaves can implement an explicit timeout to masters, and will
     * be able to detect a link disconnection even if the TCP connection
     * will not actually go down. */
747
    if (!(server.cronloops % (server.repl_ping_slave_period * REDIS_HZ))) {
748 749 750 751 752 753 754 755 756 757 758 759
        listIter li;
        listNode *ln;

        listRewind(server.slaves,&li);
        while((ln = listNext(&li))) {
            redisClient *slave = ln->value;

            /* Don't ping slaves that are in the middle of a bulk transfer
             * with the master for first synchronization. */
            if (slave->replstate == REDIS_REPL_SEND_BULK) continue;
            if (slave->replstate == REDIS_REPL_ONLINE) {
                /* If the slave is online send a normal ping */
760
                addReplySds(slave,sdsnew("*1\r\n$4\r\nPING\r\n"));
761 762 763 764 765 766
            } else {
                /* Otherwise we are in the pre-synchronization stage.
                 * Just a newline will do the work of refreshing the
                 * connection last interaction time, and at the same time
                 * we'll be sure that being a single char there are no
                 * short-write problems. */
A
antirez 已提交
767 768 769
                if (write(slave->fd, "\n", 1) == -1) {
                    /* Don't worry, it's just a ping. */
                }
770 771 772
            }
        }
    }
773
}