replication.c 30.9 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
/* Asynchronous replication implementation.
 *
 * Copyright (c) 2009-2012, Salvatore Sanfilippo <antirez at gmail dot com>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *   * Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above copyright
 *     notice, this list of conditions and the following disclaimer in the
 *     documentation and/or other materials provided with the distribution.
 *   * Neither the name of Redis nor the names of its contributors may be used
 *     to endorse or promote products derived from this software without
 *     specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */


32 33 34 35 36
#include "redis.h"

#include <sys/time.h>
#include <unistd.h>
#include <fcntl.h>
37
#include <sys/socket.h>
38 39
#include <sys/stat.h>

40 41
/* ---------------------------------- MASTER -------------------------------- */

42 43 44
void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) {
    listNode *ln;
    listIter li;
45
    int j;
46 47 48 49 50 51 52 53

    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;

54
        /* Feed slaves that are waiting for the initial SYNC (so these commands
G
guiquanz 已提交
55
         * are queued in the output buffer until the initial SYNC completes),
56
         * or are already in sync with the master. */
57 58 59
        if (slave->slaveseldb != dictid) {
            robj *selectcmd;

60 61
            if (dictid >= 0 && dictid < REDIS_SHARED_SELECT_CMDS) {
                selectcmd = shared.select[dictid];
A
antirez 已提交
62
                incrRefCount(selectcmd);
63
            } else {
64 65 66 67
                selectcmd = createObject(REDIS_STRING,
                    sdscatprintf(sdsempty(),"select %d\r\n",dictid));
            }
            addReply(slave,selectcmd);
68
            decrRefCount(selectcmd);
69 70
            slave->slaveseldb = dictid;
        }
71 72
        addReplyMultiBulkLen(slave,argc);
        for (j = 0; j < argc; j++) addReplyBulk(slave,argv[j]);
73 74 75
    }
}

76
void replicationFeedMonitors(redisClient *c, list *monitors, int dictid, robj **argv, int argc) {
77 78
    listNode *ln;
    listIter li;
79
    int j, port;
80 81
    sds cmdrepr = sdsnew("+");
    robj *cmdobj;
82
    char ip[32];
83 84 85
    struct timeval tv;

    gettimeofday(&tv,NULL);
86
    cmdrepr = sdscatprintf(cmdrepr,"%ld.%06ld ",(long)tv.tv_sec,(long)tv.tv_usec);
87
    if (c->flags & REDIS_LUA_CLIENT) {
88 89 90
        cmdrepr = sdscatprintf(cmdrepr,"[%d lua] ",dictid);
    } else if (c->flags & REDIS_UNIX_SOCKET) {
        cmdrepr = sdscatprintf(cmdrepr,"[%d unix:%s] ",dictid,server.unixsocket);
91 92
    } else {
        anetPeerToString(c->fd,ip,&port);
93
        cmdrepr = sdscatprintf(cmdrepr,"[%d %s:%d] ",dictid,ip,port);
94
    }
95 96 97

    for (j = 0; j < argc; j++) {
        if (argv[j]->encoding == REDIS_ENCODING_INT) {
98
            cmdrepr = sdscatprintf(cmdrepr, "\"%ld\"", (long)argv[j]->ptr);
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
        } 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) {
G
guiquanz 已提交
118
    /* ignore SYNC if already slave or in monitor mode */
119 120
    if (c->flags & REDIS_SLAVE) return;

121 122
    /* Refuse SYNC requests if we are a slave but the link with our master
     * is not ok... */
A
antirez 已提交
123
    if (server.masterhost && server.repl_state != REDIS_REPL_CONNECTED) {
124
        addReplyError(c,"Can't SYNC while not connected with my master");
125 126 127
        return;
    }

128 129 130 131 132
    /* 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) {
133
        addReplyError(c,"SYNC is invalid with pending input");
134 135 136 137 138 139
        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
140
    if (server.rdb_child_pid != -1) {
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
        /* 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. */
156
            copyClientOutputBuffer(c,slave);
157 158 159 160 161 162 163 164 165 166 167
            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
168
        if (rdbSaveBackground(server.rdb_filename) != REDIS_OK) {
169
            redisLog(REDIS_NOTICE,"Replication failed, can't BGSAVE");
170
            addReplyError(c,"Unable to perform background save");
171 172 173 174
            return;
        }
        c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
    }
175 176 177

    if (server.repl_disable_tcp_nodelay)
        anetDisableTcpNoDelay(NULL, c->fd); /* Non critical if it fails. */
178 179 180 181 182 183 184
    c->repldbfd = -1;
    c->flags |= REDIS_SLAVE;
    c->slaveseldb = 0;
    listAddNodeTail(server.slaves,c);
    return;
}

A
antirez 已提交
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
/* 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);
}

225 226 227 228 229 230 231 232 233 234
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
G
guiquanz 已提交
235
         * socket, but in practice SO_SNDLOWAT (the minimum count for output
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
         * 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");
    }
}

G
guiquanz 已提交
278
/* This function is called at the end of every background saving.
279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
 * 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
304
            if ((slave->repldbfd = open(server.rdb_filename,O_RDONLY)) == -1 ||
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
                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
321
        if (rdbSaveBackground(server.rdb_filename) != REDIS_OK) {
322 323 324 325 326 327 328 329 330 331 332 333 334 335
            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);
            }
        }
    }
}

336 337 338 339
/* ----------------------------------- SLAVE -------------------------------- */

/* Abort the async download of the bulk dataset while SYNC-ing with master */
void replicationAbortSyncTransfer(void) {
A
antirez 已提交
340
    redisAssert(server.repl_state == REDIS_REPL_TRANSFER);
341 342 343 344 345 346

    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
347
    server.repl_state = REDIS_REPL_CONNECT;
348 349 350
}

/* Asynchronously read the SYNC payload we receive from a master */
351
#define REPL_MAX_WRITTEN_BEFORE_FSYNC (1024*1024*8) /* 8 MB */
352
void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) {
353
    char buf[4096];
354
    ssize_t nread, readlen;
355
    off_t left;
356 357 358
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(privdata);
    REDIS_NOTUSED(mask);
359

360
    /* If repl_transfer_size == -1 we still have to read the bulk length
361
     * from the master reply. */
362
    if (server.repl_transfer_size == -1) {
363
        if (syncReadLine(fd,buf,1024,server.repl_syncio_timeout*1000) == -1) {
364 365 366
            redisLog(REDIS_WARNING,
                "I/O error reading bulk count from MASTER: %s",
                strerror(errno));
367
            goto error;
368
        }
369

370 371 372 373
        if (buf[0] == '-') {
            redisLog(REDIS_WARNING,
                "MASTER aborted replication with an error: %s",
                buf+1);
374
            goto error;
375 376 377 378
        } 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. */
379
            server.repl_transfer_lastio = server.unixtime;
380
            return;
381 382
        } 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?");
383
            goto error;
384
        }
385
        server.repl_transfer_size = strtol(buf+1,NULL,10);
386 387
        redisLog(REDIS_NOTICE,
            "MASTER <-> SLAVE sync: receiving %ld bytes from master",
388
            server.repl_transfer_size);
389 390 391 392
        return;
    }

    /* Read bulk data */
393 394
    left = server.repl_transfer_size - server.repl_transfer_read;
    readlen = (left < (signed)sizeof(buf)) ? left : (signed)sizeof(buf);
395 396 397 398 399 400 401
    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;
    }
402
    server.repl_transfer_lastio = server.unixtime;
403
    if (write(server.repl_transfer_fd,buf,nread) != nread) {
404
        redisLog(REDIS_WARNING,"Write error or short write writing to the DB dump file needed for MASTER <-> SLAVE synchronization: %s", strerror(errno));
405
        goto error;
406
    }
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
    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;
    }

422
    /* Check if the transfer is now complete */
423
    if (server.repl_transfer_read == server.repl_transfer_size) {
A
antirez 已提交
424
        if (rename(server.repl_transfer_tmpfile,server.rdb_filename) == -1) {
425 426 427 428
            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
429
        redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Loading DB in memory");
430
        emptyDb();
431 432 433 434 435
        /* 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
436
        if (rdbLoad(server.rdb_filename) != REDIS_OK) {
437 438 439 440 441 442 443 444 445 446
            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
447
        server.repl_state = REDIS_REPL_CONNECTED;
448
        redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Finished with success");
449 450 451
        /* 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. */
452
        if (server.aof_state != REDIS_AOF_OFF) {
453 454 455 456
            int retry = 10;

            stopAppendOnly();
            while (retry-- && startAppendOnly() == REDIS_ERR) {
G
guiquanz 已提交
457
                redisLog(REDIS_WARNING,"Failed enabling the AOF after successful master synchronization! Trying it again in one second.");
458 459 460 461 462 463 464
                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);
            }
        }
465
    }
466 467 468 469 470 471

    return;

error:
    replicationAbortSyncTransfer();
    return;
472 473
}

A
antirez 已提交
474
/* Send a synchronous command to the master. Used to send AUTH and
A
antirez 已提交
475
 * REPLCONF commands before starting the replication with SYNC.
A
antirez 已提交
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519
 *
 * 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. */
}

520
void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) {
A
antirez 已提交
521
    char tmpfile[256], *err;
522
    int dfd, maxtries = 5;
523 524
    int sockerr = 0;
    socklen_t errlen = sizeof(sockerr);
525 526 527
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(privdata);
    REDIS_NOTUSED(mask);
528

529 530
    /* 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
531
    if (server.repl_state == REDIS_REPL_NONE) {
532 533 534 535
        close(fd);
        return;
    }

536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591
    /* 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...");
        }
    }
592 593 594

    /* AUTH with the master if required. */
    if(server.masterauth) {
A
antirez 已提交
595 596 597 598
        err = sendSynchronousCommand(fd,"AUTH",server.masterauth,NULL);
        if (err) {
            redisLog(REDIS_WARNING,"Unable to AUTH to MASTER: %s",err);
            sdsfree(err);
599 600
            goto error;
        }
A
antirez 已提交
601 602 603 604 605 606 607 608 609 610 611 612 613 614
    }

    /* 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);
615 616 617 618
        }
    }

    /* Issue the SYNC command */
619
    if (syncWrite(fd,"SYNC\r\n",6,server.repl_syncio_timeout*1000) == -1) {
620 621
        redisLog(REDIS_WARNING,"I/O error writing to MASTER: %s",
            strerror(errno));
622
        goto error;
623
    }
624 625

    /* Prepare a suitable temp file for bulk transfer */
626 627
    while(maxtries--) {
        snprintf(tmpfile,256,
628
            "temp-%d.%ld.rdb",(int)server.unixtime,(long int)getpid());
629 630 631 632 633 634
        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));
635
        goto error;
636 637
    }

638
    /* Setup the non blocking download of the bulk file. */
639
    if (aeCreateFileEvent(server.el,fd, AE_READABLE,readSyncBulkPayload,NULL)
640
            == AE_ERR)
641
    {
642 643 644
        redisLog(REDIS_WARNING,
            "Can't create readable event for SYNC: %s (fd=%d)",
            strerror(errno),fd);
645
        goto error;
646
    }
647

A
antirez 已提交
648
    server.repl_state = REDIS_REPL_TRANSFER;
649 650 651
    server.repl_transfer_size = -1;
    server.repl_transfer_read = 0;
    server.repl_transfer_last_fsync_off = 0;
652
    server.repl_transfer_fd = dfd;
653
    server.repl_transfer_lastio = server.unixtime;
654
    server.repl_transfer_tmpfile = zstrdup(tmpfile);
655 656 657 658
    return;

error:
    close(fd);
659 660
    server.repl_transfer_s = -1;
    server.repl_state = REDIS_REPL_CONNECT;
661 662 663 664 665 666 667 668 669 670 671 672 673
    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;
    }

674
    if (aeCreateFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE,syncWithMaster,NULL) ==
675
            AE_ERR)
676 677 678 679 680 681
    {
        close(fd);
        redisLog(REDIS_WARNING,"Can't create readable event for SYNC");
        return REDIS_ERR;
    }

682
    server.repl_transfer_lastio = server.unixtime;
683
    server.repl_transfer_s = fd;
A
antirez 已提交
684
    server.repl_state = REDIS_REPL_CONNECTING;
685 686 687
    return REDIS_OK;
}

688 689 690 691 692
/* 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;

693 694
    redisAssert(server.repl_state == REDIS_REPL_CONNECTING ||
                server.repl_state == REDIS_REPL_RECEIVE_PONG);
695 696 697
    aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE);
    close(fd);
    server.repl_transfer_s = -1;
A
antirez 已提交
698
    server.repl_state = REDIS_REPL_CONNECT;
699 700
}

701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721
/* This function aborts a non blocking replication attempt if there is one
 * in progress, by canceling the non-blocking connect attempt or
 * the initial bulk transfer.
 *
 * If there was a replication handshake in progress 1 is returned and
 * the replication state (server.repl_state) set to REDIS_REPL_CONNECT.
 *
 * Otherwise zero is returned and no operation is perforemd at all. */
int cancelReplicationHandshake(void) {
    if (server.repl_state == REDIS_REPL_TRANSFER) {
        replicationAbortSyncTransfer();
    } else if (server.repl_state == REDIS_REPL_CONNECTING ||
             server.repl_state == REDIS_REPL_RECEIVE_PONG)
    {
        undoConnectWithMaster();
    } else {
        return 0;
    }
    return 1;
}

722 723 724 725 726 727 728
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);
729
            cancelReplicationHandshake();
A
antirez 已提交
730
            server.repl_state = REDIS_REPL_NONE;
731 732 733
            redisLog(REDIS_NOTICE,"MASTER MODE enabled (user request)");
        }
    } else {
734 735 736 737 738 739 740 741 742 743 744 745 746 747
        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. */
748 749
        sdsfree(server.masterhost);
        server.masterhost = sdsdup(c->argv[1]->ptr);
750
        server.masterport = port;
751
        if (server.master) freeClient(server.master);
752
        disconnectSlaves(); /* Force our slaves to resync with us as well. */
753
        cancelReplicationHandshake();
A
antirez 已提交
754
        server.repl_state = REDIS_REPL_CONNECT;
755 756 757 758 759
        redisLog(REDIS_NOTICE,"SLAVE OF %s:%d enabled (user request)",
            server.masterhost, server.masterport);
    }
    addReply(c,shared.ok);
}
760 761 762 763

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

void replicationCron(void) {
764
    /* Non blocking connection timeout? */
765 766 767
    if (server.masterhost &&
        (server.repl_state == REDIS_REPL_CONNECTING ||
         server.repl_state == REDIS_REPL_RECEIVE_PONG) &&
768 769 770 771 772 773
        (time(NULL)-server.repl_transfer_lastio) > server.repl_timeout)
    {
        redisLog(REDIS_WARNING,"Timeout connecting to the MASTER...");
        undoConnectWithMaster();
    }

774
    /* Bulk transfer I/O timeout? */
A
antirez 已提交
775
    if (server.masterhost && server.repl_state == REDIS_REPL_TRANSFER &&
A
7c6da73  
antirez 已提交
776
        (time(NULL)-server.repl_transfer_lastio) > server.repl_timeout)
777
    {
778
        redisLog(REDIS_WARNING,"Timeout receiving bulk data from MASTER... If the problem persists try to set the 'repl-timeout' parameter in redis.conf to a larger value.");
779 780 781
        replicationAbortSyncTransfer();
    }

782
    /* Timed out master when we are an already connected slave? */
A
antirez 已提交
783
    if (server.masterhost && server.repl_state == REDIS_REPL_CONNECTED &&
A
7c6da73  
antirez 已提交
784
        (time(NULL)-server.master->lastinteraction) > server.repl_timeout)
785 786 787 788 789
    {
        redisLog(REDIS_WARNING,"MASTER time out: no data nor PING received...");
        freeClient(server.master);
    }

790
    /* Check if we should connect to a MASTER */
A
antirez 已提交
791
    if (server.repl_state == REDIS_REPL_CONNECT) {
792
        redisLog(REDIS_NOTICE,"Connecting to MASTER...");
793 794
        if (connectWithMaster() == REDIS_OK) {
            redisLog(REDIS_NOTICE,"MASTER <-> SLAVE sync started");
795 796
        }
    }
797 798 799 800 801
    
    /* 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. */
802
    if (!(server.cronloops % (server.repl_ping_slave_period * server.hz))) {
803 804 805 806 807 808 809 810 811 812 813 814
        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 */
815
                addReplySds(slave,sdsnew("*1\r\n$4\r\nPING\r\n"));
816 817 818 819 820 821
            } 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
822 823 824
                if (write(slave->fd, "\n", 1) == -1) {
                    /* Don't worry, it's just a ping. */
                }
825 826 827
            }
        }
    }
828
}