X-Git-Url: https://git.saurik.com/redis.git/blobdiff_plain/3b5e72d402345507856b00bb67a91942f4f1a49d..63fd13996d5ff19a642709aef78d655939cf8ced:/src/replication.c diff --git a/src/replication.c b/src/replication.c index f6ecc1c6..712af110 100644 --- a/src/replication.c +++ b/src/replication.c @@ -10,38 +10,8 @@ void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) { listNode *ln; listIter li; - int outc = 0, j; - robj **outv; - /* We need 1+(ARGS*3) objects since commands are using the new protocol - * and we one 1 object for the first "*\r\n" multibulk count, then - * for every additional object we have "$\r\n" + object + "\r\n". */ - robj *static_outv[REDIS_STATIC_ARGS*3+1]; - robj *lenobj; - - if (argc <= REDIS_STATIC_ARGS) { - outv = static_outv; - } else { - outv = zmalloc(sizeof(robj*)*(argc*3+1)); - } - - lenobj = createObject(REDIS_STRING, - sdscatprintf(sdsempty(), "*%d\r\n", argc)); - lenobj->refcount = 0; - outv[outc++] = lenobj; - for (j = 0; j < argc; j++) { - lenobj = createObject(REDIS_STRING, - sdscatprintf(sdsempty(),"$%lu\r\n", - (unsigned long) stringObjectLen(argv[j]))); - lenobj->refcount = 0; - outv[outc++] = lenobj; - outv[outc++] = argv[j]; - outv[outc++] = shared.crlf; - } + int j; - /* Increment all the refcounts at start and decrement at end in order to - * be sure to free objects if there is no slave in a replication state - * able to be feed with commands */ - for (j = 0; j < outc; j++) incrRefCount(outv[j]); listRewind(slaves,&li); while((ln = listNext(&li))) { redisClient *slave = ln->value; @@ -49,7 +19,9 @@ void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) { /* Don't feed slaves that are still waiting for BGSAVE to start */ if (slave->replstate == REDIS_REPL_WAIT_BGSAVE_START) continue; - /* Feed all the other slaves, MONITORs and so on */ + /* 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. */ if (slave->slaveseldb != dictid) { robj *selectcmd; @@ -73,10 +45,9 @@ void replicationFeedSlaves(list *slaves, int dictid, robj **argv, int argc) { addReply(slave,selectcmd); slave->slaveseldb = dictid; } - for (j = 0; j < outc; j++) addReply(slave,outv[j]); + addReplyMultiBulkLen(slave,argc); + for (j = 0; j < argc; j++) addReplyBulk(slave,argv[j]); } - for (j = 0; j < outc; j++) decrRefCount(outv[j]); - if (outv != static_outv) zfree(outv); } void replicationFeedMonitors(list *monitors, int dictid, robj **argv, int argc) { @@ -88,7 +59,7 @@ void replicationFeedMonitors(list *monitors, int dictid, robj **argv, int argc) struct timeval tv; gettimeofday(&tv,NULL); - cmdrepr = sdscatprintf(cmdrepr,"%ld.%ld ",(long)tv.tv_sec,(long)tv.tv_usec); + cmdrepr = sdscatprintf(cmdrepr,"%ld.%06ld ",(long)tv.tv_sec,(long)tv.tv_usec); if (dictid != 0) cmdrepr = sdscatprintf(cmdrepr,"(db %d) ", dictid); for (j = 0; j < argc; j++) { @@ -315,23 +286,27 @@ void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) { /* If repl_transfer_left == -1 we still have to read the bulk length * from the master reply. */ if (server.repl_transfer_left == -1) { - if (syncReadLine(fd,buf,1024,3600) == -1) { + if (syncReadLine(fd,buf,1024,server.repl_syncio_timeout) == -1) { redisLog(REDIS_WARNING, "I/O error reading bulk count from MASTER: %s", strerror(errno)); - replicationAbortSyncTransfer(); - return; + goto error; } + if (buf[0] == '-') { redisLog(REDIS_WARNING, "MASTER aborted replication with an error: %s", buf+1); - replicationAbortSyncTransfer(); + goto error; + } 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. */ + server.repl_transfer_lastio = time(NULL); return; } 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?"); - replicationAbortSyncTransfer(); - return; + goto error; } server.repl_transfer_left = strtol(buf+1,NULL,10); redisLog(REDIS_NOTICE, @@ -353,8 +328,7 @@ void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) { server.repl_transfer_lastio = time(NULL); if (write(server.repl_transfer_fd,buf,nread) != nread) { redisLog(REDIS_WARNING,"Write error or short write writing to the DB dump file needed for MASTER <-> SLAVE synchrnonization: %s", strerror(errno)); - replicationAbortSyncTransfer(); - return; + goto error; } server.repl_transfer_left -= nread; /* Check if the transfer is now complete */ @@ -366,13 +340,17 @@ void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) { } redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Loading DB in memory"); emptyDb(); + /* 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); if (rdbLoad(server.dbfilename) != REDIS_OK) { redisLog(REDIS_WARNING,"Failed trying to load the MASTER synchronization DB from disk"); replicationAbortSyncTransfer(); return; } /* Final setup of the connected slave <- master link */ - aeDeleteFileEvent(server.el,server.repl_transfer_s,AE_READABLE); zfree(server.repl_transfer_tmpfile); close(server.repl_transfer_fd); server.master = createClient(server.repl_transfer_s); @@ -380,49 +358,65 @@ void readSyncBulkPayload(aeEventLoop *el, int fd, void *privdata, int mask) { server.master->authenticated = 1; server.replstate = REDIS_REPL_CONNECTED; redisLog(REDIS_NOTICE, "MASTER <-> SLAVE sync: Finished with success"); + /* Rewrite the AOF file now that the dataset changed. */ + if (server.appendonly) rewriteAppendOnlyFileBackground(); } + + return; + +error: + replicationAbortSyncTransfer(); + return; } -int syncWithMaster(void) { - char buf[1024], tmpfile[256], authcmd[1024]; - int fd = anetTcpConnect(NULL,server.masterhost,server.masterport); +void syncWithMaster(aeEventLoop *el, int fd, void *privdata, int mask) { + char buf[1024], tmpfile[256]; int dfd, maxtries = 5; + REDIS_NOTUSED(el); + REDIS_NOTUSED(privdata); + REDIS_NOTUSED(mask); - if (fd == -1) { - redisLog(REDIS_WARNING,"Unable to connect to MASTER: %s", - strerror(errno)); - return REDIS_ERR; + /* If this event fired after the user turned the instance into a master + * with SLAVEOF NO ONE we must just return ASAP. */ + if (server.replstate == REDIS_REPL_NONE) { + close(fd); + return; } + redisLog(REDIS_NOTICE,"Non blocking connect for SYNC fired the event."); + /* This event should only be triggered once since it is used to have a + * non-blocking connect(2) to the master. It has been triggered when this + * function is called, so we can delete it. */ + aeDeleteFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE); + /* AUTH with the master if required. */ if(server.masterauth) { - snprintf(authcmd, 1024, "AUTH %s\r\n", server.masterauth); - if (syncWrite(fd, authcmd, strlen(server.masterauth)+7, 5) == -1) { - close(fd); + char authcmd[1024]; + size_t authlen; + + authlen = snprintf(authcmd,sizeof(authcmd),"AUTH %s\r\n",server.masterauth); + if (syncWrite(fd,authcmd,authlen,server.repl_syncio_timeout) == -1) { redisLog(REDIS_WARNING,"Unable to AUTH to MASTER: %s", strerror(errno)); - return REDIS_ERR; - } + goto error; + } /* Read the AUTH result. */ - if (syncReadLine(fd,buf,1024,3600) == -1) { - close(fd); + if (syncReadLine(fd,buf,1024,server.repl_syncio_timeout) == -1) { redisLog(REDIS_WARNING,"I/O error reading auth result from MASTER: %s", strerror(errno)); - return REDIS_ERR; + goto error; } if (buf[0] != '+') { - close(fd); redisLog(REDIS_WARNING,"Cannot AUTH to MASTER, is the masterauth password correct?"); - return REDIS_ERR; + goto error; } } /* Issue the SYNC command */ - if (syncWrite(fd,"SYNC \r\n",7,5) == -1) { - close(fd); + if (syncWrite(fd,"SYNC \r\n",7,server.repl_syncio_timeout) == -1) { redisLog(REDIS_WARNING,"I/O error writing to MASTER: %s", strerror(errno)); - return REDIS_ERR; + goto error; } /* Prepare a suitable temp file for bulk transfer */ @@ -434,25 +428,51 @@ int syncWithMaster(void) { sleep(1); } if (dfd == -1) { - close(fd); redisLog(REDIS_WARNING,"Opening the temp file needed for MASTER <-> SLAVE synchronization: %s",strerror(errno)); - return REDIS_ERR; + goto error; } /* Setup the non blocking download of the bulk file. */ - if (aeCreateFileEvent(server.el, fd, AE_READABLE, readSyncBulkPayload, NULL) + if (aeCreateFileEvent(server.el,fd, AE_READABLE,readSyncBulkPayload,NULL) == AE_ERR) { - close(fd); redisLog(REDIS_WARNING,"Can't create readable event for SYNC"); - return REDIS_ERR; + goto error; } + server.replstate = REDIS_REPL_TRANSFER; server.repl_transfer_left = -1; - server.repl_transfer_s = fd; server.repl_transfer_fd = dfd; server.repl_transfer_lastio = time(NULL); server.repl_transfer_tmpfile = zstrdup(tmpfile); + return; + +error: + server.replstate = REDIS_REPL_CONNECT; + close(fd); + 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; + } + + if (aeCreateFileEvent(server.el,fd,AE_READABLE|AE_WRITABLE,syncWithMaster,NULL) == + AE_ERR) + { + close(fd); + redisLog(REDIS_WARNING,"Can't create readable event for SYNC"); + return REDIS_ERR; + } + + server.repl_transfer_s = fd; + server.replstate = REDIS_REPL_CONNECTING; return REDIS_OK; } @@ -484,23 +504,59 @@ void slaveofCommand(redisClient *c) { /* --------------------------- REPLICATION CRON ---------------------------- */ -#define REDIS_REPL_TRANSFER_TIMEOUT 60 - void replicationCron(void) { /* Bulk transfer I/O timeout? */ if (server.masterhost && server.replstate == REDIS_REPL_TRANSFER && - (time(NULL)-server.repl_transfer_lastio) > REDIS_REPL_TRANSFER_TIMEOUT) + (time(NULL)-server.repl_transfer_lastio) > server.repl_timeout) { redisLog(REDIS_WARNING,"Timeout receiving bulk data from MASTER..."); replicationAbortSyncTransfer(); } + /* Timed out master when we are an already connected slave? */ + if (server.masterhost && server.replstate == REDIS_REPL_CONNECTED && + (time(NULL)-server.master->lastinteraction) > server.repl_timeout) + { + redisLog(REDIS_WARNING,"MASTER time out: no data nor PING received..."); + freeClient(server.master); + } + /* Check if we should connect to a MASTER */ if (server.replstate == REDIS_REPL_CONNECT) { redisLog(REDIS_NOTICE,"Connecting to MASTER..."); - if (syncWithMaster() == REDIS_OK) { - redisLog(REDIS_NOTICE,"MASTER <-> SLAVE sync started: SYNC sent"); - if (server.appendonly) rewriteAppendOnlyFileBackground(); + if (connectWithMaster() == REDIS_OK) { + redisLog(REDIS_NOTICE,"MASTER <-> SLAVE sync started"); + } + } + + /* 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. */ + if (!(server.cronloops % (server.repl_ping_slave_period*10))) { + 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 */ + addReplySds(slave,sdsnew("PING\r\n")); + } 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. */ + if (write(slave->fd, "\n", 1) == -1) { + /* Don't worry, it's just a ping. */ + } + } } } }