+ if (server.vm_enabled) {
+ lockThreadedIO();
+ info = sdscatprintf(info,
+ "vm_conf_max_memory:%llu\r\n"
+ "vm_conf_page_size:%llu\r\n"
+ "vm_conf_pages:%llu\r\n"
+ "vm_stats_used_pages:%llu\r\n"
+ "vm_stats_swapped_objects:%llu\r\n"
+ "vm_stats_swappin_count:%llu\r\n"
+ "vm_stats_swappout_count:%llu\r\n"
+ "vm_stats_io_newjobs_len:%lu\r\n"
+ "vm_stats_io_processing_len:%lu\r\n"
+ "vm_stats_io_processed_len:%lu\r\n"
+ "vm_stats_io_active_threads:%lu\r\n"
+ "vm_stats_blocked_clients:%lu\r\n"
+ ,(unsigned long long) server.vm_max_memory,
+ (unsigned long long) server.vm_page_size,
+ (unsigned long long) server.vm_pages,
+ (unsigned long long) server.vm_stats_used_pages,
+ (unsigned long long) server.vm_stats_swapped_objects,
+ (unsigned long long) server.vm_stats_swapins,
+ (unsigned long long) server.vm_stats_swapouts,
+ (unsigned long) listLength(server.io_newjobs),
+ (unsigned long) listLength(server.io_processing),
+ (unsigned long) listLength(server.io_processed),
+ (unsigned long) server.io_active_threads,
+ (unsigned long) server.vm_blocked_clients
+ );
+ unlockThreadedIO();
+ }
+ for (j = 0; j < server.dbnum; j++) {
+ long long keys, vkeys;
+
+ keys = dictSize(server.db[j].dict);
+ vkeys = dictSize(server.db[j].expires);
+ if (keys || vkeys) {
+ info = sdscatprintf(info, "db%d:keys=%lld,expires=%lld\r\n",
+ j, keys, vkeys);
+ }
+ }
+ return info;
+}
+
+static void infoCommand(redisClient *c) {
+ sds info = genRedisInfoString();
+ addReplySds(c,sdscatprintf(sdsempty(),"$%lu\r\n",
+ (unsigned long)sdslen(info)));
+ addReplySds(c,info);
+ addReply(c,shared.crlf);
+}
+
+static void monitorCommand(redisClient *c) {
+ /* ignore MONITOR if aleady slave or in monitor mode */
+ if (c->flags & REDIS_SLAVE) return;
+
+ c->flags |= (REDIS_SLAVE|REDIS_MONITOR);
+ c->slaveseldb = 0;
+ listAddNodeTail(server.monitors,c);
+ addReply(c,shared.ok);
+}
+
+/* ================================= Expire ================================= */
+static int removeExpire(redisDb *db, robj *key) {
+ if (dictDelete(db->expires,key) == DICT_OK) {
+ return 1;
+ } else {
+ return 0;
+ }
+}
+
+static int setExpire(redisDb *db, robj *key, time_t when) {
+ if (dictAdd(db->expires,key,(void*)when) == DICT_ERR) {
+ return 0;
+ } else {
+ incrRefCount(key);
+ return 1;
+ }
+}
+
+/* Return the expire time of the specified key, or -1 if no expire
+ * is associated with this key (i.e. the key is non volatile) */
+static time_t getExpire(redisDb *db, robj *key) {
+ dictEntry *de;
+
+ /* No expire? return ASAP */
+ if (dictSize(db->expires) == 0 ||
+ (de = dictFind(db->expires,key)) == NULL) return -1;
+
+ return (time_t) dictGetEntryVal(de);
+}
+
+static int expireIfNeeded(redisDb *db, robj *key) {
+ time_t when;
+ dictEntry *de;
+
+ /* No expire? return ASAP */
+ if (dictSize(db->expires) == 0 ||
+ (de = dictFind(db->expires,key)) == NULL) return 0;
+
+ /* Lookup the expire */
+ when = (time_t) dictGetEntryVal(de);
+ if (time(NULL) <= when) return 0;
+
+ /* Delete the key */
+ dictDelete(db->expires,key);
+ return dictDelete(db->dict,key) == DICT_OK;
+}
+
+static int deleteIfVolatile(redisDb *db, robj *key) {
+ dictEntry *de;
+
+ /* No expire? return ASAP */
+ if (dictSize(db->expires) == 0 ||
+ (de = dictFind(db->expires,key)) == NULL) return 0;
+
+ /* Delete the key */
+ server.dirty++;
+ dictDelete(db->expires,key);
+ return dictDelete(db->dict,key) == DICT_OK;
+}
+
+static void expireGenericCommand(redisClient *c, robj *key, time_t seconds) {
+ dictEntry *de;
+
+ de = dictFind(c->db->dict,key);
+ if (de == NULL) {
+ addReply(c,shared.czero);
+ return;
+ }
+ if (seconds < 0) {
+ if (deleteKey(c->db,key)) server.dirty++;
+ addReply(c, shared.cone);
+ return;
+ } else {
+ time_t when = time(NULL)+seconds;
+ if (setExpire(c->db,key,when)) {
+ addReply(c,shared.cone);
+ server.dirty++;
+ } else {
+ addReply(c,shared.czero);
+ }
+ return;
+ }
+}
+
+static void expireCommand(redisClient *c) {
+ expireGenericCommand(c,c->argv[1],strtol(c->argv[2]->ptr,NULL,10));
+}
+
+static void expireatCommand(redisClient *c) {
+ expireGenericCommand(c,c->argv[1],strtol(c->argv[2]->ptr,NULL,10)-time(NULL));
+}
+
+static void ttlCommand(redisClient *c) {
+ time_t expire;
+ int ttl = -1;
+
+ expire = getExpire(c->db,c->argv[1]);
+ if (expire != -1) {
+ ttl = (int) (expire-time(NULL));
+ if (ttl < 0) ttl = -1;
+ }
+ addReplySds(c,sdscatprintf(sdsempty(),":%d\r\n",ttl));
+}
+
+/* ================================ MULTI/EXEC ============================== */
+
+/* Client state initialization for MULTI/EXEC */
+static void initClientMultiState(redisClient *c) {
+ c->mstate.commands = NULL;
+ c->mstate.count = 0;
+}
+
+/* Release all the resources associated with MULTI/EXEC state */
+static void freeClientMultiState(redisClient *c) {
+ int j;
+
+ for (j = 0; j < c->mstate.count; j++) {
+ int i;
+ multiCmd *mc = c->mstate.commands+j;
+
+ for (i = 0; i < mc->argc; i++)
+ decrRefCount(mc->argv[i]);
+ zfree(mc->argv);
+ }
+ zfree(c->mstate.commands);
+}
+
+/* Add a new command into the MULTI commands queue */
+static void queueMultiCommand(redisClient *c, struct redisCommand *cmd) {
+ multiCmd *mc;
+ int j;
+
+ c->mstate.commands = zrealloc(c->mstate.commands,
+ sizeof(multiCmd)*(c->mstate.count+1));
+ mc = c->mstate.commands+c->mstate.count;
+ mc->cmd = cmd;
+ mc->argc = c->argc;
+ mc->argv = zmalloc(sizeof(robj*)*c->argc);
+ memcpy(mc->argv,c->argv,sizeof(robj*)*c->argc);
+ for (j = 0; j < c->argc; j++)
+ incrRefCount(mc->argv[j]);
+ c->mstate.count++;
+}
+
+static void multiCommand(redisClient *c) {
+ c->flags |= REDIS_MULTI;
+ addReply(c,shared.ok);
+}
+
+static void discardCommand(redisClient *c) {
+ if (!(c->flags & REDIS_MULTI)) {
+ addReplySds(c,sdsnew("-ERR DISCARD without MULTI\r\n"));
+ return;
+ }
+
+ freeClientMultiState(c);
+ initClientMultiState(c);
+ c->flags &= (~REDIS_MULTI);
+ addReply(c,shared.ok);
+}
+
+static void execCommand(redisClient *c) {
+ int j;
+ robj **orig_argv;
+ int orig_argc;
+
+ if (!(c->flags & REDIS_MULTI)) {
+ addReplySds(c,sdsnew("-ERR EXEC without MULTI\r\n"));
+ return;
+ }
+
+ orig_argv = c->argv;
+ orig_argc = c->argc;
+ addReplySds(c,sdscatprintf(sdsempty(),"*%d\r\n",c->mstate.count));
+ for (j = 0; j < c->mstate.count; j++) {
+ c->argc = c->mstate.commands[j].argc;
+ c->argv = c->mstate.commands[j].argv;
+ call(c,c->mstate.commands[j].cmd);
+ }
+ c->argv = orig_argv;
+ c->argc = orig_argc;
+ freeClientMultiState(c);
+ initClientMultiState(c);
+ c->flags &= (~REDIS_MULTI);
+}
+
+/* =========================== Blocking Operations ========================= */
+
+/* Currently Redis blocking operations support is limited to list POP ops,
+ * so the current implementation is not fully generic, but it is also not
+ * completely specific so it will not require a rewrite to support new
+ * kind of blocking operations in the future.
+ *
+ * Still it's important to note that list blocking operations can be already
+ * used as a notification mechanism in order to implement other blocking
+ * operations at application level, so there must be a very strong evidence
+ * of usefulness and generality before new blocking operations are implemented.
+ *
+ * This is how the current blocking POP works, we use BLPOP as example:
+ * - If the user calls BLPOP and the key exists and contains a non empty list
+ * then LPOP is called instead. So BLPOP is semantically the same as LPOP
+ * if there is not to block.
+ * - If instead BLPOP is called and the key does not exists or the list is
+ * empty we need to block. In order to do so we remove the notification for
+ * new data to read in the client socket (so that we'll not serve new
+ * requests if the blocking request is not served). Also we put the client
+ * in a dictionary (db->blockingkeys) mapping keys to a list of clients
+ * blocking for this keys.
+ * - If a PUSH operation against a key with blocked clients waiting is
+ * performed, we serve the first in the list: basically instead to push
+ * the new element inside the list we return it to the (first / oldest)
+ * blocking client, unblock the client, and remove it form the list.
+ *
+ * The above comment and the source code should be enough in order to understand
+ * the implementation and modify / fix it later.
+ */
+
+/* Set a client in blocking mode for the specified key, with the specified
+ * timeout */
+static void blockForKeys(redisClient *c, robj **keys, int numkeys, time_t timeout) {
+ dictEntry *de;
+ list *l;
+ int j;
+
+ c->blockingkeys = zmalloc(sizeof(robj*)*numkeys);
+ c->blockingkeysnum = numkeys;
+ c->blockingto = timeout;
+ for (j = 0; j < numkeys; j++) {
+ /* Add the key in the client structure, to map clients -> keys */
+ c->blockingkeys[j] = keys[j];
+ incrRefCount(keys[j]);
+
+ /* And in the other "side", to map keys -> clients */
+ de = dictFind(c->db->blockingkeys,keys[j]);
+ if (de == NULL) {
+ int retval;
+
+ /* For every key we take a list of clients blocked for it */
+ l = listCreate();
+ retval = dictAdd(c->db->blockingkeys,keys[j],l);
+ incrRefCount(keys[j]);
+ assert(retval == DICT_OK);
+ } else {
+ l = dictGetEntryVal(de);
+ }
+ listAddNodeTail(l,c);
+ }
+ /* Mark the client as a blocked client */
+ c->flags |= REDIS_BLOCKED;
+ server.blpop_blocked_clients++;
+}
+
+/* Unblock a client that's waiting in a blocking operation such as BLPOP */
+static void unblockClientWaitingData(redisClient *c) {
+ dictEntry *de;
+ list *l;
+ int j;
+
+ assert(c->blockingkeys != NULL);
+ /* The client may wait for multiple keys, so unblock it for every key. */
+ for (j = 0; j < c->blockingkeysnum; j++) {
+ /* Remove this client from the list of clients waiting for this key. */
+ de = dictFind(c->db->blockingkeys,c->blockingkeys[j]);
+ assert(de != NULL);
+ l = dictGetEntryVal(de);
+ listDelNode(l,listSearchKey(l,c));
+ /* If the list is empty we need to remove it to avoid wasting memory */
+ if (listLength(l) == 0)
+ dictDelete(c->db->blockingkeys,c->blockingkeys[j]);
+ decrRefCount(c->blockingkeys[j]);
+ }
+ /* Cleanup the client structure */
+ zfree(c->blockingkeys);
+ c->blockingkeys = NULL;
+ c->flags &= (~REDIS_BLOCKED);
+ server.blpop_blocked_clients--;
+ /* We want to process data if there is some command waiting
+ * in the input buffer. Note that this is safe even if
+ * unblockClientWaitingData() gets called from freeClient() because
+ * freeClient() will be smart enough to call this function
+ * *after* c->querybuf was set to NULL. */
+ if (c->querybuf && sdslen(c->querybuf) > 0) processInputBuffer(c);
+}
+
+/* This should be called from any function PUSHing into lists.
+ * 'c' is the "pushing client", 'key' is the key it is pushing data against,
+ * 'ele' is the element pushed.
+ *
+ * If the function returns 0 there was no client waiting for a list push
+ * against this key.
+ *
+ * If the function returns 1 there was a client waiting for a list push
+ * against this key, the element was passed to this client thus it's not
+ * needed to actually add it to the list and the caller should return asap. */
+static int handleClientsWaitingListPush(redisClient *c, robj *key, robj *ele) {
+ struct dictEntry *de;
+ redisClient *receiver;
+ list *l;
+ listNode *ln;
+
+ de = dictFind(c->db->blockingkeys,key);
+ if (de == NULL) return 0;
+ l = dictGetEntryVal(de);
+ ln = listFirst(l);
+ assert(ln != NULL);
+ receiver = ln->value;
+
+ addReplySds(receiver,sdsnew("*2\r\n"));
+ addReplyBulkLen(receiver,key);
+ addReply(receiver,key);
+ addReply(receiver,shared.crlf);
+ addReplyBulkLen(receiver,ele);
+ addReply(receiver,ele);
+ addReply(receiver,shared.crlf);
+ unblockClientWaitingData(receiver);
+ return 1;
+}
+
+/* Blocking RPOP/LPOP */
+static void blockingPopGenericCommand(redisClient *c, int where) {
+ robj *o;
+ time_t timeout;
+ int j;
+
+ for (j = 1; j < c->argc-1; j++) {
+ o = lookupKeyWrite(c->db,c->argv[j]);
+ if (o != NULL) {
+ if (o->type != REDIS_LIST) {
+ addReply(c,shared.wrongtypeerr);
+ return;
+ } else {
+ list *list = o->ptr;
+ if (listLength(list) != 0) {
+ /* If the list contains elements fall back to the usual
+ * non-blocking POP operation */
+ robj *argv[2], **orig_argv;
+ int orig_argc;
+
+ /* We need to alter the command arguments before to call
+ * popGenericCommand() as the command takes a single key. */
+ orig_argv = c->argv;
+ orig_argc = c->argc;
+ argv[1] = c->argv[j];
+ c->argv = argv;
+ c->argc = 2;
+
+ /* Also the return value is different, we need to output
+ * the multi bulk reply header and the key name. The
+ * "real" command will add the last element (the value)
+ * for us. If this souds like an hack to you it's just
+ * because it is... */
+ addReplySds(c,sdsnew("*2\r\n"));
+ addReplyBulkLen(c,argv[1]);
+ addReply(c,argv[1]);
+ addReply(c,shared.crlf);
+ popGenericCommand(c,where);
+
+ /* Fix the client structure with the original stuff */
+ c->argv = orig_argv;
+ c->argc = orig_argc;
+ return;
+ }
+ }
+ }
+ }
+ /* If the list is empty or the key does not exists we must block */
+ timeout = strtol(c->argv[c->argc-1]->ptr,NULL,10);
+ if (timeout > 0) timeout += time(NULL);
+ blockForKeys(c,c->argv+1,c->argc-2,timeout);
+}
+
+static void blpopCommand(redisClient *c) {
+ blockingPopGenericCommand(c,REDIS_HEAD);
+}
+
+static void brpopCommand(redisClient *c) {
+ blockingPopGenericCommand(c,REDIS_TAIL);
+}
+
+/* =============================== Replication ============================= */
+
+static int syncWrite(int fd, char *ptr, ssize_t size, int timeout) {
+ ssize_t nwritten, ret = size;
+ time_t start = time(NULL);
+
+ timeout++;
+ while(size) {
+ if (aeWait(fd,AE_WRITABLE,1000) & AE_WRITABLE) {
+ nwritten = write(fd,ptr,size);
+ if (nwritten == -1) return -1;
+ ptr += nwritten;
+ size -= nwritten;
+ }
+ if ((time(NULL)-start) > timeout) {
+ errno = ETIMEDOUT;
+ return -1;
+ }
+ }
+ return ret;
+}
+
+static int syncRead(int fd, char *ptr, ssize_t size, int timeout) {
+ ssize_t nread, totread = 0;
+ time_t start = time(NULL);
+
+ timeout++;
+ while(size) {
+ if (aeWait(fd,AE_READABLE,1000) & AE_READABLE) {
+ nread = read(fd,ptr,size);
+ if (nread == -1) return -1;
+ ptr += nread;
+ size -= nread;
+ totread += nread;
+ }
+ if ((time(NULL)-start) > timeout) {
+ errno = ETIMEDOUT;
+ return -1;
+ }
+ }
+ return totread;
+}
+
+static int syncReadLine(int fd, char *ptr, ssize_t size, int timeout) {
+ ssize_t nread = 0;
+
+ size--;
+ while(size) {
+ char c;
+
+ if (syncRead(fd,&c,1,timeout) == -1) return -1;
+ if (c == '\n') {
+ *ptr = '\0';
+ if (nread && *(ptr-1) == '\r') *(ptr-1) = '\0';
+ return nread;
+ } else {
+ *ptr++ = c;
+ *ptr = '\0';
+ nread++;
+ }
+ }
+ return nread;
+}
+
+static void syncCommand(redisClient *c) {
+ /* ignore SYNC if aleady slave or in monitor mode */
+ if (c->flags & REDIS_SLAVE) return;
+
+ /* 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) {
+ addReplySds(c,sdsnew("-ERR SYNC is invalid with pending input\r\n"));
+ 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 */
+ if (server.bgsavechildpid != -1) {
+ /* 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. */
+ listRelease(c->reply);
+ c->reply = listDup(slave->reply);
+ 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");
+ if (rdbSaveBackground(server.dbfilename) != REDIS_OK) {
+ redisLog(REDIS_NOTICE,"Replication failed, can't BGSAVE");
+ addReplySds(c,sdsnew("-ERR Unalbe to perform background save\r\n"));
+ return;
+ }
+ c->replstate = REDIS_REPL_WAIT_BGSAVE_END;
+ }
+ c->repldbfd = -1;
+ c->flags |= REDIS_SLAVE;
+ c->slaveseldb = 0;
+ listAddNodeTail(server.slaves,c);
+ return;
+}
+
+static 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;
+ }
+ addReplySds(slave,sdsempty());
+ 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. */
+static 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;
+ }
+ if ((slave->repldbfd = open(server.dbfilename,O_RDONLY)) == -1 ||
+ 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) {
+ if (rdbSaveBackground(server.dbfilename) != REDIS_OK) {
+ 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);
+ }
+ }
+ }
+}
+
+static int syncWithMaster(void) {
+ char buf[1024], tmpfile[256], authcmd[1024];
+ int dumpsize;
+ int fd = anetTcpConnect(NULL,server.masterhost,server.masterport);
+ int dfd;
+
+ if (fd == -1) {
+ redisLog(REDIS_WARNING,"Unable to connect to MASTER: %s",
+ strerror(errno));
+ return REDIS_ERR;
+ }
+
+ /* 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);
+ redisLog(REDIS_WARNING,"Unable to AUTH to MASTER: %s",
+ strerror(errno));
+ return REDIS_ERR;
+ }
+ /* Read the AUTH result. */
+ if (syncReadLine(fd,buf,1024,3600) == -1) {
+ close(fd);
+ redisLog(REDIS_WARNING,"I/O error reading auth result from MASTER: %s",
+ strerror(errno));
+ return REDIS_ERR;
+ }
+ if (buf[0] != '+') {
+ close(fd);
+ redisLog(REDIS_WARNING,"Cannot AUTH to MASTER, is the masterauth password correct?");
+ return REDIS_ERR;
+ }
+ }
+
+ /* Issue the SYNC command */
+ if (syncWrite(fd,"SYNC \r\n",7,5) == -1) {
+ close(fd);
+ redisLog(REDIS_WARNING,"I/O error writing to MASTER: %s",
+ strerror(errno));
+ return REDIS_ERR;
+ }
+ /* Read the bulk write count */
+ if (syncReadLine(fd,buf,1024,3600) == -1) {
+ close(fd);
+ redisLog(REDIS_WARNING,"I/O error reading bulk count from MASTER: %s",
+ strerror(errno));
+ return REDIS_ERR;
+ }
+ if (buf[0] != '$') {
+ close(fd);
+ redisLog(REDIS_WARNING,"Bad protocol from MASTER, the first byte is not '$', are you sure the host and port are right?");
+ return REDIS_ERR;
+ }
+ dumpsize = atoi(buf+1);
+ redisLog(REDIS_NOTICE,"Receiving %d bytes data dump from MASTER",dumpsize);
+ /* Read the bulk write data on a temp file */
+ snprintf(tmpfile,256,"temp-%d.%ld.rdb",(int)time(NULL),(long int)random());
+ dfd = open(tmpfile,O_CREAT|O_WRONLY,0644);
+ if (dfd == -1) {
+ close(fd);
+ redisLog(REDIS_WARNING,"Opening the temp file needed for MASTER <-> SLAVE synchronization: %s",strerror(errno));
+ return REDIS_ERR;
+ }
+ while(dumpsize) {
+ int nread, nwritten;
+
+ nread = read(fd,buf,(dumpsize < 1024)?dumpsize:1024);
+ if (nread == -1) {
+ redisLog(REDIS_WARNING,"I/O error trying to sync with MASTER: %s",
+ strerror(errno));
+ close(fd);
+ close(dfd);
+ return REDIS_ERR;
+ }
+ nwritten = write(dfd,buf,nread);
+ if (nwritten == -1) {
+ redisLog(REDIS_WARNING,"Write error writing to the DB dump file needed for MASTER <-> SLAVE synchrnonization: %s", strerror(errno));
+ close(fd);
+ close(dfd);
+ return REDIS_ERR;
+ }
+ dumpsize -= nread;
+ }
+ close(dfd);
+ if (rename(tmpfile,server.dbfilename) == -1) {
+ redisLog(REDIS_WARNING,"Failed trying to rename the temp DB into dump.rdb in MASTER <-> SLAVE synchronization: %s", strerror(errno));
+ unlink(tmpfile);
+ close(fd);
+ return REDIS_ERR;
+ }
+ emptyDb();
+ if (rdbLoad(server.dbfilename) != REDIS_OK) {
+ redisLog(REDIS_WARNING,"Failed trying to load the MASTER synchronization DB from disk");
+ close(fd);
+ return REDIS_ERR;
+ }
+ server.master = createClient(fd);
+ server.master->flags |= REDIS_MASTER;
+ server.master->authenticated = 1;
+ server.replstate = REDIS_REPL_CONNECTED;
+ return REDIS_OK;
+}
+
+static 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);
+ server.replstate = REDIS_REPL_NONE;
+ redisLog(REDIS_NOTICE,"MASTER MODE enabled (user request)");
+ }
+ } else {
+ sdsfree(server.masterhost);
+ server.masterhost = sdsdup(c->argv[1]->ptr);
+ server.masterport = atoi(c->argv[2]->ptr);
+ if (server.master) freeClient(server.master);
+ server.replstate = REDIS_REPL_CONNECT;
+ redisLog(REDIS_NOTICE,"SLAVE OF %s:%d enabled (user request)",
+ server.masterhost, server.masterport);
+ }
+ addReply(c,shared.ok);
+}
+
+/* ============================ Maxmemory directive ======================== */
+
+/* Try to free one object form the pre-allocated objects free list.
+ * This is useful under low mem conditions as by default we take 1 million
+ * free objects allocated. On success REDIS_OK is returned, otherwise
+ * REDIS_ERR. */
+static int tryFreeOneObjectFromFreelist(void) {
+ robj *o;
+
+ if (server.vm_enabled) pthread_mutex_lock(&server.obj_freelist_mutex);
+ if (listLength(server.objfreelist)) {
+ listNode *head = listFirst(server.objfreelist);
+ o = listNodeValue(head);
+ listDelNode(server.objfreelist,head);
+ if (server.vm_enabled) pthread_mutex_unlock(&server.obj_freelist_mutex);
+ zfree(o);
+ return REDIS_OK;
+ } else {
+ if (server.vm_enabled) pthread_mutex_unlock(&server.obj_freelist_mutex);
+ return REDIS_ERR;
+ }
+}
+
+/* This function gets called when 'maxmemory' is set on the config file to limit
+ * the max memory used by the server, and we are out of memory.
+ * This function will try to, in order:
+ *
+ * - Free objects from the free list
+ * - Try to remove keys with an EXPIRE set
+ *
+ * It is not possible to free enough memory to reach used-memory < maxmemory
+ * the server will start refusing commands that will enlarge even more the
+ * memory usage.
+ */
+static void freeMemoryIfNeeded(void) {
+ while (server.maxmemory && zmalloc_used_memory() > server.maxmemory) {
+ int j, k, freed = 0;
+
+ if (tryFreeOneObjectFromFreelist() == REDIS_OK) continue;
+ for (j = 0; j < server.dbnum; j++) {
+ int minttl = -1;
+ robj *minkey = NULL;
+ struct dictEntry *de;
+
+ if (dictSize(server.db[j].expires)) {
+ freed = 1;
+ /* From a sample of three keys drop the one nearest to
+ * the natural expire */
+ for (k = 0; k < 3; k++) {
+ time_t t;
+
+ de = dictGetRandomKey(server.db[j].expires);
+ t = (time_t) dictGetEntryVal(de);
+ if (minttl == -1 || t < minttl) {
+ minkey = dictGetEntryKey(de);
+ minttl = t;
+ }
+ }
+ deleteKey(server.db+j,minkey);
+ }
+ }
+ if (!freed) return; /* nothing to free... */
+ }
+}
+
+/* ============================== Append Only file ========================== */
+
+static void feedAppendOnlyFile(struct redisCommand *cmd, int dictid, robj **argv, int argc) {
+ sds buf = sdsempty();
+ int j;
+ ssize_t nwritten;
+ time_t now;
+ robj *tmpargv[3];
+
+ /* The DB this command was targetting is not the same as the last command
+ * we appendend. To issue a SELECT command is needed. */
+ if (dictid != server.appendseldb) {
+ char seldb[64];
+
+ snprintf(seldb,sizeof(seldb),"%d",dictid);
+ buf = sdscatprintf(buf,"*2\r\n$6\r\nSELECT\r\n$%lu\r\n%s\r\n",
+ (unsigned long)strlen(seldb),seldb);
+ server.appendseldb = dictid;
+ }
+
+ /* "Fix" the argv vector if the command is EXPIRE. We want to translate
+ * EXPIREs into EXPIREATs calls */
+ if (cmd->proc == expireCommand) {
+ long when;
+
+ tmpargv[0] = createStringObject("EXPIREAT",8);
+ tmpargv[1] = argv[1];
+ incrRefCount(argv[1]);
+ when = time(NULL)+strtol(argv[2]->ptr,NULL,10);
+ tmpargv[2] = createObject(REDIS_STRING,
+ sdscatprintf(sdsempty(),"%ld",when));
+ argv = tmpargv;
+ }
+
+ /* Append the actual command */
+ buf = sdscatprintf(buf,"*%d\r\n",argc);
+ for (j = 0; j < argc; j++) {
+ robj *o = argv[j];
+
+ o = getDecodedObject(o);
+ buf = sdscatprintf(buf,"$%lu\r\n",(unsigned long)sdslen(o->ptr));
+ buf = sdscatlen(buf,o->ptr,sdslen(o->ptr));
+ buf = sdscatlen(buf,"\r\n",2);
+ decrRefCount(o);
+ }
+
+ /* Free the objects from the modified argv for EXPIREAT */
+ if (cmd->proc == expireCommand) {
+ for (j = 0; j < 3; j++)
+ decrRefCount(argv[j]);
+ }
+
+ /* We want to perform a single write. This should be guaranteed atomic
+ * at least if the filesystem we are writing is a real physical one.
+ * While this will save us against the server being killed I don't think
+ * there is much to do about the whole server stopping for power problems
+ * or alike */
+ nwritten = write(server.appendfd,buf,sdslen(buf));
+ if (nwritten != (signed)sdslen(buf)) {
+ /* Ooops, we are in troubles. The best thing to do for now is
+ * to simply exit instead to give the illusion that everything is
+ * working as expected. */
+ if (nwritten == -1) {
+ redisLog(REDIS_WARNING,"Exiting on error writing to the append-only file: %s",strerror(errno));
+ } else {
+ redisLog(REDIS_WARNING,"Exiting on short write while writing to the append-only file: %s",strerror(errno));
+ }
+ exit(1);
+ }
+ /* If a background append only file rewriting is in progress we want to
+ * accumulate the differences between the child DB and the current one
+ * in a buffer, so that when the child process will do its work we
+ * can append the differences to the new append only file. */
+ if (server.bgrewritechildpid != -1)
+ server.bgrewritebuf = sdscatlen(server.bgrewritebuf,buf,sdslen(buf));
+
+ sdsfree(buf);
+ now = time(NULL);
+ if (server.appendfsync == APPENDFSYNC_ALWAYS ||
+ (server.appendfsync == APPENDFSYNC_EVERYSEC &&
+ now-server.lastfsync > 1))
+ {
+ fsync(server.appendfd); /* Let's try to get this data on the disk */
+ server.lastfsync = now;
+ }
+}
+
+/* In Redis commands are always executed in the context of a client, so in
+ * order to load the append only file we need to create a fake client. */
+static struct redisClient *createFakeClient(void) {
+ struct redisClient *c = zmalloc(sizeof(*c));
+
+ selectDb(c,0);
+ c->fd = -1;
+ c->querybuf = sdsempty();
+ c->argc = 0;
+ c->argv = NULL;
+ c->flags = 0;
+ /* We set the fake client as a slave waiting for the synchronization
+ * so that Redis will not try to send replies to this client. */
+ c->replstate = REDIS_REPL_WAIT_BGSAVE_START;
+ c->reply = listCreate();
+ listSetFreeMethod(c->reply,decrRefCount);
+ listSetDupMethod(c->reply,dupClientReplyValue);
+ return c;
+}
+
+static void freeFakeClient(struct redisClient *c) {
+ sdsfree(c->querybuf);
+ listRelease(c->reply);
+ zfree(c);
+}
+
+/* Replay the append log file. On error REDIS_OK is returned. On non fatal
+ * error (the append only file is zero-length) REDIS_ERR is returned. On
+ * fatal error an error message is logged and the program exists. */
+int loadAppendOnlyFile(char *filename) {
+ struct redisClient *fakeClient;
+ FILE *fp = fopen(filename,"r");
+ struct redis_stat sb;
+ unsigned long long loadedkeys = 0;
+
+ if (redis_fstat(fileno(fp),&sb) != -1 && sb.st_size == 0)
+ return REDIS_ERR;
+
+ if (fp == NULL) {
+ redisLog(REDIS_WARNING,"Fatal error: can't open the append log file for reading: %s",strerror(errno));
+ exit(1);
+ }
+
+ fakeClient = createFakeClient();
+ while(1) {
+ int argc, j;
+ unsigned long len;
+ robj **argv;
+ char buf[128];
+ sds argsds;
+ struct redisCommand *cmd;
+
+ if (fgets(buf,sizeof(buf),fp) == NULL) {
+ if (feof(fp))
+ break;
+ else
+ goto readerr;
+ }
+ if (buf[0] != '*') goto fmterr;
+ argc = atoi(buf+1);
+ argv = zmalloc(sizeof(robj*)*argc);
+ for (j = 0; j < argc; j++) {
+ if (fgets(buf,sizeof(buf),fp) == NULL) goto readerr;
+ if (buf[0] != '$') goto fmterr;
+ len = strtol(buf+1,NULL,10);
+ argsds = sdsnewlen(NULL,len);
+ if (len && fread(argsds,len,1,fp) == 0) goto fmterr;
+ argv[j] = createObject(REDIS_STRING,argsds);
+ if (fread(buf,2,1,fp) == 0) goto fmterr; /* discard CRLF */
+ }
+
+ /* Command lookup */
+ cmd = lookupCommand(argv[0]->ptr);
+ if (!cmd) {
+ redisLog(REDIS_WARNING,"Unknown command '%s' reading the append only file", argv[0]->ptr);
+ exit(1);
+ }
+ /* Try object sharing and encoding */
+ if (server.shareobjects) {
+ int j;
+ for(j = 1; j < argc; j++)
+ argv[j] = tryObjectSharing(argv[j]);
+ }
+ if (cmd->flags & REDIS_CMD_BULK)
+ tryObjectEncoding(argv[argc-1]);
+ /* Run the command in the context of a fake client */
+ fakeClient->argc = argc;
+ fakeClient->argv = argv;
+ cmd->proc(fakeClient);
+ /* Discard the reply objects list from the fake client */
+ while(listLength(fakeClient->reply))
+ listDelNode(fakeClient->reply,listFirst(fakeClient->reply));
+ /* Clean up, ready for the next command */
+ for (j = 0; j < argc; j++) decrRefCount(argv[j]);
+ zfree(argv);
+ /* Handle swapping while loading big datasets when VM is on */
+ loadedkeys++;
+ if (server.vm_enabled && (loadedkeys % 5000) == 0) {
+ while (zmalloc_used_memory() > server.vm_max_memory) {
+ if (vmSwapOneObjectBlocking() == REDIS_ERR) break;
+ }
+ }
+ }
+ fclose(fp);
+ freeFakeClient(fakeClient);
+ return REDIS_OK;
+
+readerr:
+ if (feof(fp)) {
+ redisLog(REDIS_WARNING,"Unexpected end of file reading the append only file");
+ } else {
+ redisLog(REDIS_WARNING,"Unrecoverable error reading the append only file: %s", strerror(errno));
+ }
+ exit(1);
+fmterr:
+ redisLog(REDIS_WARNING,"Bad file format reading the append only file");
+ exit(1);
+}
+
+/* Write an object into a file in the bulk format $<count>\r\n<payload>\r\n */
+static int fwriteBulk(FILE *fp, robj *obj) {
+ char buf[128];
+ int decrrc = 0;
+
+ /* Avoid the incr/decr ref count business if possible to help
+ * copy-on-write (we are often in a child process when this function
+ * is called).
+ * Also makes sure that key objects don't get incrRefCount-ed when VM
+ * is enabled */
+ if (obj->encoding != REDIS_ENCODING_RAW) {
+ obj = getDecodedObject(obj);
+ decrrc = 1;
+ }
+ snprintf(buf,sizeof(buf),"$%ld\r\n",(long)sdslen(obj->ptr));
+ if (fwrite(buf,strlen(buf),1,fp) == 0) goto err;
+ if (sdslen(obj->ptr) && fwrite(obj->ptr,sdslen(obj->ptr),1,fp) == 0)
+ goto err;
+ if (fwrite("\r\n",2,1,fp) == 0) goto err;
+ if (decrrc) decrRefCount(obj);
+ return 1;
+err:
+ if (decrrc) decrRefCount(obj);
+ return 0;
+}
+
+/* Write a double value in bulk format $<count>\r\n<payload>\r\n */
+static int fwriteBulkDouble(FILE *fp, double d) {
+ char buf[128], dbuf[128];
+
+ snprintf(dbuf,sizeof(dbuf),"%.17g\r\n",d);
+ snprintf(buf,sizeof(buf),"$%lu\r\n",(unsigned long)strlen(dbuf)-2);
+ if (fwrite(buf,strlen(buf),1,fp) == 0) return 0;
+ if (fwrite(dbuf,strlen(dbuf),1,fp) == 0) return 0;
+ return 1;
+}
+
+/* Write a long value in bulk format $<count>\r\n<payload>\r\n */
+static int fwriteBulkLong(FILE *fp, long l) {
+ char buf[128], lbuf[128];
+
+ snprintf(lbuf,sizeof(lbuf),"%ld\r\n",l);
+ snprintf(buf,sizeof(buf),"$%lu\r\n",(unsigned long)strlen(lbuf)-2);
+ if (fwrite(buf,strlen(buf),1,fp) == 0) return 0;
+ if (fwrite(lbuf,strlen(lbuf),1,fp) == 0) return 0;
+ return 1;
+}
+
+/* Write a sequence of commands able to fully rebuild the dataset into
+ * "filename". Used both by REWRITEAOF and BGREWRITEAOF. */
+static int rewriteAppendOnlyFile(char *filename) {
+ dictIterator *di = NULL;
+ dictEntry *de;
+ FILE *fp;
+ char tmpfile[256];
+ int j;
+ time_t now = time(NULL);
+
+ /* Note that we have to use a different temp name here compared to the
+ * one used by rewriteAppendOnlyFileBackground() function. */
+ snprintf(tmpfile,256,"temp-rewriteaof-%d.aof", (int) getpid());
+ fp = fopen(tmpfile,"w");
+ if (!fp) {
+ redisLog(REDIS_WARNING, "Failed rewriting the append only file: %s", strerror(errno));
+ return REDIS_ERR;
+ }
+ for (j = 0; j < server.dbnum; j++) {
+ char selectcmd[] = "*2\r\n$6\r\nSELECT\r\n";
+ redisDb *db = server.db+j;
+ dict *d = db->dict;
+ if (dictSize(d) == 0) continue;
+ di = dictGetIterator(d);
+ if (!di) {
+ fclose(fp);
+ return REDIS_ERR;
+ }
+
+ /* SELECT the new DB */
+ if (fwrite(selectcmd,sizeof(selectcmd)-1,1,fp) == 0) goto werr;
+ if (fwriteBulkLong(fp,j) == 0) goto werr;
+
+ /* Iterate this DB writing every entry */
+ while((de = dictNext(di)) != NULL) {
+ robj *key, *o;
+ time_t expiretime;
+ int swapped;
+
+ key = dictGetEntryKey(de);
+ /* If the value for this key is swapped, load a preview in memory.
+ * We use a "swapped" flag to remember if we need to free the
+ * value object instead to just increment the ref count anyway
+ * in order to avoid copy-on-write of pages if we are forked() */
+ if (!server.vm_enabled || key->storage == REDIS_VM_MEMORY ||
+ key->storage == REDIS_VM_SWAPPING) {
+ o = dictGetEntryVal(de);
+ swapped = 0;
+ } else {
+ o = vmPreviewObject(key);
+ swapped = 1;
+ }
+ expiretime = getExpire(db,key);
+
+ /* Save the key and associated value */
+ if (o->type == REDIS_STRING) {
+ /* Emit a SET command */
+ char cmd[]="*3\r\n$3\r\nSET\r\n";
+ if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
+ /* Key and value */
+ if (fwriteBulk(fp,key) == 0) goto werr;
+ if (fwriteBulk(fp,o) == 0) goto werr;
+ } else if (o->type == REDIS_LIST) {
+ /* Emit the RPUSHes needed to rebuild the list */
+ list *list = o->ptr;
+ listNode *ln;
+ listIter li;
+
+ listRewind(list,&li);
+ while((ln = listNext(&li))) {
+ char cmd[]="*3\r\n$5\r\nRPUSH\r\n";
+ robj *eleobj = listNodeValue(ln);
+
+ if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
+ if (fwriteBulk(fp,key) == 0) goto werr;
+ if (fwriteBulk(fp,eleobj) == 0) goto werr;
+ }
+ } else if (o->type == REDIS_SET) {
+ /* Emit the SADDs needed to rebuild the set */
+ dict *set = o->ptr;
+ dictIterator *di = dictGetIterator(set);
+ dictEntry *de;
+
+ while((de = dictNext(di)) != NULL) {
+ char cmd[]="*3\r\n$4\r\nSADD\r\n";
+ robj *eleobj = dictGetEntryKey(de);
+
+ if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
+ if (fwriteBulk(fp,key) == 0) goto werr;
+ if (fwriteBulk(fp,eleobj) == 0) goto werr;
+ }
+ dictReleaseIterator(di);
+ } else if (o->type == REDIS_ZSET) {
+ /* Emit the ZADDs needed to rebuild the sorted set */
+ zset *zs = o->ptr;
+ dictIterator *di = dictGetIterator(zs->dict);
+ dictEntry *de;
+
+ while((de = dictNext(di)) != NULL) {
+ char cmd[]="*4\r\n$4\r\nZADD\r\n";
+ robj *eleobj = dictGetEntryKey(de);
+ double *score = dictGetEntryVal(de);
+
+ if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
+ if (fwriteBulk(fp,key) == 0) goto werr;
+ if (fwriteBulkDouble(fp,*score) == 0) goto werr;
+ if (fwriteBulk(fp,eleobj) == 0) goto werr;
+ }
+ dictReleaseIterator(di);
+ } else {
+ redisAssert(0 != 0);
+ }
+ /* Save the expire time */
+ if (expiretime != -1) {
+ char cmd[]="*3\r\n$8\r\nEXPIREAT\r\n";
+ /* If this key is already expired skip it */
+ if (expiretime < now) continue;
+ if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
+ if (fwriteBulk(fp,key) == 0) goto werr;
+ if (fwriteBulkLong(fp,expiretime) == 0) goto werr;
+ }
+ if (swapped) decrRefCount(o);
+ }
+ dictReleaseIterator(di);
+ }
+
+ /* Make sure data will not remain on the OS's output buffers */
+ fflush(fp);
+ fsync(fileno(fp));
+ fclose(fp);
+
+ /* Use RENAME to make sure the DB file is changed atomically only
+ * if the generate DB file is ok. */
+ if (rename(tmpfile,filename) == -1) {
+ redisLog(REDIS_WARNING,"Error moving temp append only file on the final destination: %s", strerror(errno));
+ unlink(tmpfile);
+ return REDIS_ERR;
+ }
+ redisLog(REDIS_NOTICE,"SYNC append only file rewrite performed");
+ return REDIS_OK;
+
+werr:
+ fclose(fp);
+ unlink(tmpfile);
+ redisLog(REDIS_WARNING,"Write error writing append only file on disk: %s", strerror(errno));
+ if (di) dictReleaseIterator(di);
+ return REDIS_ERR;
projects / redis.git / blobdiff