aof.c

#include "redis.h"
#include "bio.h"

#include <signal.h>
#include <fcntl.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <sys/time.h>
#ifdef _WIN32
  #include "win32fixes.h"
#else
  #include <sys/resource.h>
  #include <sys/wait.h>
#endif

void aofUpdateCurrentSize(void);

void aof_background_fsync(int fd) {
#ifdef _WIN32
    bioCreateBackgroundJob(REDIS_BIO_AOF_FSYNC,(void*)(size_t)fd,NULL,NULL);
#else
    bioCreateBackgroundJob(REDIS_BIO_AOF_FSYNC,(void*)(long)fd,NULL,NULL);
#endif
}

/* Called when the user switches from "appendonly yes" to "appendonly no"
 * at runtime using the CONFIG command. */
void stopAppendOnly(void) {
    flushAppendOnlyFile(1);
    aof_fsync(server.appendfd);
    close(server.appendfd);

    server.appendfd = -1;
    server.appendseldb = -1;
    server.appendonly = 0;
    /* rewrite operation in progress? kill it, wait child exit */
    if (server.bgrewritechildpid != -1) {
#ifdef _WIN32
        /* Windows placeholder for killing whatever lounched instead of fork()  */
        w32CeaseAndDesist(server.bgsavechildpid);
#else
      int statloc;
      
        if (kill(server.bgrewritechildpid,SIGKILL) != -1)
            wait3(&statloc,0,NULL);
#endif
        /* reset the buffer accumulating changes while the child saves */
        sdsfree(server.bgrewritebuf);
        server.bgrewritebuf = sdsempty();
        server.bgrewritechildpid = -1;
    }
}

/* Called when the user switches from "appendonly no" to "appendonly yes"
 * at runtime using the CONFIG command. */
int startAppendOnly(void) {
    server.appendonly = 1;
    server.lastfsync = time(NULL);
#ifdef _WIN32
    server.appendfd = open(server.appendfilename,O_WRONLY|O_APPEND|O_CREAT|_O_BINARY,_S_IREAD|_S_IWRITE);
#else
    server.appendfd = open(server.appendfilename,O_WRONLY|O_APPEND|O_CREAT,0644);
#endif
    if (server.appendfd == -1) {
        redisLog(REDIS_WARNING,"Used tried to switch on AOF via CONFIG, but I can't open the AOF file: %s",strerror(errno));
        return REDIS_ERR;
    }
    if (rewriteAppendOnlyFileBackground() == REDIS_ERR) {
        server.appendonly = 0;
        close(server.appendfd);
        redisLog(REDIS_WARNING,"Used tried to switch on AOF via CONFIG, I can't trigger a background AOF rewrite operation. Check the above logs for more info about the error.",strerror(errno));
        return REDIS_ERR;
    }
    return REDIS_OK;
}

/* Write the append only file buffer on disk.
 *
 * Since we are required to write the AOF before replying to the client,
 * and the only way the client socket can get a write is entering when the
 * the event loop, we accumulate all the AOF writes in a memory
 * buffer and write it on disk using this function just before entering
 * the event loop again.
 *
 * About the 'force' argument:
 *
 * When the fsync policy is set to 'everysec' we may delay the flush if there
 * is still an fsync() going on in the background thread, since for instance
 * on Linux write(2) will be blocked by the background fsync anyway.
 * When this happens we remember that there is some aof buffer to be
 * flushed ASAP, and will try to do that in the serverCron() function.
 *
 * However if force is set to 1 we'll write regardless of the background
 * fsync. */
void flushAppendOnlyFile(int force) {
    ssize_t nwritten;
    int sync_in_progress = 0;

    if (sdslen(server.aofbuf) == 0) return;

    if (server.appendfsync == APPENDFSYNC_EVERYSEC)
        sync_in_progress = bioPendingJobsOfType(REDIS_BIO_AOF_FSYNC) != 0;

    if (server.appendfsync == APPENDFSYNC_EVERYSEC && !force) {
        /* With this append fsync policy we do background fsyncing.
         * If the fsync is still in progress we can try to delay
         * the write for a couple of seconds. */
        if (sync_in_progress) {
            if (server.aof_flush_postponed_start == 0) {
                /* No previous write postponinig, remember that we are
                 * postponing the flush and return. */
                server.aof_flush_postponed_start = server.unixtime;
                return;
            } else if (server.unixtime - server.aof_flush_postponed_start < 2) {
                /* We were already waiting for fsync to finish, but for less
                 * than two seconds this is still ok. Postpone again. */
                return;
            }
            /* Otherwise fall trough, and go write since we can't wait
             * over two seconds. */
            redisLog(REDIS_NOTICE,"Asynchronous AOF fsync is taking too long (disk is busy?). Writing the AOF buffer without waiting for fsync to complete, this may slow down Redis.");
        }
    }
    /* If you are following this code path, then we are going to write so
     * set reset the postponed flush sentinel to zero. */
    server.aof_flush_postponed_start = 0;

    /* 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,server.aofbuf,sdslen(server.aofbuf));
    if (nwritten != (signed)sdslen(server.aofbuf)) {
        /* Ooops, we are in troubles. The best thing to do for now is
         * aborting instead of giving 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);
    }
    server.appendonly_current_size += nwritten;

    /* Re-use AOF buffer when it is small enough. The maximum comes from the
     * arena size of 4k minus some overhead (but is otherwise arbitrary). */
    if ((sdslen(server.aofbuf)+sdsavail(server.aofbuf)) < 4000) {
        sdsclear(server.aofbuf);
    } else {
        sdsfree(server.aofbuf);
        server.aofbuf = sdsempty();
    }

    /* Don't fsync if no-appendfsync-on-rewrite is set to yes and there are
     * children doing I/O in the background. */
    if (server.no_appendfsync_on_rewrite &&
        (server.bgrewritechildpid != -1 || server.bgsavechildpid != -1))
            return;

    /* Perform the fsync if needed. */
    if (server.appendfsync == APPENDFSYNC_ALWAYS) {
        /* aof_fsync is defined as fdatasync() for Linux in order to avoid
         * flushing metadata. */
        aof_fsync(server.appendfd); /* Let's try to get this data on the disk */
        server.lastfsync = server.unixtime;
    } else if ((server.appendfsync == APPENDFSYNC_EVERYSEC &&
                server.unixtime > server.lastfsync)) {
        if (!sync_in_progress) aof_background_fsync(server.appendfd);
        server.lastfsync = server.unixtime;
    }
}

sds catAppendOnlyGenericCommand(sds dst, int argc, robj **argv) {
    char buf[32];
    int len, j;
    robj *o;

    buf[0] = '*';
    len = 1+ll2string(buf+1,sizeof(buf)-1,argc);
    buf[len++] = '\r';
    buf[len++] = '\n';
    dst = sdscatlen(dst,buf,len);

    for (j = 0; j < argc; j++) {
        o = getDecodedObject(argv[j]);
        buf[0] = '$';
        len = 1+ll2string(buf+1,sizeof(buf)-1,sdslen(o->ptr));
        buf[len++] = '\r';
        buf[len++] = '\n';
        dst = sdscatlen(dst,buf,len);
        dst = sdscatlen(dst,o->ptr,sdslen(o->ptr));
        dst = sdscatlen(dst,"\r\n",2);
        decrRefCount(o);
    }
    return dst;
}

sds catAppendOnlyExpireAtCommand(sds buf, robj *key, robj *seconds) {
    int argc = 3;
    time_t when;
    robj *argv[3];

    /* Make sure we can use strtol */
    seconds = getDecodedObject(seconds);
    when = time(NULL)+strtol(seconds->ptr,NULL,10);
    decrRefCount(seconds);

    argv[0] = createStringObject("EXPIREAT",8);
    argv[1] = key;
#ifdef _WIN32
    argv[2] = createObject(REDIS_STRING,
        sdscatprintf(sdsempty(),"%lld",(long long)when));
#else
    argv[2] = createObject(REDIS_STRING,
        sdscatprintf(sdsempty(),"%ld",when));
#endif
    buf = catAppendOnlyGenericCommand(buf, argc, argv);
    decrRefCount(argv[0]);
    decrRefCount(argv[2]);
    return buf;
}

void feedAppendOnlyFile(struct redisCommand *cmd, int dictid, robj **argv, int argc) {
    sds buf = sdsempty();
    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;
    }

    if (cmd->proc == expireCommand) {
        /* Translate EXPIRE into EXPIREAT */
        buf = catAppendOnlyExpireAtCommand(buf,argv[1],argv[2]);
    } else if (cmd->proc == setexCommand) {
        /* Translate SETEX to SET and EXPIREAT */
        tmpargv[0] = createStringObject("SET",3);
        tmpargv[1] = argv[1];
        tmpargv[2] = argv[3];
        buf = catAppendOnlyGenericCommand(buf,3,tmpargv);
        decrRefCount(tmpargv[0]);
        buf = catAppendOnlyExpireAtCommand(buf,argv[1],argv[2]);
    } else {
        buf = catAppendOnlyGenericCommand(buf,argc,argv);
    }

    /* Append to the AOF buffer. This will be flushed on disk just before
     * of re-entering the event loop, so before the client will get a
     * positive reply about the operation performed. */
    server.aofbuf = sdscatlen(server.aofbuf,buf,sdslen(buf));

    /* 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);
}

/* 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. */
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->bufpos = 0;
    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();
    c->watched_keys = listCreate();
    listSetFreeMethod(c->reply,decrRefCount);
    listSetDupMethod(c->reply,dupClientReplyValue);
    initClientMultiState(c);
    return c;
}

void freeFakeClient(struct redisClient *c) {
    sdsfree(c->querybuf);
    listRelease(c->reply);
    listRelease(c->watched_keys);
    freeClientMultiState(c);
    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;
#ifdef _WIN32
    FILE *fp = fopen(filename,"rb");
#else
    FILE *fp = fopen(filename,"r");
#endif
    struct redis_stat sb;
    int appendonly = server.appendonly;
#ifdef _WIN64
    long long loops = 0;
#else
    long loops = 0;
#endif

    if (fp && redis_fstat(fileno(fp),&sb) != -1 && sb.st_size == 0) {
        server.appendonly_current_size = 0;
        fclose(fp);
        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);
    }

    /* Temporarily disable AOF, to prevent EXEC from feeding a MULTI
     * to the same file we're about to read. */
    server.appendonly = 0;

    fakeClient = createFakeClient();
    startLoading(fp);

    while(1) {
        int argc, j;
#ifdef _WIN32
        size_t len;
#else
        unsigned long len;
#endif
        robj **argv;
        char buf[128];
        sds argsds;
        struct redisCommand *cmd;
        int force_swapout;

        /* Serve the clients from time to time */
#ifdef _WIN64
        if (!(loops++ % (long long)1000)) {
            loadingProgress(ftello(fp));
            aeProcessEvents(server.el, AE_FILE_EVENTS|AE_DONT_WAIT);
        }
#else
        if (!(loops++ % 1000)) {
            loadingProgress(ftello(fp));
            aeProcessEvents(server.el, AE_FILE_EVENTS|AE_DONT_WAIT);
        }
#endif

        if (fgets(buf,sizeof(buf),fp) == NULL) {
            if (feof(fp))
                break;
            else
                goto readerr;
        }
        if (buf[0] != '*') goto fmterr;
        argc = atoi(buf+1);
        if (argc < 1) goto fmterr;

        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);
        }
        /* Run the command in the context of a fake client */
        fakeClient->argc = argc;
        fakeClient->argv = argv;
        cmd->proc(fakeClient);

        /* The fake client should not have a reply */
        redisAssert(fakeClient->bufpos == 0 && listLength(fakeClient->reply) == 0);
        /* The fake client should never get blocked */
        redisAssert((fakeClient->flags & REDIS_BLOCKED) == 0);

        /* Clean up. Command code may have changed argv/argc so we use the
         * argv/argc of the client instead of the local variables. */
        for (j = 0; j < fakeClient->argc; j++)
            decrRefCount(fakeClient->argv[j]);
        zfree(fakeClient->argv);

        /* Handle swapping while loading big datasets when VM is on */
        force_swapout = 0;
        if ((zmalloc_used_memory() - server.vm_max_memory) > 1024*1024*32)
            force_swapout = 1;

        if (server.vm_enabled && force_swapout) {
            while (zmalloc_used_memory() > server.vm_max_memory) {
                if (vmSwapOneObjectBlocking() == REDIS_ERR) break;
            }
        }
    }

    /* This point can only be reached when EOF is reached without errors.
     * If the client is in the middle of a MULTI/EXEC, log error and quit. */
    if (fakeClient->flags & REDIS_MULTI) goto readerr;

    fclose(fp);
    freeFakeClient(fakeClient);
    server.appendonly = appendonly;
    stopLoading();
    aofUpdateCurrentSize();
    server.auto_aofrewrite_base_size = server.appendonly_current_size;
    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: make a backup of your AOF file, then use ./redis-check-aof --fix <filename>");
    exit(1);
}

/* Write a sequence of commands able to fully rebuild the dataset into
 * "filename". Used both by REWRITEAOF and BGREWRITEAOF. */
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. */
#ifdef _WIN32
    snprintf(tmpfile,256,"temp-rewriteaof-%lld.aof", (long long int) getpid());
    fp = fopen(tmpfile,"wb");
#else
    snprintf(tmpfile,256,"temp-rewriteaof-%d.aof", (int) getpid());
    fp = fopen(tmpfile,"w");
#endif
    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 = dictGetSafeIterator(d);
        if (!di) {
            fclose(fp);
            return REDIS_ERR;
        }

        /* SELECT the new DB */
        if (fwrite(selectcmd,sizeof(selectcmd)-1,1,fp) == 0) goto werr;
        if (fwriteBulkLongLong(fp,j) == 0) goto werr;

        /* Iterate this DB writing every entry */
        while((de = dictNext(di)) != NULL) {
            sds keystr = dictGetEntryKey(de);
            robj key, *o;
            time_t expiretime;
            int swapped;

            keystr = dictGetEntryKey(de);
            o = dictGetEntryVal(de);
            initStaticStringObject(key,keystr);
            /* 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 || o->storage == REDIS_VM_MEMORY ||
                o->storage == REDIS_VM_SWAPPING) {
                swapped = 0;
            } else {
                o = vmPreviewObject(o);
                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 (fwriteBulkObject(fp,&key) == 0) goto werr;
                if (fwriteBulkObject(fp,o) == 0) goto werr;
            } else if (o->type == REDIS_LIST) {
                /* Emit the RPUSHes needed to rebuild the list */
                char cmd[]="*3\r\n$5\r\nRPUSH\r\n";
                if (o->encoding == REDIS_ENCODING_ZIPLIST) {
                    unsigned char *zl = o->ptr;
                    unsigned char *p = ziplistIndex(zl,0);
                    unsigned char *vstr;
                    unsigned int vlen;
                    long long vlong;

                    while(ziplistGet(p,&vstr,&vlen,&vlong)) {
                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (vstr) {
                            if (fwriteBulkString(fp,(char*)vstr,vlen) == 0)
                                goto werr;
                        } else {
                            if (fwriteBulkLongLong(fp,vlong) == 0)
                                goto werr;
                        }
                        p = ziplistNext(zl,p);
                    }
                } else if (o->encoding == REDIS_ENCODING_LINKEDLIST) {
                    list *list = o->ptr;
                    listNode *ln;
                    listIter li;

                    listRewind(list,&li);
                    while((ln = listNext(&li))) {
                        robj *eleobj = listNodeValue(ln);

                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkObject(fp,eleobj) == 0) goto werr;
                    }
                } else {
                    redisPanic("Unknown list encoding");
                }
            } else if (o->type == REDIS_SET) {
                char cmd[]="*3\r\n$4\r\nSADD\r\n";

                /* Emit the SADDs needed to rebuild the set */
                if (o->encoding == REDIS_ENCODING_INTSET) {
                    int ii = 0;
                    int64_t llval;
                    while(intsetGet(o->ptr,ii++,&llval)) {
                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkLongLong(fp,llval) == 0) goto werr;
                    }
                } else if (o->encoding == REDIS_ENCODING_HT) {
                    dictIterator *di = dictGetIterator(o->ptr);
                    dictEntry *de;
                    while((de = dictNext(di)) != NULL) {
                        robj *eleobj = dictGetEntryKey(de);
                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkObject(fp,eleobj) == 0) goto werr;
                    }
                    dictReleaseIterator(di);
                } else {
                    redisPanic("Unknown set encoding");
                }
            } else if (o->type == REDIS_ZSET) {
                /* Emit the ZADDs needed to rebuild the sorted set */
                char cmd[]="*4\r\n$4\r\nZADD\r\n";

                if (o->encoding == REDIS_ENCODING_ZIPLIST) {
                    unsigned char *zl = o->ptr;
                    unsigned char *eptr, *sptr;
                    unsigned char *vstr;
                    unsigned int vlen;
                    long long vll;
                    double score;

                    eptr = ziplistIndex(zl,0);
                    redisAssert(eptr != NULL);
                    sptr = ziplistNext(zl,eptr);
                    redisAssert(sptr != NULL);

                    while (eptr != NULL) {
                        redisAssert(ziplistGet(eptr,&vstr,&vlen,&vll));
                        score = zzlGetScore(sptr);

                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkDouble(fp,score) == 0) goto werr;
                        if (vstr != NULL) {
                            if (fwriteBulkString(fp,(char*)vstr,vlen) == 0)
                                goto werr;
                        } else {
                            if (fwriteBulkLongLong(fp,vll) == 0)
                                goto werr;
                        }
                        zzlNext(zl,&eptr,&sptr);
                    }
                } else if (o->encoding == REDIS_ENCODING_SKIPLIST) {
                    zset *zs = o->ptr;
                    dictIterator *di = dictGetIterator(zs->dict);
                    dictEntry *de;

                    while((de = dictNext(di)) != NULL) {
                        robj *eleobj = dictGetEntryKey(de);
                        double *score = dictGetEntryVal(de);

                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkDouble(fp,*score) == 0) goto werr;
                        if (fwriteBulkObject(fp,eleobj) == 0) goto werr;
                    }
                    dictReleaseIterator(di);
                } else {
                    redisPanic("Unknown sorted set encoding");
                }
            } else if (o->type == REDIS_HASH) {
                char cmd[]="*4\r\n$4\r\nHSET\r\n";

                /* Emit the HSETs needed to rebuild the hash */
                if (o->encoding == REDIS_ENCODING_ZIPMAP) {
                    unsigned char *p = zipmapRewind(o->ptr);
                    unsigned char *field, *val;
                    unsigned int flen, vlen;

                    while((p = zipmapNext(p,&field,&flen,&val,&vlen)) != NULL) {
                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkString(fp,(char*)field,flen) == 0)
                            goto werr;
                        if (fwriteBulkString(fp,(char*)val,vlen) == 0)
                            goto werr;
                    }
                } else {
                    dictIterator *di = dictGetIterator(o->ptr);
                    dictEntry *de;

                    while((de = dictNext(di)) != NULL) {
                        robj *field = dictGetEntryKey(de);
                        robj *val = dictGetEntryVal(de);

                        if (fwrite(cmd,sizeof(cmd)-1,1,fp) == 0) goto werr;
                        if (fwriteBulkObject(fp,&key) == 0) goto werr;
                        if (fwriteBulkObject(fp,field) == 0) goto werr;
                        if (fwriteBulkObject(fp,val) == 0) goto werr;
                    }
                    dictReleaseIterator(di);
                }
            } else {
                redisPanic("Unknown object type");
            }
            /* 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 (fwriteBulkObject(fp,&key) == 0) goto werr;
                if (fwriteBulkLongLong(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);
    aof_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;
}

/* This is how rewriting of the append only file in background works:
 *
 * 1) The user calls BGREWRITEAOF
 * 2) Redis calls this function, that forks():
 *    2a) the child rewrite the append only file in a temp file.
 *    2b) the parent accumulates differences in server.bgrewritebuf.
 * 3) When the child finished '2a' exists.
 * 4) The parent will trap the exit code, if it's OK, will append the
 *    data accumulated into server.bgrewritebuf into the temp file, and
 *    finally will rename(2) the temp file in the actual file name.
 *    The the new file is reopened as the new append only file. Profit!
 */
int rewriteAppendOnlyFileBackground(void) {
    pid_t childpid;
    long long start;

    if (server.bgrewritechildpid != -1) return REDIS_ERR;
    if (server.vm_enabled) waitEmptyIOJobsQueue();
    start = ustime();
    if ((childpid = fork()) == 0) {
        char tmpfile[256];

        /* Child */
        if (server.vm_enabled) vmReopenSwapFile();
        if (server.ipfd > 0) close(server.ipfd);
        if (server.sofd > 0) close(server.sofd);
        snprintf(tmpfile,256,"temp-rewriteaof-bg-%d.aof", (int) getpid());
        if (rewriteAppendOnlyFile(tmpfile) == REDIS_OK) {
            _exit(0);
        } else {
            _exit(1);
        }
    } else {
        /* Parent */
        server.stat_fork_time = ustime()-start;
#ifdef _WIN32
        if (childpid == -1) {
            char tmpfile[256];

            childpid = getpid();
            snprintf(tmpfile,256,"temp-rewriteaof-bg-%lld.aof", (long long)childpid);
            server.bgrewritechildpid = childpid;
            updateDictResizePolicy();
            server.appendseldb = -1;

            redisLog(REDIS_NOTICE,
                "Foreground append only file rewriting started by pid %lld",(long long)childpid);

            if (rewriteAppendOnlyFile(tmpfile) == REDIS_OK) {
                backgroundRewriteDoneHandler(0);
                return REDIS_OK;
            } else {
                backgroundRewriteDoneHandler(0xff);
                redisLog(REDIS_WARNING,
                    "Can't rewrite append only file in background: spoon: %s",
                    strerror(errno));
                return REDIS_ERR;
            }
        }
#else
        if (childpid == -1) {
            redisLog(REDIS_WARNING,
                "Can't rewrite append only file in background: fork: %s",
                strerror(errno));
            return REDIS_ERR;
        }
        redisLog(REDIS_NOTICE,
            "Background append only file rewriting started by pid %d",childpid);
        server.aofrewrite_scheduled = 0;
        server.bgrewritechildpid = childpid;
#endif
        updateDictResizePolicy();
        /* We set appendseldb to -1 in order to force the next call to the
         * feedAppendOnlyFile() to issue a SELECT command, so the differences
         * accumulated by the parent into server.bgrewritebuf will start
         * with a SELECT statement and it will be safe to merge. */
        server.appendseldb = -1;
        return REDIS_OK;
    }
    return REDIS_OK; /* unreached */
}

void bgrewriteaofCommand(redisClient *c) {
    if (server.bgrewritechildpid != -1) {
        addReplyError(c,"Background append only file rewriting already in progress");
    } else if (server.bgsavechildpid != -1) {
        server.aofrewrite_scheduled = 1;
        addReplyStatus(c,"Background append only file rewriting scheduled");
    } else if (rewriteAppendOnlyFileBackground() == REDIS_OK) {
        addReplyStatus(c,"Background append only file rewriting started");
    } else {
        addReply(c,shared.err);
    }
}

void aofRemoveTempFile(pid_t childpid) {
    char tmpfile[256];

    snprintf(tmpfile,256,"temp-rewriteaof-bg-%d.aof", (int) childpid);
    unlink(tmpfile);
}

/* Update the server.appendonly_current_size filed explicitly using stat(2)
 * to check the size of the file. This is useful after a rewrite or after
 * a restart, normally the size is updated just adding the write length
 * to the current lenght, that is much faster. */
void aofUpdateCurrentSize(void) {
    struct redis_stat sb;

    if (redis_fstat(server.appendfd,&sb) == -1) {
        redisLog(REDIS_WARNING,"Unable to check the AOF length: %s",
            strerror(errno));
    } else {
        server.appendonly_current_size = sb.st_size;
    }
}

/* A background append only file rewriting (BGREWRITEAOF) terminated its work.
 * Handle this. */
void backgroundRewriteDoneHandler(int statloc) {
    int exitcode = WEXITSTATUS(statloc);
    int bysignal = WIFSIGNALED(statloc);

    if (!bysignal && exitcode == 0) {
        int newfd, oldfd;
        int nwritten;
        char tmpfile[256];

        redisLog(REDIS_NOTICE,
            "Background AOF rewrite terminated with success");

        /* Flush the differences accumulated by the parent to the
         * rewritten AOF. */
        snprintf(tmpfile,256,"temp-rewriteaof-bg-%d.aof",
            (int)server.bgrewritechildpid);
#ifdef _WIN32
        newfd = open(tmpfile,O_WRONLY|O_APPEND|O_CREAT|_O_BINARY,_S_IREAD|_S_IWRITE);
#else
        newfd = open(tmpfile,O_WRONLY|O_APPEND);
#endif
        if (newfd == -1) {
#ifdef _WIN32
            /* Windows fix: More info */
            redisLog(REDIS_WARNING, 
                "Not able to open the temp append only file (%s) produced by the child: %s", tmpfile, strerror(errno));
#else
            redisLog(REDIS_WARNING,
                "Unable to open the temporary AOF produced by the child: %s", strerror(errno));
#endif
            goto cleanup;
        }

        nwritten = write(newfd,server.bgrewritebuf,sdslen(server.bgrewritebuf));
        if (nwritten != (signed)sdslen(server.bgrewritebuf)) {
            if (nwritten == -1) {
                redisLog(REDIS_WARNING,
                    "Error trying to flush the parent diff to the rewritten AOF: %s", strerror(errno));
            } else {
                redisLog(REDIS_WARNING,
                    "Short write trying to flush the parent diff to the rewritten AOF: %s", strerror(errno));
            }
            close(newfd);
            goto cleanup;
        }

        redisLog(REDIS_NOTICE,
            "Parent diff successfully flushed to the rewritten AOF (%lu bytes)", nwritten);
#ifdef _WIN32
        /* Close files before renaming */
        close(newfd);
        if (server.appendfd != -1) close(server.appendfd);
#endif

        /* The only remaining thing to do is to rename the temporary file to
         * the configured file and switch the file descriptor used to do AOF
         * writes. We don't want close(2) or rename(2) calls to block the
         * server on old file deletion.
         *
         * There are two possible scenarios:
         *
         * 1) AOF is DISABLED and this was a one time rewrite. The temporary
         * file will be renamed to the configured file. When this file already
         * exists, it will be unlinked, which may block the server.
         *
         * 2) AOF is ENABLED and the rewritten AOF will immediately start
         * receiving writes. After the temporary file is renamed to the
         * configured file, the original AOF file descriptor will be closed.
         * Since this will be the last reference to that file, closing it
         * causes the underlying file to be unlinked, which may block the
         * server.
         *
         * To mitigate the blocking effect of the unlink operation (either
         * caused by rename(2) in scenario 1, or by close(2) in scenario 2), we
         * use a background thread to take care of this. First, we
         * make scenario 1 identical to scenario 2 by opening the target file
         * when it exists. The unlink operation after the rename(2) will then
         * be executed upon calling close(2) for its descriptor. Everything to
         * guarantee atomicity for this switch has already happened by then, so
         * we don't care what the outcome or duration of that close operation
         * is, as long as the file descriptor is released again. */
        if (server.appendfd == -1) {
            /* AOF disabled */

             /* Don't care if this fails: oldfd will be -1 and we handle that.
              * One notable case of -1 return is if the old file does
              * not exist. */
#ifdef _WIN32
             oldfd = open(server.appendfilename,O_RDONLY|_O_BINARY,0);
#else
             oldfd = open(server.appendfilename,O_RDONLY|O_NONBLOCK);
#endif
        } else {
            /* AOF enabled */
            oldfd = -1; /* We'll set this to the current AOF filedes later. */
        }

        /* Rename the temporary file. This will not unlink the target file if
         * it exists, because we reference it with "oldfd". */
        if (rename(tmpfile,server.appendfilename) == -1) {
            redisLog(REDIS_WARNING,
                "Error trying to rename the temporary AOF: %s", strerror(errno));
#ifndef _WIN32
            if (oldfd != -1) close(oldfd);
#endif
            goto cleanup;
        }

        if (server.appendfd == -1) {
            /* AOF disabled, we don't need to set the AOF file descriptor
             * to this new file, so we can close it. */
            close(newfd);
        } else {
            /* AOF enabled, replace the old fd with the new one. */
            oldfd = server.appendfd;
            server.appendfd = newfd;
            if (server.appendfsync == APPENDFSYNC_ALWAYS)
                aof_fsync(newfd);
            else if (server.appendfsync == APPENDFSYNC_EVERYSEC)
                aof_background_fsync(newfd);
            server.appendseldb = -1; /* Make sure SELECT is re-issued */
            aofUpdateCurrentSize();
            server.auto_aofrewrite_base_size = server.appendonly_current_size;            
            /* Clear regular AOF buffer since its contents was just written to
             * the new AOF from the background rewrite buffer. */
            sdsfree(server.aofbuf);
            server.aofbuf = sdsempty();
        }

        redisLog(REDIS_NOTICE, "Background AOF rewrite successful");

        /* Asynchronously close the overwritten AOF. */
#ifdef _WIN32
        if (oldfd != -1) bioCreateBackgroundJob(REDIS_BIO_CLOSE_FILE,(void*)(size_t)oldfd,NULL,NULL);
#else        
        if (oldfd != -1) bioCreateBackgroundJob(REDIS_BIO_CLOSE_FILE,(void*)(long)oldfd,NULL,NULL);
#endif

    } else if (!bysignal && exitcode != 0) {
        redisLog(REDIS_WARNING,
            "Background AOF rewrite terminated with error");
    } else {
        redisLog(REDIS_WARNING,
            "Background AOF rewrite terminated by signal %d",
            WTERMSIG(statloc));
    }

cleanup:
    sdsfree(server.bgrewritebuf);
    server.bgrewritebuf = sdsempty();
    aofRemoveTempFile(server.bgrewritechildpid);
    server.bgrewritechildpid = -1;
}