static int createAllCFSfromConfig()
{
dictEntry *de = NULL;
dictIterator *di = NULL;
if(!g_dataModel.cfsHashTable)
return -1;
di = dictGetIterator(g_cfmDict);
if(!di){
return -1;
}
de = dictNext(di);
while(de){
columnFamilyMetadata *cfmd = (columnFamilyMetadata *)
dictGetEntryKey(de);
columnFamilyStore *cfs = NULL;
if(!cfmd || !(cfs = getCFStore(cfmd))){
dictReleaseIterator(di);
return -1;
}
if(insertDMCFS(cfs) < 0){
dictReleaseIterator(di);
freeHeapCFStore(cfs);
return -1;
}
de = dictNext(di);
}
dictReleaseIterator(di);
return 0;
}
void dbPrintSets(FILE *f)
{
dictIterator *iter = NULL;
dictEntry *entry = NULL;
if (NULL == f)
return;
lockRead(sets);
if (0 == dictSize(sets))
{
unlockRead(sets);
fprintf(f, "No sets in db.\r\n");
return;
}
if (NULL == (iter = dictGetIterator(sets)))
{
unlockRead(sets);
return;
}
while (NULL != (entry = dictNext(iter)))
{
fprintf(f, "%s\r\n", dictGetEntryKey(entry));
lockRead(dictGetEntryVal(entry));
setPrint((set *) dictGetEntryVal(entry), f, 1);
unlockRead(dictGetEntryVal(entry));
fprintf(f, "\r\n");
}
dictReleaseIterator(iter);
unlockRead(sets);
}
void keysCommand(redisClient *c) {
dictIterator *di;
dictEntry *de;
sds pattern = c->argv[1]->ptr;
int plen = sdslen(pattern);
unsigned long numkeys = 0;
robj *lenobj = createObject(REDIS_STRING,NULL);
di = dictGetIterator(c->db->dict);
addReply(c,lenobj);
decrRefCount(lenobj);
while((de = dictNext(di)) != NULL) {
sds key = dictGetEntryKey(de);
robj *keyobj;
if ((pattern[0] == '*' && pattern[1] == '\0') ||
stringmatchlen(pattern,plen,key,sdslen(key),0)) {
keyobj = createStringObject(key,sdslen(key));
if (expireIfNeeded(c->db,keyobj) == 0) {
addReplyBulk(c,keyobj);
numkeys++;
}
decrRefCount(keyobj);
}
}
dictReleaseIterator(di);
lenobj->ptr = sdscatprintf(sdsempty(),"*%lu\r\n",numkeys);
}
void keysCommand(redisClient *c) {
dictIterator *di;
dictEntry *de;
sds pattern = c->argv[1]->ptr;
int plen = sdslen(pattern), allkeys;
unsigned long numkeys = 0;
void *replylen = addDeferredMultiBulkLength(c);
di = dictGetIterator(c->db->dict);
allkeys = (pattern[0] == '*' && pattern[1] == '\0');
while((de = dictNext(di)) != NULL) {
sds key = dictGetEntryKey(de);
robj *keyobj;
if (allkeys || stringmatchlen(pattern,plen,key,sdslen(key),0)) {
keyobj = createStringObject(key,sdslen(key));
if (expireIfNeeded(c->db,keyobj) == 0) {
addReplyBulk(c,keyobj);
numkeys++;
}
decrRefCount(keyobj);
}
}
dictReleaseIterator(di);
setDeferredMultiBulkLength(c,replylen,numkeys);
}
void delkeysCommand(redisClient *c) {
dictIterator *di;
dictEntry *de;
sds pattern = c->argv[1]->ptr;
int plen = sdslen(pattern), allkeys;
unsigned long deleted = 0;
di = dictGetIterator(c->db->dict);
allkeys = (pattern[0] == '*' && pattern[1] == '\0');
while((de = dictNext(di)) != NULL) {
sds key = dictGetEntryKey(de);
robj *keyobj;
if (allkeys || stringmatchlen(pattern,plen,key,sdslen(key),0)) {
keyobj = createStringObject(key,sdslen(key));
if (dbDelete(c->db,keyobj)) {
touchWatchedKey(c->db,keyobj);
server.dirty++;
deleted++;
}
decrRefCount(keyobj);
}
}
dictReleaseIterator(di);
addReplyLongLong(c,deleted);
}
static robj *lookupKey(redisDb *db, robj *key) {
dictEntry *de = dictFind(db->dict,key);
if (de) {
robj *key = dictGetEntryKey(de);
robj *val = dictGetEntryVal(de);
if (server.vm_enabled) {
if (key->storage == REDIS_VM_MEMORY ||
key->storage == REDIS_VM_SWAPPING)
{
/* If we were swapping the object out, stop it, this key
* was requested. */
if (key->storage == REDIS_VM_SWAPPING)
vmCancelThreadedIOJob(key);
/* Update the access time of the key for the aging algorithm. */
key->vm.atime = server.unixtime;
} else {
int notify = (key->storage == REDIS_VM_LOADING);
/* Our value was swapped on disk. Bring it at home. */
redisAssert(val == NULL);
val = vmLoadObject(key);
dictGetEntryVal(de) = val;
/* Clients blocked by the VM subsystem may be waiting for
* this key... */
if (notify) handleClientsBlockedOnSwappedKey(db,key);
}
}
return val;
} else {
return NULL;
}
}
void setExpire(redisDb *db, robj *key, time_t when) {
dictEntry *de;
/* Reuse the sds from the main dict in the expire dict */
de = dictFind(db->dict,key->ptr);
redisAssert(de != NULL);
dictReplace(db->expires,dictGetEntryKey(de),(void*)when);
}
static void freeDictOfIndRow(dict *d, int num_cols, bool is_ob) {
dictEntry *ide;
dictIterator *idi = dictGetIterator(d);
while((ide = dictNext(idi)) != NULL) {
robj *ikey = dictGetEntryKey(ide);
freeIndRow(ikey->ptr, num_cols, is_ob, 0);
}
dictReleaseIterator(idi);
}
sds _masterGetStatus() {
/*TODO: calculate cache increase speed,
* then adopt a suitable stale-cache freeing strategy
* Three involved params:
* (1) master sleep,
* (2) ae loop wait,
* and (3) number of stale entries in one ae loop
*/
sds status = sdsempty();//sdsfromlonglong(master_total_mem);
status = sdscatprintf(status,"TOL RAM: %-6.2lfMB\tUSED RAM: %-6.2lf\n",
BYTES_TO_MEGABYTES(MASTER_MAX_AVAIL_MEM),
BYTES_TO_MEGABYTES(master_total_mem));
#if (CCACHE_LOG_LEVEL == CCACHE_DEBUG)
status = sdscatprintf(status,"Detail:\n");
status = sdscatprintf(status,"%-3s %-32s: %-6s\n"," ","KEY","MEM");
dictIterator *di = dictGetIterator(master_cache);
dictEntry *de;
int idx = 1;
while((de = dictNext(di)) != NULL) {
objSds *value = (objSds*)dictGetEntryVal(de);
if(value) {
if(value->ptr) {
status = sdscatprintf(status,"%-3d %-32s: %-6ld\n",
idx++,
(char*)dictGetEntryKey(de),
sdslen(value->ptr));
}
else {
status = sdscatprintf(status,"%-3d %-32s: %-6s\n",
idx++,
(char*)dictGetEntryKey(de),
"WAITING");
}
}
}
dictReleaseIterator(di);
#endif
sds status_reply = sdsnew("HTTP/1.1 200 OK\r\n");
status_reply = sdscatprintf(status_reply,"Content-Length: %ld\r\n\r\n%s",sdslen(status),status);
sdsfree(status);
return status_reply;
}
/* Return random element from a non empty set.
* The returned element can be a int64_t value if the set is encoded
* as an "intset" blob of integers, or a redis object if the set
* is a regular set.
*
* The caller provides both pointers to be populated with the right
* object. The return value of the function is the object->encoding
* field of the object and is used by the caller to check if the
* int64_t pointer or the redis object pointere was populated.
*
* When an object is returned (the set was a real set) the ref count
* of the object is not incremented so this function can be considered
* copy on write friendly. */
int setTypeRandomElement(robj *setobj, robj **objele, int64_t *llele) {
if (setobj->encoding == REDIS_ENCODING_HT) {
dictEntry *de = dictGetRandomKey(setobj->ptr);
*objele = dictGetEntryKey(de);
} else if (setobj->encoding == REDIS_ENCODING_INTSET) {
*llele = intsetRandom(setobj->ptr);
} else {
redisPanic("Unknown set encoding");
}
return setobj->encoding;
}
/* Move to the next entry in the set. Returns the object at the current
* position.
*
* Since set elements can be internally be stored as redis objects or
* simple arrays of integers, setTypeNext returns the encoding of the
* set object you are iterating, and will populate the appropriate pointer
* (eobj) or (llobj) accordingly.
*
* When there are no longer elements -1 is returned.
* Returned objects ref count is not incremented, so this function is
* copy on write friendly. */
int setTypeNext(setTypeIterator *si, robj **objele, int64_t *llele) {
if (si->encoding == REDIS_ENCODING_HT) {
dictEntry *de = dictNext(si->di);
if (de == NULL) return -1;
*objele = dictGetEntryKey(de);
} else if (si->encoding == REDIS_ENCODING_INTSET) {
if (!intsetGet(si->subject->ptr,si->ii++,llele))
return -1;
}
return si->encoding;
}
void movekeysCommand(redisClient *c) {
redisDb *src, *dst;
int srcid;
dictIterator *di;
dictEntry *de;
sds pattern = c->argv[1]->ptr;
int plen = sdslen(pattern), allkeys;
unsigned long numkeys = 0;
/* Obtain source and target DB pointers */
src = c->db;
srcid = c->db->id;
if (selectDb(c,atoi(c->argv[2]->ptr)) == REDIS_ERR) {
addReply(c,shared.outofrangeerr);
return;
}
dst = c->db;
selectDb(c,srcid); /* Back to the source DB */
/* If the user is moving using as target the same
* DB as the source DB it is probably an error. */
if (src == dst) {
addReply(c,shared.sameobjecterr);
return;
}
di = dictGetIterator(c->db->dict);
allkeys = (pattern[0] == '*' && pattern[1] == '\0');
while((de = dictNext(di)) != NULL) {
sds key = dictGetEntryKey(de);
robj *keyobj;
if (allkeys || stringmatchlen(pattern,plen,key,sdslen(key),0)) {
keyobj = createStringObject(key,sdslen(key));
if (expireIfNeeded(c->db,keyobj) == 0) {
robj *val = dictGetEntryVal(de);
/* Try to add the element to the target DB */
if (dbAdd(dst,keyobj,val) != REDIS_ERR) {
incrRefCount(val);
/* OK! key moved, free the entry in the source DB */
dbDelete(src,keyobj);
server.dirty++;
numkeys++;
}
}
decrRefCount(keyobj);
}
}
dictReleaseIterator(di);
addReplyLongLong(c,numkeys);
}
/* Return the logic clock of the specified key, or 0 if no key exist */
uint16_t getLogiClock(redisDb *db, robj *key) {
uint16_t logiclock;
dictEntry *de;
de = dictFind(db->dict,key->ptr);
if (de == NULL) {
return 0;
}
sds skey = (sds) dictGetEntryKey(de);
logiclock = sdslogiclock(skey);
redisAssert(logiclock != 0);
return logiclock;
}
/* Unsubscribe from all the channels. Return the number of channels the
* client was subscribed from. */
int pubsubUnsubscribeAllChannels(redisClient *c, int notify) {
dictIterator *di = dictGetSafeIterator(c->pubsub_channels);
dictEntry *de;
int count = 0;
while((de = dictNext(di)) != NULL) {
robj *channel = dictGetEntryKey(de);
count += pubsubUnsubscribeChannel(c,channel,notify);
}
dictReleaseIterator(di);
return count;
}
robj *lookupKeyWithVersion(redisDb *db, robj *key, uint16_t* version) {
dictEntry *de = dictFind(db->dict, key->ptr);
if(de) {
robj *val = dictGetEntryVal(de);
sds key_tmp = dictGetEntryKey(de);
*version = sdsversion(key_tmp);
val->lru = shared.lruclock;
db->stat_keyspace_hits++;
return val;
} else {
db->stat_keyspace_misses++;
return NULL;
}
}
void expirekeysGenericCommand(redisClient *c, robj *keypattern, robj *param, long offset) {
dictIterator *di;
dictEntry *de;
sds pattern = keypattern->ptr;
int plen = sdslen(pattern);
unsigned long numkeys = 0, allkeys;
time_t seconds;
time_t when;
if (getLongFromObjectOrReply(c, param, &seconds, NULL) != REDIS_OK) return;
seconds -= offset;
when = time(NULL)+seconds;
di = dictGetIterator(c->db->dict);
allkeys = (pattern[0] == '*' && pattern[1] == '\0');
while((de = dictNext(di)) != NULL) {
sds key = dictGetEntryKey(de);
robj *keyobj;
if (allkeys || stringmatchlen(pattern,plen,key,sdslen(key),0)) {
keyobj = createStringObject(key,sdslen(key));
if (seconds <= 0 && !server.loading && !server.masterhost) {
robj *aux;
redisAssert(dbDelete(c->db,keyobj));
server.dirty++;
numkeys++;
/* Replicate/AOF this as an explicit DEL. */
aux = createStringObject("DEL",3);
rewriteClientCommandVector(c,2,aux,keyobj);
decrRefCount(aux);
touchWatchedKey(c->db,keyobj);
} else {
time_t when = time(NULL)+seconds;
setExpire(c->db,keyobj,when);
touchWatchedKey(c->db,keyobj);
server.dirty++;
numkeys++;
}
decrRefCount(keyobj);
}
}
dictReleaseIterator(di);
addReplyLongLong(c,numkeys);
}
columnFamilyStore *findDMCFS(char *keyspaceName, char *cfName)
{
columnFamilyStore cfstmp = {0};
columnFamilyMetadata tm = {keyspaceName, cfName};
dictEntry *de = NULL;
if(!keyspaceName || !cfName || !g_dataModel.cfsHashTable){
return NULL;
}
cfstmp.cfmd = &tm;
if(!(de = dictFind(g_dataModel.cfsHashTable, &cfstmp))){
return NULL;
}
return (columnFamilyStore *)dictGetEntryKey(de);
}
/* Try to share an object against the shared objects pool */
static robj *tryObjectSharing(robj *o) {
struct dictEntry *de;
unsigned long c;
if (o == NULL || server.shareobjects == 0) return o;
de = dictFind(server.sharingpool,o);
if (de) {
robj *shared = dictGetEntryKey(de);
c = ((unsigned long) dictGetEntryVal(de))+1;
dictGetEntryVal(de) = (void*) c;
incrRefCount(shared);
decrRefCount(o);
return shared;
} else {
/* Here we are using a stream algorihtm: Every time an object is
* shared we increment its count, everytime there is a miss we
* recrement the counter of a random object. If this object reaches
* zero we remove the object and put the current object instead. */
if (dictSize(server.sharingpool) >=
server.sharingpoolsize) {
de = dictGetRandomKey(server.sharingpool);
c = ((unsigned long) dictGetEntryVal(de))-1;
dictGetEntryVal(de) = (void*) c;
if (c == 0) {
dictDelete(server.sharingpool,de->key);
}
} else {
c = 0; /* If the pool is empty we want to add this object */
}
if (c == 0) {
int retval;
retval = dictAdd(server.sharingpool,o,(void*)1);
redisAssert(retval == DICT_OK);
incrRefCount(o);
}
return o;
}
}
/* Return a random key, in form of a Redis object.
* If there are no keys, NULL is returned.
*
* The function makes sure to return keys not already expired. */
robj *dbRandomKey(redisDb *db) {
struct dictEntry *de;
while(1) {
sds key;
robj *keyobj;
de = dictGetRandomKey(db->dict);
if (de == NULL) return NULL;
key = dictGetEntryKey(de);
keyobj = createStringObject(key,sdslen(key));
if (dictFind(db->expires,key)) {
if (expireIfNeeded(db,keyobj)) {
decrRefCount(keyobj);
continue; /* search for another key. This expired. */
}
}
return keyobj;
}
}
/* This generic command implements both ZADD and ZINCRBY. */
void zaddGenericCommand(redisClient *c, robj *key, robj *ele, double score, int incr) {
robj *zsetobj;
zset *zs;
zskiplistNode *znode;
zsetobj = lookupKeyWrite(c->db,key);
if (zsetobj == NULL) {
zsetobj = createZsetObject();
dbAdd(c->db,key,zsetobj);
} else {
if (zsetobj->type != REDIS_ZSET) {
addReply(c,shared.wrongtypeerr);
return;
}
}
zs = zsetobj->ptr;
/* Since both ZADD and ZINCRBY are implemented here, we need to increment
* the score first by the current score if ZINCRBY is called. */
if (incr) {
/* Read the old score. If the element was not present starts from 0 */
dictEntry *de = dictFind(zs->dict,ele);
if (de != NULL)
score += *(double*)dictGetEntryVal(de);
if (isnan(score)) {
addReplyError(c,"resulting score is not a number (NaN)");
/* Note that we don't need to check if the zset may be empty and
* should be removed here, as we can only obtain Nan as score if
* there was already an element in the sorted set. */
return;
}
}
/* We need to remove and re-insert the element when it was already present
* in the dictionary, to update the skiplist. Note that we delay adding a
* pointer to the score because we want to reference the score in the
* skiplist node. */
if (dictAdd(zs->dict,ele,NULL) == DICT_OK) {
dictEntry *de;
/* New element */
incrRefCount(ele); /* added to hash */
znode = zslInsert(zs->zsl,score,ele);
incrRefCount(ele); /* added to skiplist */
/* Update the score in the dict entry */
de = dictFind(zs->dict,ele);
redisAssert(de != NULL);
dictGetEntryVal(de) = &znode->score;
touchWatchedKey(c->db,c->argv[1]);
server.dirty++;
if (incr)
addReplyDouble(c,score);
else
addReply(c,shared.cone);
} else {
dictEntry *de;
robj *curobj;
double *curscore;
int deleted;
/* Update score */
de = dictFind(zs->dict,ele);
redisAssert(de != NULL);
curobj = dictGetEntryKey(de);
curscore = dictGetEntryVal(de);
/* When the score is updated, reuse the existing string object to
* prevent extra alloc/dealloc of strings on ZINCRBY. */
if (score != *curscore) {
deleted = zslDelete(zs->zsl,*curscore,curobj);
redisAssert(deleted != 0);
znode = zslInsert(zs->zsl,score,curobj);
incrRefCount(curobj);
/* Update the score in the current dict entry */
dictGetEntryVal(de) = &znode->score;
touchWatchedKey(c->db,c->argv[1]);
server.dirty++;
}
if (incr)
addReplyDouble(c,score);
else
addReply(c,shared.czero);
}
}
/* The SORT command is the most complex command in Redis. Warning: this code
* is optimized for speed and a bit less for readability */
void sortCommand(redisClient *c) {
list *operations;
unsigned int outputlen = 0;
int desc = 0, alpha = 0;
int limit_start = 0, limit_count = -1, start, end;
int j, dontsort = 0, vectorlen;
int getop = 0; /* GET operation counter */
robj *sortval, *sortby = NULL, *storekey = NULL;
redisSortObject *vector; /* Resulting vector to sort */
/* Lookup the key to sort. It must be of the right types */
sortval = lookupKeyRead(c->db,c->argv[1]);
if (sortval == NULL) {
addReply(c,shared.emptymultibulk);
return;
}
if (sortval->type != REDIS_SET && sortval->type != REDIS_LIST &&
sortval->type != REDIS_ZSET)
{
addReply(c,shared.wrongtypeerr);
return;
}
/* Create a list of operations to perform for every sorted element.
* Operations can be GET/DEL/INCR/DECR */
operations = listCreate();
listSetFreeMethod(operations,zfree);
j = 2;
/* Now we need to protect sortval incrementing its count, in the future
* SORT may have options able to overwrite/delete keys during the sorting
* and the sorted key itself may get destroied */
incrRefCount(sortval);
/* The SORT command has an SQL-alike syntax, parse it */
while(j < c->argc) {
int leftargs = c->argc-j-1;
if (!strcasecmp(c->argv[j]->ptr,"asc")) {
desc = 0;
} else if (!strcasecmp(c->argv[j]->ptr,"desc")) {
desc = 1;
} else if (!strcasecmp(c->argv[j]->ptr,"alpha")) {
alpha = 1;
} else if (!strcasecmp(c->argv[j]->ptr,"limit") && leftargs >= 2) {
limit_start = atoi(c->argv[j+1]->ptr);
limit_count = atoi(c->argv[j+2]->ptr);
j+=2;
} else if (!strcasecmp(c->argv[j]->ptr,"store") && leftargs >= 1) {
storekey = c->argv[j+1];
j++;
} else if (!strcasecmp(c->argv[j]->ptr,"by") && leftargs >= 1) {
sortby = c->argv[j+1];
/* If the BY pattern does not contain '*', i.e. it is constant,
* we don't need to sort nor to lookup the weight keys. */
if (strchr(c->argv[j+1]->ptr,'*') == NULL) dontsort = 1;
j++;
} else if (!strcasecmp(c->argv[j]->ptr,"get") && leftargs >= 1) {
listAddNodeTail(operations,createSortOperation(
REDIS_SORT_GET,c->argv[j+1]));
getop++;
j++;
} else {
decrRefCount(sortval);
listRelease(operations);
addReply(c,shared.syntaxerr);
return;
}
j++;
}
/* Load the sorting vector with all the objects to sort */
switch(sortval->type) {
case REDIS_LIST: vectorlen = (int)listTypeLength(sortval); break;
case REDIS_SET: vectorlen = (int)setTypeSize(sortval); break;
case REDIS_ZSET: vectorlen = (int)dictSize(((zset*)sortval->ptr)->dict); break;
default: vectorlen = 0; redisPanic("Bad SORT type"); /* Avoid GCC warning */
}
vector = zmalloc(sizeof(redisSortObject)*vectorlen);
j = 0;
if (sortval->type == REDIS_LIST) {
listTypeIterator *li = listTypeInitIterator(sortval,0,REDIS_TAIL);
listTypeEntry entry;
while(listTypeNext(li,&entry)) {
vector[j].obj = listTypeGet(&entry);
vector[j].u.score = 0;
vector[j].u.cmpobj = NULL;
j++;
}
listTypeReleaseIterator(li);
} else if (sortval->type == REDIS_SET) {
setTypeIterator *si = setTypeInitIterator(sortval);
robj *ele;
while((ele = setTypeNextObject(si)) != NULL) {
vector[j].obj = ele;
vector[j].u.score = 0;
vector[j].u.cmpobj = NULL;
j++;
}
setTypeReleaseIterator(si);
} else if (sortval->type == REDIS_ZSET) {
dict *set = ((zset*)sortval->ptr)->dict;
dictIterator *di;
dictEntry *setele;
di = dictGetIterator(set);
while((setele = dictNext(di)) != NULL) {
vector[j].obj = dictGetEntryKey(setele);
vector[j].u.score = 0;
vector[j].u.cmpobj = NULL;
j++;
}
dictReleaseIterator(di);
} else {
redisPanic("Unknown type");
}
redisAssert(j == vectorlen);
/* Now it's time to load the right scores in the sorting vector */
if (dontsort == 0) {
for (j = 0; j < vectorlen; j++) {
robj *byval;
if (sortby) {
/* lookup value to sort by */
byval = lookupKeyByPattern(c->db,sortby,vector[j].obj);
if (!byval) continue;
} else {
/* use object itself to sort by */
byval = vector[j].obj;
}
if (alpha) {
if (sortby) vector[j].u.cmpobj = getDecodedObject(byval);
} else {
if (byval->encoding == REDIS_ENCODING_RAW) {
vector[j].u.score = strtod(byval->ptr,NULL);
} else if (byval->encoding == REDIS_ENCODING_INT) {
/* Don't need to decode the object if it's
* integer-encoded (the only encoding supported) so
* far. We can just cast it */
#ifdef _WIN64
vector[j].u.score = (long long)byval->ptr;
#else
vector[j].u.score = (long)byval->ptr;
#endif
} else {
redisAssert(1 != 1);
}
}
/* when the object was retrieved using lookupKeyByPattern,
* its refcount needs to be decreased. */
if (sortby) {
decrRefCount(byval);
}
}
}
/* We are ready to sort the vector... perform a bit of sanity check
* on the LIMIT option too. We'll use a partial version of quicksort. */
start = (limit_start < 0) ? 0 : limit_start;
end = (limit_count < 0) ? vectorlen-1 : start+limit_count-1;
if (start >= vectorlen) {
start = vectorlen-1;
end = vectorlen-2;
}
if (end >= vectorlen) end = vectorlen-1;
if (dontsort == 0) {
server.sort_desc = desc;
server.sort_alpha = alpha;
server.sort_bypattern = sortby ? 1 : 0;
if (sortby && (start != 0 || end != vectorlen-1))
pqsort(vector,vectorlen,sizeof(redisSortObject),sortCompare, start,end);
else
qsort(vector,vectorlen,sizeof(redisSortObject),sortCompare);
}
/* Send command output to the output buffer, performing the specified
* GET/DEL/INCR/DECR operations if any. */
outputlen = getop ? getop*(end-start+1) : end-start+1;
if (storekey == NULL) {
/* STORE option not specified, sent the sorting result to client */
addReplyMultiBulkLen(c,outputlen);
for (j = start; j <= end; j++) {
listNode *ln;
listIter li;
if (!getop) addReplyBulk(c,vector[j].obj);
listRewind(operations,&li);
while((ln = listNext(&li))) {
redisSortOperation *sop = ln->value;
robj *val = lookupKeyByPattern(c->db,sop->pattern,
vector[j].obj);
if (sop->type == REDIS_SORT_GET) {
if (!val) {
addReply(c,shared.nullbulk);
} else {
addReplyBulk(c,val);
decrRefCount(val);
}
} else {
redisAssert(sop->type == REDIS_SORT_GET); /* always fails */
}
}
}
} else {
robj *sobj = createZiplistObject();
/* STORE option specified, set the sorting result as a List object */
for (j = start; j <= end; j++) {
listNode *ln;
listIter li;
if (!getop) {
listTypePush(sobj,vector[j].obj,REDIS_TAIL);
} else {
listRewind(operations,&li);
while((ln = listNext(&li))) {
redisSortOperation *sop = ln->value;
robj *val = lookupKeyByPattern(c->db,sop->pattern,
vector[j].obj);
if (sop->type == REDIS_SORT_GET) {
if (!val) val = createStringObject("",0);
/* listTypePush does an incrRefCount, so we should take care
* care of the incremented refcount caused by either
* lookupKeyByPattern or createStringObject("",0) */
listTypePush(sobj,val,REDIS_TAIL);
decrRefCount(val);
} else {
/* always fails */
redisAssert(sop->type == REDIS_SORT_GET);
}
}
}
}
dbReplace(c->db,storekey,sobj);
/* Note: we add 1 because the DB is dirty anyway since even if the
* SORT result is empty a new key is set and maybe the old content
* replaced. */
server.dirty += 1+outputlen;
touchWatchedKey(c->db,storekey);
addReplyLongLong(c,outputlen);
}
/* Cleanup */
if (sortval->type == REDIS_LIST || sortval->type == REDIS_SET)
for (j = 0; j < vectorlen; j++)
decrRefCount(vector[j].obj);
decrRefCount(sortval);
listRelease(operations);
for (j = 0; j < vectorlen; j++) {
if (alpha && vector[j].u.cmpobj)
decrRefCount(vector[j].u.cmpobj);
}
zfree(vector);
}
/* 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. */
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 = 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;
}
/* Try to swap an object that's a good candidate for swapping.
* Returns REDIS_OK if the object was swapped, REDIS_ERR if it's not possible
* to swap any object at all.
*
* If 'usethreaded' is true, Redis will try to swap the object in background
* using I/O threads. */
int vmSwapOneObject(int usethreads) {
int j, i;
struct dictEntry *best = NULL;
double best_swappability = 0;
redisDb *best_db = NULL;
robj *val;
sds key;
for (j = 0; j < server.dbnum; j++) {
redisDb *db = server.db+j;
/* Why maxtries is set to 100?
* Because this way (usually) we'll find 1 object even if just 1% - 2%
* are swappable objects */
int maxtries = 100;
if (dictSize(db->dict) == 0) continue;
for (i = 0; i < 5; i++) {
dictEntry *de;
double swappability;
if (maxtries) maxtries--;
de = dictGetRandomKey(db->dict);
val = dictGetEntryVal(de);
/* Only swap objects that are currently in memory.
*
* Also don't swap shared objects: not a good idea in general and
* we need to ensure that the main thread does not touch the
* object while the I/O thread is using it, but we can't
* control other keys without adding additional mutex. */
if (val->storage != REDIS_VM_MEMORY || val->refcount != 1) {
if (maxtries) i--; /* don't count this try */
continue;
}
swappability = computeObjectSwappability(val);
if (!best || swappability > best_swappability) {
best = de;
best_swappability = swappability;
best_db = db;
}
}
}
if (best == NULL) return REDIS_ERR;
key = dictGetEntryKey(best);
val = dictGetEntryVal(best);
redisLog(REDIS_DEBUG,"Key with best swappability: %s, %f",
key, best_swappability);
/* Swap it */
if (usethreads) {
robj *keyobj = createStringObject(key,sdslen(key));
vmSwapObjectThreaded(keyobj,val,best_db);
decrRefCount(keyobj);
return REDIS_OK;
} else {
vmpointer *vp;
if ((vp = vmSwapObjectBlocking(val)) != NULL) {
dictGetEntryVal(best) = vp;
return REDIS_OK;
} else {
return REDIS_ERR;
}
}
}
void zunionInterGenericCommand(redisClient *c, robj *dstkey, int op) {
int i, j, setnum;
int aggregate = REDIS_AGGR_SUM;
zsetopsrc *src;
robj *dstobj;
zset *dstzset;
zskiplistNode *znode;
dictIterator *di;
dictEntry *de;
int touched = 0;
/* expect setnum input keys to be given */
setnum = atoi(c->argv[2]->ptr);
if (setnum < 1) {
addReplyError(c,
"at least 1 input key is needed for ZUNIONSTORE/ZINTERSTORE");
return;
}
/* test if the expected number of keys would overflow */
if (3+setnum > c->argc) {
addReply(c,shared.syntaxerr);
return;
}
/* read keys to be used for input */
src = zmalloc(sizeof(zsetopsrc) * setnum);
for (i = 0, j = 3; i < setnum; i++, j++) {
robj *obj = lookupKeyWrite(c->db,c->argv[j]);
if (!obj) {
src[i].dict = NULL;
} else {
if (obj->type == REDIS_ZSET) {
src[i].dict = ((zset*)obj->ptr)->dict;
} else if (obj->type == REDIS_SET) {
src[i].dict = (obj->ptr);
} else {
zfree(src);
addReply(c,shared.wrongtypeerr);
return;
}
}
/* default all weights to 1 */
src[i].weight = 1.0;
}
/* parse optional extra arguments */
if (j < c->argc) {
int remaining = c->argc - j;
while (remaining) {
if (remaining >= (setnum + 1) && !strcasecmp(c->argv[j]->ptr,"weights")) {
j++; remaining--;
for (i = 0; i < setnum; i++, j++, remaining--) {
if (getDoubleFromObjectOrReply(c,c->argv[j],&src[i].weight,
"weight value is not a double") != REDIS_OK)
{
zfree(src);
return;
}
}
} else if (remaining >= 2 && !strcasecmp(c->argv[j]->ptr,"aggregate")) {
j++; remaining--;
if (!strcasecmp(c->argv[j]->ptr,"sum")) {
aggregate = REDIS_AGGR_SUM;
} else if (!strcasecmp(c->argv[j]->ptr,"min")) {
aggregate = REDIS_AGGR_MIN;
} else if (!strcasecmp(c->argv[j]->ptr,"max")) {
aggregate = REDIS_AGGR_MAX;
} else {
zfree(src);
addReply(c,shared.syntaxerr);
return;
}
j++; remaining--;
} else {
zfree(src);
addReply(c,shared.syntaxerr);
return;
}
}
}
/* sort sets from the smallest to largest, this will improve our
* algorithm's performance */
qsort(src,setnum,sizeof(zsetopsrc),qsortCompareZsetopsrcByCardinality);
dstobj = createZsetObject();
dstzset = dstobj->ptr;
if (op == REDIS_OP_INTER) {
/* skip going over all entries if the smallest zset is NULL or empty */
if (src[0].dict && dictSize(src[0].dict) > 0) {
/* precondition: as src[0].dict is non-empty and the zsets are ordered
* from small to large, all src[i > 0].dict are non-empty too */
di = dictGetIterator(src[0].dict);
while((de = dictNext(di)) != NULL) {
double score, value;
score = src[0].weight * zunionInterDictValue(de);
for (j = 1; j < setnum; j++) {
dictEntry *other = dictFind(src[j].dict,dictGetEntryKey(de));
if (other) {
value = src[j].weight * zunionInterDictValue(other);
zunionInterAggregate(&score,value,aggregate);
} else {
break;
}
}
/* Only continue when present in every source dict. */
if (j == setnum) {
robj *o = dictGetEntryKey(de);
znode = zslInsert(dstzset->zsl,score,o);
incrRefCount(o); /* added to skiplist */
dictAdd(dstzset->dict,o,&znode->score);
incrRefCount(o); /* added to dictionary */
}
}
dictReleaseIterator(di);
}
} else if (op == REDIS_OP_UNION) {
for (i = 0; i < setnum; i++) {
if (!src[i].dict) continue;
di = dictGetIterator(src[i].dict);
while((de = dictNext(di)) != NULL) {
double score, value;
/* skip key when already processed */
if (dictFind(dstzset->dict,dictGetEntryKey(de)) != NULL)
continue;
/* initialize score */
score = src[i].weight * zunionInterDictValue(de);
/* because the zsets are sorted by size, its only possible
* for sets at larger indices to hold this entry */
for (j = (i+1); j < setnum; j++) {
dictEntry *other = dictFind(src[j].dict,dictGetEntryKey(de));
if (other) {
value = src[j].weight * zunionInterDictValue(other);
zunionInterAggregate(&score,value,aggregate);
}
}
robj *o = dictGetEntryKey(de);
znode = zslInsert(dstzset->zsl,score,o);
incrRefCount(o); /* added to skiplist */
dictAdd(dstzset->dict,o,&znode->score);
incrRefCount(o); /* added to dictionary */
}
dictReleaseIterator(di);
}
} else {
/* unknown operator */
redisAssert(op == REDIS_OP_INTER || op == REDIS_OP_UNION);
}
if (dbDelete(c->db,dstkey)) {
touchWatchedKey(c->db,dstkey);
touched = 1;
server.dirty++;
}
if (dstzset->zsl->length) {
dbAdd(c->db,dstkey,dstobj);
addReplyLongLong(c, dstzset->zsl->length);
if (!touched) touchWatchedKey(c->db,dstkey);
server.dirty++;
} else {
decrRefCount(dstobj);
addReply(c, shared.czero);
}
zfree(src);
}
/* 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");
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);
/* SELECT the new DB */
fwrite(selectcmd,sizeof(selectcmd)-1,1,fp);
fwriteBulkLong(fp,j);
/* 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";
fwrite(cmd, sizeof(cmd)-1,1,fp);
/* Key and value */
fwriteBulkObject(fp,key);
fwriteBulkObject(fp,o);
}
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);
fwrite(cmd, sizeof(cmd)-1, 1, fp);
fwriteBulkObject(fp,key);
fwriteBulkObject(fp,eleobj);
}
}
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);
fwrite(cmd,sizeof(cmd)-1,1,fp) ;
fwriteBulkObject(fp,key) ;
fwriteBulkObject(fp,eleobj);
}
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 (fwriteBulkObject(fp,key) == 0) goto werr;
if (fwriteBulkDouble(fp,*score) == 0) goto werr;
if (fwriteBulkObject(fp,eleobj) == 0) goto werr;
}
dictReleaseIterator(di);
} 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) == -1)
return -1;
if (fwriteBulkString(fp,(char*)val,vlen) == -1)
return -1;
}
} 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) == -1) return -1;
if (fwriteBulkObject(fp,val) == -1) return -1;
}
dictReleaseIterator(di);
}
} else {
redisAssert(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;
fwrite(cmd,sizeof(cmd)-1,1,fp) ;
fwriteBulkObject(fp,key) ;
fwriteBulkLong(fp,expiretime) ;
}
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. */
rename(tmpfile,filename) ;
redisLog(REDIS_NOTICE,"SYNC append only file rewrite performed");
return REDIS_OK;
}
/* How a good candidate is this object for swapping?
* The better candidate it is, the greater the returned value.
*
* Currently we try to perform a fast estimation of the object size in
* memory, and combine it with aging informations.
*
* Basically swappability = idle-time * log(estimated size)
*
* Bigger objects are preferred over smaller objects, but not
* proportionally, this is why we use the logarithm. This algorithm is
* just a first try and will probably be tuned later. */
double computeObjectSwappability(robj *o) {
/* actual age can be >= minage, but not < minage. As we use wrapping
* 21 bit clocks with minutes resolution for the LRU. */
time_t minage = estimateObjectIdleTime(o);
#ifdef _WIN32
ssize_t asize = 0, elesize;
#else
long asize = 0, elesize;
#endif
robj *ele;
list *l;
listNode *ln;
dict *d;
struct dictEntry *de;
if (minage <= 0) return 0;
switch(o->type) {
case REDIS_STRING:
if (o->encoding != REDIS_ENCODING_RAW) {
asize = sizeof(*o);
} else {
#ifdef _WIN32
asize = sdslen(o->ptr)+sizeof(*o)+sizeof(size_t)*2;
#else
asize = sdslen(o->ptr)+sizeof(*o)+sizeof(long)*2;
#endif
}
break;
case REDIS_LIST:
if (o->encoding == REDIS_ENCODING_ZIPLIST) {
asize = sizeof(*o)+ziplistBlobLen(o->ptr);
} else {
l = o->ptr;
ln = listFirst(l);
asize = sizeof(list);
if (ln) {
ele = ln->value;
elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
(sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
asize += (sizeof(listNode)+elesize)*listLength(l);
}
}
break;
case REDIS_SET:
if (o->encoding == REDIS_ENCODING_INTSET) {
intset *is = o->ptr;
asize = sizeof(*is)+is->encoding*is->length;
} else {
d = o->ptr;
asize = sizeof(dict)+(sizeof(struct dictEntry*)*dictSlots(d));
if (dictSize(d)) {
de = dictGetRandomKey(d);
ele = dictGetEntryKey(de);
elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
(sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
asize += (sizeof(struct dictEntry)+elesize)*dictSize(d);
}
}
break;
case REDIS_ZSET:
if (o->encoding == REDIS_ENCODING_ZIPLIST) {
asize = sizeof(*o)+(ziplistBlobLen(o->ptr) / 2);
} else {
d = ((zset*)o->ptr)->dict;
asize = sizeof(zset)+(sizeof(struct dictEntry*)*dictSlots(d));
if (dictSize(d)) {
de = dictGetRandomKey(d);
ele = dictGetEntryKey(de);
elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
(sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
asize += (sizeof(struct dictEntry)+elesize)*dictSize(d);
asize += sizeof(zskiplistNode)*dictSize(d);
}
}
break;
case REDIS_HASH:
if (o->encoding == REDIS_ENCODING_ZIPMAP) {
unsigned char *p = zipmapRewind((unsigned char*)o->ptr);
unsigned int len = zipmapLen((unsigned char*)o->ptr);
unsigned int klen, vlen;
unsigned char *key, *val;
if ((p = zipmapNext(p,&key,&klen,&val,&vlen)) == NULL) {
klen = 0;
vlen = 0;
}
asize = len*(klen+vlen+3);
} else if (o->encoding == REDIS_ENCODING_HT) {
d = o->ptr;
asize = sizeof(dict)+(sizeof(struct dictEntry*)*dictSlots(d));
if (dictSize(d)) {
de = dictGetRandomKey(d);
ele = dictGetEntryKey(de);
elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
(sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
ele = dictGetEntryVal(de);
elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
(sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
asize += (sizeof(struct dictEntry)+elesize)*dictSize(d);
}
}
break;
}
return (double)minage*log(1+(double)asize);
}
