-#include "redis.h"
-#include "lzf.h" /* LZF compression library */
-
#include <math.h>
#include <sys/types.h>
#include <sys/time.h>
#include <sys/resource.h>
#include <sys/wait.h>
#include <arpa/inet.h>
+#include <sys/stat.h>
+#include "rdb.h"
+#include "lzf.h" /* LZF compression library */
+
+static int rdbWriteRaw(rio *rdb, void *p, size_t len) {
+ if (rioWrite(rdb,p,len) == 0)
+ return -1;
+ return 1;
+}
-int rdbSaveType(FILE *fp, unsigned char type) {
- if (fwrite(&type,1,1,fp) == 0) return -1;
- return 0;
+int rdbSaveType(rio *rdb, unsigned char type) {
+ return rdbWriteRaw(rdb,&type,1);
}
-int rdbSaveTime(FILE *fp, time_t t) {
+int rdbSaveTime(rio *rdb, time_t t) {
int32_t t32 = (int32_t) t;
- if (fwrite(&t32,4,1,fp) == 0) return -1;
- return 0;
+ return rdbWriteRaw(rdb,&t32,4);
}
/* check rdbLoadLen() comments for more info */
-int rdbSaveLen(FILE *fp, uint32_t len) {
+int rdbSaveLen(rio *rdb, uint32_t len) {
unsigned char buf[2];
+ size_t nwritten;
if (len < (1<<6)) {
/* Save a 6 bit len */
buf[0] = (len&0xFF)|(REDIS_RDB_6BITLEN<<6);
- if (fwrite(buf,1,1,fp) == 0) return -1;
+ if (rdbWriteRaw(rdb,buf,1) == -1) return -1;
+ nwritten = 1;
} else if (len < (1<<14)) {
/* Save a 14 bit len */
buf[0] = ((len>>8)&0xFF)|(REDIS_RDB_14BITLEN<<6);
buf[1] = len&0xFF;
- if (fwrite(buf,2,1,fp) == 0) return -1;
+ if (rdbWriteRaw(rdb,buf,2) == -1) return -1;
+ nwritten = 2;
} else {
/* Save a 32 bit len */
buf[0] = (REDIS_RDB_32BITLEN<<6);
- if (fwrite(buf,1,1,fp) == 0) return -1;
+ if (rdbWriteRaw(rdb,buf,1) == -1) return -1;
len = htonl(len);
- if (fwrite(&len,4,1,fp) == 0) return -1;
+ if (rdbWriteRaw(rdb,&len,4) == -4) return -1;
+ nwritten = 1+4;
}
- return 0;
+ return nwritten;
}
/* Encode 'value' as an integer if possible (if integer will fit the
return rdbEncodeInteger(value,enc);
}
-int rdbSaveLzfStringObject(FILE *fp, unsigned char *s, size_t len) {
+int rdbSaveLzfStringObject(rio *rdb, unsigned char *s, size_t len) {
size_t comprlen, outlen;
unsigned char byte;
+ int n, nwritten = 0;
void *out;
/* We require at least four bytes compression for this to be worth it */
}
/* Data compressed! Let's save it on disk */
byte = (REDIS_RDB_ENCVAL<<6)|REDIS_RDB_ENC_LZF;
- if (fwrite(&byte,1,1,fp) == 0) goto writeerr;
- if (rdbSaveLen(fp,comprlen) == -1) goto writeerr;
- if (rdbSaveLen(fp,len) == -1) goto writeerr;
- if (fwrite(out,comprlen,1,fp) == 0) goto writeerr;
+ if ((n = rdbWriteRaw(rdb,&byte,1)) == -1) goto writeerr;
+ nwritten += n;
+
+ if ((n = rdbSaveLen(rdb,comprlen)) == -1) goto writeerr;
+ nwritten += n;
+
+ if ((n = rdbSaveLen(rdb,len)) == -1) goto writeerr;
+ nwritten += n;
+
+ if ((n = rdbWriteRaw(rdb,out,comprlen)) == -1) goto writeerr;
+ nwritten += n;
+
zfree(out);
- return comprlen;
+ return nwritten;
writeerr:
zfree(out);
}
/* Save a string objet as [len][data] on disk. If the object is a string
- * representation of an integer value we try to safe it in a special form */
-int rdbSaveRawString(FILE *fp, unsigned char *s, size_t len) {
+ * representation of an integer value we try to save it in a special form */
+int rdbSaveRawString(rio *rdb, unsigned char *s, size_t len) {
int enclen;
+ int n, nwritten = 0;
/* Try integer encoding */
if (len <= 11) {
unsigned char buf[5];
if ((enclen = rdbTryIntegerEncoding((char*)s,len,buf)) > 0) {
- if (fwrite(buf,enclen,1,fp) == 0) return -1;
- return 0;
+ if (rdbWriteRaw(rdb,buf,enclen) == -1) return -1;
+ return enclen;
}
}
/* Try LZF compression - under 20 bytes it's unable to compress even
* aaaaaaaaaaaaaaaaaa so skip it */
if (server.rdbcompression && len > 20) {
- int retval;
-
- retval = rdbSaveLzfStringObject(fp,s,len);
- if (retval == -1) return -1;
- if (retval > 0) return 0;
- /* retval == 0 means data can't be compressed, save the old way */
+ n = rdbSaveLzfStringObject(rdb,s,len);
+ if (n == -1) return -1;
+ if (n > 0) return n;
+ /* Return value of 0 means data can't be compressed, save the old way */
}
/* Store verbatim */
- if (rdbSaveLen(fp,len) == -1) return -1;
- if (len && fwrite(s,len,1,fp) == 0) return -1;
- return 0;
+ if ((n = rdbSaveLen(rdb,len)) == -1) return -1;
+ nwritten += n;
+ if (len > 0) {
+ if (rdbWriteRaw(rdb,s,len) == -1) return -1;
+ nwritten += len;
+ }
+ return nwritten;
}
/* Save a long long value as either an encoded string or a string. */
-int rdbSaveLongLongAsStringObject(FILE *fp, long long value) {
+int rdbSaveLongLongAsStringObject(rio *rdb, long long value) {
unsigned char buf[32];
+ int n, nwritten = 0;
int enclen = rdbEncodeInteger(value,buf);
if (enclen > 0) {
- if (fwrite(buf,enclen,1,fp) == 0) return -1;
+ return rdbWriteRaw(rdb,buf,enclen);
} else {
/* Encode as string */
enclen = ll2string((char*)buf,32,value);
redisAssert(enclen < 32);
- if (rdbSaveLen(fp,enclen) == -1) return -1;
- if (fwrite(buf,enclen,1,fp) == 0) return -1;
+ if ((n = rdbSaveLen(rdb,enclen)) == -1) return -1;
+ nwritten += n;
+ if ((n = rdbWriteRaw(rdb,buf,enclen)) == -1) return -1;
+ nwritten += n;
}
- return 0;
+ return nwritten;
}
/* Like rdbSaveStringObjectRaw() but handle encoded objects */
-int rdbSaveStringObject(FILE *fp, robj *obj) {
+int rdbSaveStringObject(rio *rdb, robj *obj) {
/* Avoid to decode the object, then encode it again, if the
* object is alrady integer encoded. */
if (obj->encoding == REDIS_ENCODING_INT) {
- return rdbSaveLongLongAsStringObject(fp,(long)obj->ptr);
+ return rdbSaveLongLongAsStringObject(rdb,(long)obj->ptr);
} else {
redisAssert(obj->encoding == REDIS_ENCODING_RAW);
- return rdbSaveRawString(fp,obj->ptr,sdslen(obj->ptr));
+ return rdbSaveRawString(rdb,obj->ptr,sdslen(obj->ptr));
}
}
* 254: + inf
* 255: - inf
*/
-int rdbSaveDoubleValue(FILE *fp, double val) {
+int rdbSaveDoubleValue(rio *rdb, double val) {
unsigned char buf[128];
int len;
buf[0] = strlen((char*)buf+1);
len = buf[0]+1;
}
- if (fwrite(buf,len,1,fp) == 0) return -1;
- return 0;
+ return rdbWriteRaw(rdb,buf,len);
}
-/* Save a Redis object. */
-int rdbSaveObject(FILE *fp, robj *o) {
+/* Save a Redis object. Returns -1 on error, 0 on success. */
+int rdbSaveObject(rio *rdb, robj *o) {
+ int n, nwritten = 0;
+
if (o->type == REDIS_STRING) {
/* Save a string value */
- if (rdbSaveStringObject(fp,o) == -1) return -1;
+ if ((n = rdbSaveStringObject(rdb,o)) == -1) return -1;
+ nwritten += n;
} else if (o->type == REDIS_LIST) {
/* Save a list value */
if (o->encoding == REDIS_ENCODING_ZIPLIST) {
- unsigned char *p;
- unsigned char *vstr;
- unsigned int vlen;
- long long vlong;
-
- if (rdbSaveLen(fp,ziplistLen(o->ptr)) == -1) return -1;
- p = ziplistIndex(o->ptr,0);
- while(ziplistGet(p,&vstr,&vlen,&vlong)) {
- if (vstr) {
- if (rdbSaveRawString(fp,vstr,vlen) == -1)
- return -1;
- } else {
- if (rdbSaveLongLongAsStringObject(fp,vlong) == -1)
- return -1;
- }
- p = ziplistNext(o->ptr,p);
- }
+ size_t l = ziplistBlobLen((unsigned char*)o->ptr);
+
+ if ((n = rdbSaveRawString(rdb,o->ptr,l)) == -1) return -1;
+ nwritten += n;
} else if (o->encoding == REDIS_ENCODING_LINKEDLIST) {
list *list = o->ptr;
listIter li;
listNode *ln;
- if (rdbSaveLen(fp,listLength(list)) == -1) return -1;
+ if ((n = rdbSaveLen(rdb,listLength(list))) == -1) return -1;
+ nwritten += n;
+
listRewind(list,&li);
while((ln = listNext(&li))) {
robj *eleobj = listNodeValue(ln);
- if (rdbSaveStringObject(fp,eleobj) == -1) return -1;
+ if ((n = rdbSaveStringObject(rdb,eleobj)) == -1) return -1;
+ nwritten += n;
}
} else {
redisPanic("Unknown list encoding");
dictIterator *di = dictGetIterator(set);
dictEntry *de;
- if (rdbSaveLen(fp,dictSize(set)) == -1) return -1;
+ if ((n = rdbSaveLen(rdb,dictSize(set))) == -1) return -1;
+ nwritten += n;
+
while((de = dictNext(di)) != NULL) {
robj *eleobj = dictGetEntryKey(de);
- if (rdbSaveStringObject(fp,eleobj) == -1) return -1;
+ if ((n = rdbSaveStringObject(rdb,eleobj)) == -1) return -1;
+ nwritten += n;
}
dictReleaseIterator(di);
} else if (o->encoding == REDIS_ENCODING_INTSET) {
- intset *is = o->ptr;
- int64_t llval;
- int i = 0;
+ size_t l = intsetBlobLen((intset*)o->ptr);
- if (rdbSaveLen(fp,intsetLen(is)) == -1) return -1;
- while(intsetGet(is,i++,&llval)) {
- if (rdbSaveLongLongAsStringObject(fp,llval) == -1) return -1;
- }
+ if ((n = rdbSaveRawString(rdb,o->ptr,l)) == -1) return -1;
+ nwritten += n;
} else {
redisPanic("Unknown set encoding");
}
} else if (o->type == REDIS_ZSET) {
- /* Save a set value */
- zset *zs = o->ptr;
- dictIterator *di = dictGetIterator(zs->dict);
- dictEntry *de;
+ /* Save a sorted set value */
+ if (o->encoding == REDIS_ENCODING_ZIPLIST) {
+ size_t l = ziplistBlobLen((unsigned char*)o->ptr);
- if (rdbSaveLen(fp,dictSize(zs->dict)) == -1) return -1;
- while((de = dictNext(di)) != NULL) {
- robj *eleobj = dictGetEntryKey(de);
- double *score = dictGetEntryVal(de);
+ if ((n = rdbSaveRawString(rdb,o->ptr,l)) == -1) return -1;
+ nwritten += n;
+ } else if (o->encoding == REDIS_ENCODING_SKIPLIST) {
+ zset *zs = o->ptr;
+ dictIterator *di = dictGetIterator(zs->dict);
+ dictEntry *de;
+
+ if ((n = rdbSaveLen(rdb,dictSize(zs->dict))) == -1) return -1;
+ nwritten += n;
+
+ while((de = dictNext(di)) != NULL) {
+ robj *eleobj = dictGetEntryKey(de);
+ double *score = dictGetEntryVal(de);
- if (rdbSaveStringObject(fp,eleobj) == -1) return -1;
- if (rdbSaveDoubleValue(fp,*score) == -1) return -1;
+ if ((n = rdbSaveStringObject(rdb,eleobj)) == -1) return -1;
+ nwritten += n;
+ if ((n = rdbSaveDoubleValue(rdb,*score)) == -1) return -1;
+ nwritten += n;
+ }
+ dictReleaseIterator(di);
+ } else {
+ redisPanic("Unknown sorted set encoding");
}
- dictReleaseIterator(di);
} else if (o->type == REDIS_HASH) {
/* Save a hash value */
if (o->encoding == REDIS_ENCODING_ZIPMAP) {
- unsigned char *p = zipmapRewind(o->ptr);
- unsigned int count = zipmapLen(o->ptr);
- unsigned char *key, *val;
- unsigned int klen, vlen;
-
- if (rdbSaveLen(fp,count) == -1) return -1;
- while((p = zipmapNext(p,&key,&klen,&val,&vlen)) != NULL) {
- if (rdbSaveRawString(fp,key,klen) == -1) return -1;
- if (rdbSaveRawString(fp,val,vlen) == -1) return -1;
- }
+ size_t l = zipmapBlobLen((unsigned char*)o->ptr);
+
+ if ((n = rdbSaveRawString(rdb,o->ptr,l)) == -1) return -1;
+ nwritten += n;
} else {
dictIterator *di = dictGetIterator(o->ptr);
dictEntry *de;
- if (rdbSaveLen(fp,dictSize((dict*)o->ptr)) == -1) return -1;
+ if ((n = rdbSaveLen(rdb,dictSize((dict*)o->ptr))) == -1) return -1;
+ nwritten += n;
+
while((de = dictNext(di)) != NULL) {
robj *key = dictGetEntryKey(de);
robj *val = dictGetEntryVal(de);
- if (rdbSaveStringObject(fp,key) == -1) return -1;
- if (rdbSaveStringObject(fp,val) == -1) return -1;
+ if ((n = rdbSaveStringObject(rdb,key)) == -1) return -1;
+ nwritten += n;
+ if ((n = rdbSaveStringObject(rdb,val)) == -1) return -1;
+ nwritten += n;
}
dictReleaseIterator(di);
}
} else {
redisPanic("Unknown object type");
}
- return 0;
+ return nwritten;
}
/* Return the length the object will have on disk if saved with
* the rdbSaveObject() function. Currently we use a trick to get
* this length with very little changes to the code. In the future
* we could switch to a faster solution. */
-off_t rdbSavedObjectLen(robj *o, FILE *fp) {
- if (fp == NULL) fp = server.devnull;
- rewind(fp);
- redisAssert(rdbSaveObject(fp,o) != 1);
- return ftello(fp);
+off_t rdbSavedObjectLen(robj *o) {
+ int len = rdbSaveObject(NULL,o);
+ redisAssert(len != -1);
+ return len;
}
-/* Return the number of pages required to save this object in the swap file */
-off_t rdbSavedObjectPages(robj *o, FILE *fp) {
- off_t bytes = rdbSavedObjectLen(o,fp);
-
- return (bytes+(server.vm_page_size-1))/server.vm_page_size;
+/* Save a key-value pair, with expire time, type, key, value.
+ * On error -1 is returned.
+ * On success if the key was actaully saved 1 is returned, otherwise 0
+ * is returned (the key was already expired). */
+int rdbSaveKeyValuePair(rio *rdb, robj *key, robj *val,
+ time_t expiretime, time_t now)
+{
+ int vtype;
+
+ /* Save the expire time */
+ if (expiretime != -1) {
+ /* If this key is already expired skip it */
+ if (expiretime < now) return 0;
+ if (rdbSaveType(rdb,REDIS_EXPIRETIME) == -1) return -1;
+ if (rdbSaveTime(rdb,expiretime) == -1) return -1;
+ }
+ /* Fix the object type if needed, to support saving zipmaps, ziplists,
+ * and intsets, directly as blobs of bytes: they are already serialized. */
+ vtype = val->type;
+ if (vtype == REDIS_HASH && val->encoding == REDIS_ENCODING_ZIPMAP)
+ vtype = REDIS_HASH_ZIPMAP;
+ else if (vtype == REDIS_LIST && val->encoding == REDIS_ENCODING_ZIPLIST)
+ vtype = REDIS_LIST_ZIPLIST;
+ else if (vtype == REDIS_SET && val->encoding == REDIS_ENCODING_INTSET)
+ vtype = REDIS_SET_INTSET;
+ else if (vtype == REDIS_ZSET && val->encoding == REDIS_ENCODING_ZIPLIST)
+ vtype = REDIS_ZSET_ZIPLIST;
+ /* Save type, key, value */
+ if (rdbSaveType(rdb,vtype) == -1) return -1;
+ if (rdbSaveStringObject(rdb,key) == -1) return -1;
+ if (rdbSaveObject(rdb,val) == -1) return -1;
+ return 1;
}
/* Save the DB on disk. Return REDIS_ERR on error, REDIS_OK on success */
int rdbSave(char *filename) {
dictIterator *di = NULL;
dictEntry *de;
- FILE *fp;
char tmpfile[256];
int j;
time_t now = time(NULL);
+ FILE *fp;
+ rio rdb;
- /* Wait for I/O therads to terminate, just in case this is a
- * foreground-saving, to avoid seeking the swap file descriptor at the
- * same time. */
- if (server.vm_enabled)
- waitEmptyIOJobsQueue();
+ if (server.ds_enabled) {
+ cacheForcePointInTime();
+ return dsRdbSave(filename);
+ }
snprintf(tmpfile,256,"temp-%d.rdb", (int) getpid());
fp = fopen(tmpfile,"w");
if (!fp) {
- redisLog(REDIS_WARNING, "Failed saving the DB: %s", strerror(errno));
+ redisLog(REDIS_WARNING, "Failed opening .rdb for saving: %s",
+ strerror(errno));
return REDIS_ERR;
}
- if (fwrite("REDIS0001",9,1,fp) == 0) goto werr;
+
+ rdb = rioInitWithFile(fp);
+ if (rdbWriteRaw(&rdb,"REDIS0002",9) == -1) goto werr;
+
for (j = 0; j < server.dbnum; j++) {
redisDb *db = server.db+j;
dict *d = db->dict;
}
/* Write the SELECT DB opcode */
- if (rdbSaveType(fp,REDIS_SELECTDB) == -1) goto werr;
- if (rdbSaveLen(fp,j) == -1) goto werr;
+ if (rdbSaveType(&rdb,REDIS_SELECTDB) == -1) goto werr;
+ if (rdbSaveLen(&rdb,j) == -1) goto werr;
/* Iterate this DB writing every entry */
while((de = dictNext(di)) != NULL) {
sds keystr = dictGetEntryKey(de);
robj key, *o = dictGetEntryVal(de);
- time_t expiretime;
+ time_t expire;
initStaticStringObject(key,keystr);
- expiretime = getExpire(db,&key);
-
- /* Save the expire time */
- if (expiretime != -1) {
- /* If this key is already expired skip it */
- if (expiretime < now) continue;
- if (rdbSaveType(fp,REDIS_EXPIRETIME) == -1) goto werr;
- if (rdbSaveTime(fp,expiretime) == -1) goto werr;
- }
- /* Save the key and associated value. This requires special
- * handling if the value is swapped out. */
- if (!server.vm_enabled || o->storage == REDIS_VM_MEMORY ||
- o->storage == REDIS_VM_SWAPPING) {
- /* Save type, key, value */
- if (rdbSaveType(fp,o->type) == -1) goto werr;
- if (rdbSaveStringObject(fp,&key) == -1) goto werr;
- if (rdbSaveObject(fp,o) == -1) goto werr;
- } else {
- /* REDIS_VM_SWAPPED or REDIS_VM_LOADING */
- robj *po;
- /* Get a preview of the object in memory */
- po = vmPreviewObject(o);
- /* Save type, key, value */
- if (rdbSaveType(fp,po->type) == -1) goto werr;
- if (rdbSaveStringObject(fp,&key) == -1) goto werr;
- if (rdbSaveObject(fp,po) == -1) goto werr;
- /* Remove the loaded object from memory */
- decrRefCount(po);
- }
+ expire = getExpire(db,&key);
+ if (rdbSaveKeyValuePair(&rdb,&key,o,expire,now) == -1) goto werr;
}
dictReleaseIterator(di);
}
/* EOF opcode */
- if (rdbSaveType(fp,REDIS_EOF) == -1) goto werr;
+ if (rdbSaveType(&rdb,REDIS_EOF) == -1) goto werr;
/* Make sure data will not remain on the OS's output buffers */
fflush(fp);
int rdbSaveBackground(char *filename) {
pid_t childpid;
- if (server.bgsavechildpid != -1) return REDIS_ERR;
- if (server.vm_enabled) waitEmptyIOJobsQueue();
+ if (server.bgsavechildpid != -1 ||
+ server.bgsavethread != (pthread_t) -1) return REDIS_ERR;
+
server.dirty_before_bgsave = server.dirty;
+
+ if (server.ds_enabled) {
+ cacheForcePointInTime();
+ return dsRdbSaveBackground(filename);
+ }
+
if ((childpid = fork()) == 0) {
+ int retval;
+
/* Child */
- if (server.vm_enabled) vmReopenSwapFile();
if (server.ipfd > 0) close(server.ipfd);
if (server.sofd > 0) close(server.sofd);
- if (rdbSave(filename) == REDIS_OK) {
- _exit(0);
- } else {
- _exit(1);
- }
+ retval = rdbSave(filename);
+ _exit((retval == REDIS_OK) ? 0 : 1);
} else {
/* Parent */
if (childpid == -1) {
unlink(tmpfile);
}
-int rdbLoadType(FILE *fp) {
+int rdbLoadType(rio *rdb) {
unsigned char type;
- if (fread(&type,1,1,fp) == 0) return -1;
+ if (rioRead(rdb,&type,1) == 0) return -1;
return type;
}
-time_t rdbLoadTime(FILE *fp) {
+time_t rdbLoadTime(rio *rdb) {
int32_t t32;
- if (fread(&t32,4,1,fp) == 0) return -1;
+ if (rioRead(rdb,&t32,4) == 0) return -1;
return (time_t) t32;
}
*
* isencoded is set to 1 if the readed length is not actually a length but
* an "encoding type", check the above comments for more info */
-uint32_t rdbLoadLen(FILE *fp, int *isencoded) {
+uint32_t rdbLoadLen(rio *rdb, int *isencoded) {
unsigned char buf[2];
uint32_t len;
int type;
if (isencoded) *isencoded = 0;
- if (fread(buf,1,1,fp) == 0) return REDIS_RDB_LENERR;
+ if (rioRead(rdb,buf,1) == 0) return REDIS_RDB_LENERR;
type = (buf[0]&0xC0)>>6;
if (type == REDIS_RDB_6BITLEN) {
/* Read a 6 bit len */
return buf[0]&0x3F;
} else if (type == REDIS_RDB_14BITLEN) {
/* Read a 14 bit len */
- if (fread(buf+1,1,1,fp) == 0) return REDIS_RDB_LENERR;
+ if (rioRead(rdb,buf+1,1) == 0) return REDIS_RDB_LENERR;
return ((buf[0]&0x3F)<<8)|buf[1];
} else {
/* Read a 32 bit len */
- if (fread(&len,4,1,fp) == 0) return REDIS_RDB_LENERR;
+ if (rioRead(rdb,&len,4) == 0) return REDIS_RDB_LENERR;
return ntohl(len);
}
}
* encoding type 'enctype'. If encode is true the function may return
* an integer-encoded object as reply, otherwise the returned object
* will always be encoded as a raw string. */
-robj *rdbLoadIntegerObject(FILE *fp, int enctype, int encode) {
+robj *rdbLoadIntegerObject(rio *rdb, int enctype, int encode) {
unsigned char enc[4];
long long val;
if (enctype == REDIS_RDB_ENC_INT8) {
- if (fread(enc,1,1,fp) == 0) return NULL;
+ if (rioRead(rdb,enc,1) == 0) return NULL;
val = (signed char)enc[0];
} else if (enctype == REDIS_RDB_ENC_INT16) {
uint16_t v;
- if (fread(enc,2,1,fp) == 0) return NULL;
+ if (rioRead(rdb,enc,2) == 0) return NULL;
v = enc[0]|(enc[1]<<8);
val = (int16_t)v;
} else if (enctype == REDIS_RDB_ENC_INT32) {
uint32_t v;
- if (fread(enc,4,1,fp) == 0) return NULL;
+ if (rioRead(rdb,enc,4) == 0) return NULL;
v = enc[0]|(enc[1]<<8)|(enc[2]<<16)|(enc[3]<<24);
val = (int32_t)v;
} else {
return createObject(REDIS_STRING,sdsfromlonglong(val));
}
-robj *rdbLoadLzfStringObject(FILE*fp) {
+robj *rdbLoadLzfStringObject(rio *rdb) {
unsigned int len, clen;
unsigned char *c = NULL;
sds val = NULL;
- if ((clen = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
- if ((len = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((clen = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((len = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
if ((c = zmalloc(clen)) == NULL) goto err;
if ((val = sdsnewlen(NULL,len)) == NULL) goto err;
- if (fread(c,clen,1,fp) == 0) goto err;
+ if (rioRead(rdb,c,clen) == 0) goto err;
if (lzf_decompress(c,clen,val,len) == 0) goto err;
zfree(c);
return createObject(REDIS_STRING,val);
return NULL;
}
-robj *rdbGenericLoadStringObject(FILE*fp, int encode) {
+robj *rdbGenericLoadStringObject(rio *rdb, int encode) {
int isencoded;
uint32_t len;
sds val;
- len = rdbLoadLen(fp,&isencoded);
+ len = rdbLoadLen(rdb,&isencoded);
if (isencoded) {
switch(len) {
case REDIS_RDB_ENC_INT8:
case REDIS_RDB_ENC_INT16:
case REDIS_RDB_ENC_INT32:
- return rdbLoadIntegerObject(fp,len,encode);
+ return rdbLoadIntegerObject(rdb,len,encode);
case REDIS_RDB_ENC_LZF:
- return rdbLoadLzfStringObject(fp);
+ return rdbLoadLzfStringObject(rdb);
default:
redisPanic("Unknown RDB encoding type");
}
if (len == REDIS_RDB_LENERR) return NULL;
val = sdsnewlen(NULL,len);
- if (len && fread(val,len,1,fp) == 0) {
+ if (len && rioRead(rdb,val,len) == 0) {
sdsfree(val);
return NULL;
}
return createObject(REDIS_STRING,val);
}
-robj *rdbLoadStringObject(FILE *fp) {
- return rdbGenericLoadStringObject(fp,0);
+robj *rdbLoadStringObject(rio *rdb) {
+ return rdbGenericLoadStringObject(rdb,0);
}
-robj *rdbLoadEncodedStringObject(FILE *fp) {
- return rdbGenericLoadStringObject(fp,1);
+robj *rdbLoadEncodedStringObject(rio *rdb) {
+ return rdbGenericLoadStringObject(rdb,1);
}
/* For information about double serialization check rdbSaveDoubleValue() */
-int rdbLoadDoubleValue(FILE *fp, double *val) {
+int rdbLoadDoubleValue(rio *rdb, double *val) {
char buf[128];
unsigned char len;
- if (fread(&len,1,1,fp) == 0) return -1;
+ if (rioRead(rdb,&len,1) == 0) return -1;
switch(len) {
case 255: *val = R_NegInf; return 0;
case 254: *val = R_PosInf; return 0;
case 253: *val = R_Nan; return 0;
default:
- if (fread(buf,len,1,fp) == 0) return -1;
+ if (rioRead(rdb,buf,len) == 0) return -1;
buf[len] = '\0';
sscanf(buf, "%lg", val);
return 0;
/* Load a Redis object of the specified type from the specified file.
* On success a newly allocated object is returned, otherwise NULL. */
-robj *rdbLoadObject(int type, FILE *fp) {
+robj *rdbLoadObject(int type, rio *rdb) {
robj *o, *ele, *dec;
size_t len;
unsigned int i;
- redisLog(REDIS_DEBUG,"LOADING OBJECT %d (at %d)\n",type,ftell(fp));
+ redisLog(REDIS_DEBUG,"LOADING OBJECT %d (at %d)\n",type,rdb->tell(rdb));
if (type == REDIS_STRING) {
/* Read string value */
- if ((o = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
+ if ((o = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
o = tryObjectEncoding(o);
} else if (type == REDIS_LIST) {
/* Read list value */
- if ((len = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((len = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
/* Use a real list when there are too many entries */
if (len > server.list_max_ziplist_entries) {
/* Load every single element of the list */
while(len--) {
- if ((ele = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
+ if ((ele = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
/* If we are using a ziplist and the value is too big, convert
* the object to a real list. */
}
} else if (type == REDIS_SET) {
/* Read list/set value */
- if ((len = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((len = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
/* Use a regular set when there are too many entries. */
if (len > server.set_max_intset_entries) {
/* Load every single element of the list/set */
for (i = 0; i < len; i++) {
long long llval;
- if ((ele = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
+ if ((ele = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
ele = tryObjectEncoding(ele);
if (o->encoding == REDIS_ENCODING_INTSET) {
} else if (type == REDIS_ZSET) {
/* Read list/set value */
size_t zsetlen;
+ size_t maxelelen = 0;
zset *zs;
- if ((zsetlen = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((zsetlen = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
o = createZsetObject();
zs = o->ptr;
+
/* Load every single element of the list/set */
while(zsetlen--) {
robj *ele;
double score;
zskiplistNode *znode;
- if ((ele = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
+ if ((ele = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
ele = tryObjectEncoding(ele);
- if (rdbLoadDoubleValue(fp,&score) == -1) return NULL;
+ if (rdbLoadDoubleValue(rdb,&score) == -1) return NULL;
+
+ /* Don't care about integer-encoded strings. */
+ if (ele->encoding == REDIS_ENCODING_RAW &&
+ sdslen(ele->ptr) > maxelelen)
+ maxelelen = sdslen(ele->ptr);
+
znode = zslInsert(zs->zsl,score,ele);
dictAdd(zs->dict,ele,&znode->score);
incrRefCount(ele); /* added to skiplist */
}
+
+ /* Convert *after* loading, since sorted sets are not stored ordered. */
+ if (zsetLength(o) <= server.zset_max_ziplist_entries &&
+ maxelelen <= server.zset_max_ziplist_value)
+ zsetConvert(o,REDIS_ENCODING_ZIPLIST);
} else if (type == REDIS_HASH) {
size_t hashlen;
- if ((hashlen = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR) return NULL;
+ if ((hashlen = rdbLoadLen(rdb,NULL)) == REDIS_RDB_LENERR) return NULL;
o = createHashObject();
/* Too many entries? Use an hash table. */
if (hashlen > server.hash_max_zipmap_entries)
while(hashlen--) {
robj *key, *val;
- if ((key = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
- if ((val = rdbLoadEncodedStringObject(fp)) == NULL) return NULL;
+ if ((key = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
+ if ((val = rdbLoadEncodedStringObject(rdb)) == NULL) return NULL;
/* If we are using a zipmap and there are too big values
* the object is converted to real hash table encoding. */
if (o->encoding != REDIS_ENCODING_HT &&
dictAdd((dict*)o->ptr,key,val);
}
}
+ } else if (type == REDIS_HASH_ZIPMAP ||
+ type == REDIS_LIST_ZIPLIST ||
+ type == REDIS_SET_INTSET ||
+ type == REDIS_ZSET_ZIPLIST)
+ {
+ robj *aux = rdbLoadStringObject(rdb);
+
+ if (aux == NULL) return NULL;
+ o = createObject(REDIS_STRING,NULL); /* string is just placeholder */
+ o->ptr = zmalloc(sdslen(aux->ptr));
+ memcpy(o->ptr,aux->ptr,sdslen(aux->ptr));
+ decrRefCount(aux);
+
+ /* Fix the object encoding, and make sure to convert the encoded
+ * data type into the base type if accordingly to the current
+ * configuration there are too many elements in the encoded data
+ * type. Note that we only check the length and not max element
+ * size as this is an O(N) scan. Eventually everything will get
+ * converted. */
+ switch(type) {
+ case REDIS_HASH_ZIPMAP:
+ o->type = REDIS_HASH;
+ o->encoding = REDIS_ENCODING_ZIPMAP;
+ if (zipmapLen(o->ptr) > server.hash_max_zipmap_entries)
+ convertToRealHash(o);
+ break;
+ case REDIS_LIST_ZIPLIST:
+ o->type = REDIS_LIST;
+ o->encoding = REDIS_ENCODING_ZIPLIST;
+ if (ziplistLen(o->ptr) > server.list_max_ziplist_entries)
+ listTypeConvert(o,REDIS_ENCODING_LINKEDLIST);
+ break;
+ case REDIS_SET_INTSET:
+ o->type = REDIS_SET;
+ o->encoding = REDIS_ENCODING_INTSET;
+ if (intsetLen(o->ptr) > server.set_max_intset_entries)
+ setTypeConvert(o,REDIS_ENCODING_HT);
+ break;
+ case REDIS_ZSET_ZIPLIST:
+ o->type = REDIS_ZSET;
+ o->encoding = REDIS_ENCODING_ZIPLIST;
+ if (zsetLength(o) > server.zset_max_ziplist_entries)
+ zsetConvert(o,REDIS_ENCODING_SKIPLIST);
+ break;
+ default:
+ redisPanic("Unknown encoding");
+ break;
+ }
} else {
redisPanic("Unknown object type");
}
return o;
}
+/* Mark that we are loading in the global state and setup the fields
+ * needed to provide loading stats. */
+void startLoading(FILE *fp) {
+ struct stat sb;
+
+ /* Load the DB */
+ server.loading = 1;
+ server.loading_start_time = time(NULL);
+ if (fstat(fileno(fp), &sb) == -1) {
+ server.loading_total_bytes = 1; /* just to avoid division by zero */
+ } else {
+ server.loading_total_bytes = sb.st_size;
+ }
+}
+
+/* Refresh the loading progress info */
+void loadingProgress(off_t pos) {
+ server.loading_loaded_bytes = pos;
+}
+
+/* Loading finished */
+void stopLoading(void) {
+ server.loading = 0;
+}
+
int rdbLoad(char *filename) {
- FILE *fp;
uint32_t dbid;
int type, retval, rdbver;
- int swap_all_values = 0;
redisDb *db = server.db+0;
char buf[1024];
time_t expiretime, now = time(NULL);
+ long loops = 0;
+ FILE *fp;
+ rio rdb;
fp = fopen(filename,"r");
if (!fp) return REDIS_ERR;
return REDIS_ERR;
}
rdbver = atoi(buf+5);
- if (rdbver != 1) {
+ if (rdbver < 1 || rdbver > 2) {
fclose(fp);
redisLog(REDIS_WARNING,"Can't handle RDB format version %d",rdbver);
return REDIS_ERR;
}
+
+ startLoading(fp);
+ rdb = rioInitWithFile(fp);
while(1) {
robj *key, *val;
- int force_swapout;
-
expiretime = -1;
+
+ /* Serve the clients from time to time */
+ if (!(loops++ % 1000)) {
+ loadingProgress(rdb.tell(&rdb));
+ aeProcessEvents(server.el, AE_FILE_EVENTS|AE_DONT_WAIT);
+ }
+
/* Read type. */
- if ((type = rdbLoadType(fp)) == -1) goto eoferr;
+ if ((type = rdbLoadType(&rdb)) == -1) goto eoferr;
if (type == REDIS_EXPIRETIME) {
- if ((expiretime = rdbLoadTime(fp)) == -1) goto eoferr;
+ if ((expiretime = rdbLoadTime(&rdb)) == -1) goto eoferr;
/* We read the time so we need to read the object type again */
- if ((type = rdbLoadType(fp)) == -1) goto eoferr;
+ if ((type = rdbLoadType(&rdb)) == -1) goto eoferr;
}
if (type == REDIS_EOF) break;
/* Handle SELECT DB opcode as a special case */
if (type == REDIS_SELECTDB) {
- if ((dbid = rdbLoadLen(fp,NULL)) == REDIS_RDB_LENERR)
+ if ((dbid = rdbLoadLen(&rdb,NULL)) == REDIS_RDB_LENERR)
goto eoferr;
if (dbid >= (unsigned)server.dbnum) {
redisLog(REDIS_WARNING,"FATAL: Data file was created with a Redis server configured to handle more than %d databases. Exiting\n", server.dbnum);
continue;
}
/* Read key */
- if ((key = rdbLoadStringObject(fp)) == NULL) goto eoferr;
+ if ((key = rdbLoadStringObject(&rdb)) == NULL) goto eoferr;
/* Read value */
- if ((val = rdbLoadObject(type,fp)) == NULL) goto eoferr;
+ if ((val = rdbLoadObject(type,&rdb)) == NULL) goto eoferr;
/* Check if the key already expired */
if (expiretime != -1 && expiretime < now) {
decrRefCount(key);
/* Set the expire time if needed */
if (expiretime != -1) setExpire(db,key,expiretime);
- /* Handle swapping while loading big datasets when VM is on */
-
- /* If we detecter we are hopeless about fitting something in memory
- * we just swap every new key on disk. Directly...
- * Note that's important to check for this condition before resorting
- * to random sampling, otherwise we may try to swap already
- * swapped keys. */
- if (swap_all_values) {
- dictEntry *de = dictFind(db->dict,key->ptr);
-
- /* de may be NULL since the key already expired */
- if (de) {
- vmpointer *vp;
- val = dictGetEntryVal(de);
-
- if (val->refcount == 1 &&
- (vp = vmSwapObjectBlocking(val)) != NULL)
- dictGetEntryVal(de) = vp;
- }
- decrRefCount(key);
- continue;
- }
decrRefCount(key);
-
- /* Flush data on disk once 32 MB of additional RAM are used... */
- force_swapout = 0;
- if ((zmalloc_used_memory() - server.vm_max_memory) > 1024*1024*32)
- force_swapout = 1;
-
- /* If we have still some hope of having some value fitting memory
- * then we try random sampling. */
- if (!swap_all_values && server.vm_enabled && force_swapout) {
- while (zmalloc_used_memory() > server.vm_max_memory) {
- if (vmSwapOneObjectBlocking() == REDIS_ERR) break;
- }
- if (zmalloc_used_memory() > server.vm_max_memory)
- swap_all_values = 1; /* We are already using too much mem */
- }
}
fclose(fp);
+ stopLoading();
return REDIS_OK;
eoferr: /* unexpected end of file is handled here with a fatal exit */
}
/* A background saving child (BGSAVE) terminated its work. Handle this. */
-void backgroundSaveDoneHandler(int statloc) {
- int exitcode = WEXITSTATUS(statloc);
- int bysignal = WIFSIGNALED(statloc);
-
+void backgroundSaveDoneHandler(int exitcode, int bysignal) {
if (!bysignal && exitcode == 0) {
redisLog(REDIS_NOTICE,
"Background saving terminated with success");
redisLog(REDIS_WARNING, "Background saving error");
} else {
redisLog(REDIS_WARNING,
- "Background saving terminated by signal %d", WTERMSIG(statloc));
+ "Background saving terminated by signal %d", bysignal);
rdbRemoveTempFile(server.bgsavechildpid);
}
server.bgsavechildpid = -1;
+ server.bgsavethread = (pthread_t) -1;
+ server.bgsavethread_state = REDIS_BGSAVE_THREAD_UNACTIVE;
/* Possibly there are slaves waiting for a BGSAVE in order to be served
* (the first stage of SYNC is a bulk transfer of dump.rdb) */
updateSlavesWaitingBgsave(exitcode == 0 ? REDIS_OK : REDIS_ERR);
}
+
+void saveCommand(redisClient *c) {
+ if (server.bgsavechildpid != -1 || server.bgsavethread != (pthread_t)-1) {
+ addReplyError(c,"Background save already in progress");
+ return;
+ }
+ if (rdbSave(server.dbfilename) == REDIS_OK) {
+ addReply(c,shared.ok);
+ } else {
+ addReply(c,shared.err);
+ }
+}
+
+void bgsaveCommand(redisClient *c) {
+ if (server.bgsavechildpid != -1 || server.bgsavethread != (pthread_t)-1) {
+ addReplyError(c,"Background save already in progress");
+ return;
+ }
+ if (rdbSaveBackground(server.dbfilename) == REDIS_OK) {
+ addReplyStatus(c,"Background saving started");
+ } else {
+ addReply(c,shared.err);
+ }
+}
projects / redis.git / blobdiff