隨著nosql風(fēng)潮興起，redis作為當(dāng)中一個(gè)耀眼的明星，也越來(lái)越多的被關(guān)注和使用，我在工作中也廣泛的用到了redis來(lái)充當(dāng)cache和key-value DB，但當(dāng)大家發(fā)現(xiàn)數(shù)據(jù)越來(lái)越多時(shí)，不禁有些擔(dān)心，redis能撐的住嗎，雖然官方已經(jīng)有漂亮的benchmark，自己也可以做做壓力測(cè)試，但是看看源碼，也是確認(rèn)問(wèn)題最直接的辦法之一。比如目前我們要確認(rèn)的一個(gè)問(wèn)題是，redis是如何刪除過(guò)期數(shù)據(jù)的？
用一個(gè)可以"find reference"的IDE,沿著setex(Set the value and expiration of a key)命令一窺究竟：
void setexCommand(redisClient *c) {
c->argv[3] = tryObjectEncoding(c->argv[3]);
setGenericCommand(c,0,c->argv[1],c->argv[3],c->argv[2]);
}
setGenericCommand是一個(gè)實(shí)現(xiàn)set,setnx,setex的通用函數(shù)，參數(shù)設(shè)置不同而已。
void setCommand(redisClient *c) {
c->argv[2] = tryObjectEncoding(c->argv[2]);
setGenericCommand(c,0,c->argv[1],c->argv[2],NULL);
}
void setnxCommand(redisClient *c) {
c->argv[2] = tryObjectEncoding(c->argv[2]);
setGenericCommand(c,1,c->argv[1],c->argv[2],NULL);
}
void setexCommand(redisClient *c) {
c->argv[3] = tryObjectEncoding(c->argv[3]);
setGenericCommand(c,0,c->argv[1],c->argv[3],c->argv[2]);
}
再看setGenericCommand：
1 void setGenericCommand(redisClient *c, int nx, robj *key, robj *val, robj *expire) {
2 long seconds = 0; /* initialized to avoid an harmness warning */
3
4 if (expire) {
5 if (getLongFromObjectOrReply(c, expire, &seconds, NULL) != REDIS_OK)
6 return;
7 if (seconds <= 0) {
8 addReplyError(c,"invalid expire time in SETEX");
9 return;
10 }
11 }
12
13 if (lookupKeyWrite(c->db,key) != NULL && nx) {
14 addReply(c,shared.czero);
15 return;
16 }
17 setKey(c->db,key,val);
18 server.dirty++;
19 if (expire) setExpire(c->db,key,time(NULL)+seconds);
20 addReply(c, nx ? shared.cone : shared.ok);
21 }
22
13行處理"Set the value of a key, only if the key does not exist"的場(chǎng)景，17行插入這個(gè)key，19行設(shè)置它的超時(shí)，注意時(shí)間戳已經(jīng)被設(shè)置成了到期時(shí)間。這里要看一下redisDb(即c->db)的定義：
typedef struct redisDb {
dict *dict; /* The keyspace for this DB */
dict *expires; /* Timeout of keys with a timeout set */
dict *blocking_keys; /* Keys with clients waiting for data (BLPOP) */
dict *io_keys; /* Keys with clients waiting for VM I/O */
dict *watched_keys; /* WATCHED keys for MULTI/EXEC CAS */
int id;
} redisDb;
僅關(guān)注dict和expires，分別來(lái)存key-value和它的超時(shí)，也就是說(shuō)如果一個(gè)key-value是有超時(shí)的，那么它會(huì)存在dict里，同時(shí)也存到expires里，類(lèi)似這樣的形式：dict[key]:value,expires[key]:timeout.
當(dāng)然key-value沒(méi)有超時(shí)，expires里就不存在這個(gè)key。剩下setKey和setExpire兩個(gè)函數(shù)無(wú)非是插數(shù)據(jù)到兩個(gè)字典里，這里不再詳述。
那么redis是如何刪除過(guò)期key的呢。
通過(guò)查看dbDelete的調(diào)用者，首先注意到這一個(gè)函數(shù)，是用來(lái)刪除過(guò)期key的。
1 int expireIfNeeded(redisDb *db, robj *key) {
2 time_t when = getExpire(db,key);
3
4 if (when < 0) return 0; /* No expire for this key */
5
6 /* Don't expire anything while loading. It will be done later. */
7 if (server.loading) return 0;
8
9 /* If we are running in the context of a slave, return ASAP:
10 * the slave key expiration is controlled by the master that will
11 * send us synthesized DEL operations for expired keys.
12 *
13 * Still we try to return the right information to the caller,
14 * that is, 0 if we think the key should be still valid, 1 if
15 * we think the key is expired at this time. */
16 if (server.masterhost != NULL) {
17 return time(NULL) > when;
18 }
19
20 /* Return when this key has not expired */
21 if (time(NULL) <= when) return 0;
22
23 /* Delete the key */
24 server.stat_expiredkeys++;
25 propagateExpire(db,key);
26 return dbDelete(db,key);
27 }
28
ifNeed表示能刪則刪，所以4行沒(méi)有設(shè)置超時(shí)不刪，7行在"loading"時(shí)不刪，16行非主庫(kù)不刪，21行未到期不刪。25行同步從庫(kù)和文件。
再看看哪些函數(shù)調(diào)用了expireIfNeeded，有lookupKeyRead，lookupKeyWrite，dbRandomKey，existsCommand，keysCommand。通過(guò)這些函數(shù)命名可以看出，只要訪問(wèn)了某一個(gè)key，順帶做的事情就是嘗試查看過(guò)期并刪除，這就保證了用戶(hù)不可能訪問(wèn)到過(guò)期的key。但是如果有大量的key過(guò)期，并且沒(méi)有被訪問(wèn)到，那么就浪費(fèi)了許多內(nèi)存。Redis是如何處理這個(gè)問(wèn)題的呢。
dbDelete的調(diào)用者里還發(fā)現(xiàn)這樣一個(gè)函數(shù)：
1 /* Try to expire a few timed out keys. The algorithm used is adaptive and
2 * will use few CPU cycles if there are few expiring keys, otherwise
3 * it will get more aggressive to avoid that too much memory is used by
4 * keys that can be removed from the keyspace. */
5 void activeExpireCycle(void) {
6 int j;
7
8 for (j = 0; j < server.dbnum; j++) {
9 int expired;
10 redisDb *db = server.db+j;
11
12 /* Continue to expire if at the end of the cycle more than 25%
13 * of the keys were expired. */
14 do {
15 long num = dictSize(db->expires);
16 time_t now = time(NULL);
17
18 expired = 0;
19 if (num > REDIS_EXPIRELOOKUPS_PER_CRON)
20 num = REDIS_EXPIRELOOKUPS_PER_CRON;
21 while (num--) {
22 dictEntry *de;
23 time_t t;
24
25 if ((de = dictGetRandomKey(db->expires)) == NULL) break;
26 t = (time_t) dictGetEntryVal(de);
27 if (now > t) {
28 sds key = dictGetEntryKey(de);
29 robj *keyobj = createStringObject(key,sdslen(key));
30
31 propagateExpire(db,keyobj);
32 dbDelete(db,keyobj);
33 decrRefCount(keyobj);
34 expired++;
35 server.stat_expiredkeys++;
36 }
37 }
38 } while (expired > REDIS_EXPIRELOOKUPS_PER_CRON/4);
39 }
40 }
41
這個(gè)函數(shù)的意圖已經(jīng)有說(shuō)明：刪一點(diǎn)點(diǎn)過(guò)期key，如果過(guò)期key較少，那也只用一點(diǎn)點(diǎn)cpu。25行隨機(jī)取一個(gè)key，38行刪key成功的概率較低就退出。這個(gè)函數(shù)被放在一個(gè)cron里，每毫秒被調(diào)用一次。這個(gè)算法保證每次會(huì)刪除一定比例的key，但是如果key總量很大，而這個(gè)比例控制的太大，就需要更多次的循環(huán)，浪費(fèi)cpu，控制的太小，過(guò)期的key就會(huì)變多，浪費(fèi)內(nèi)存——這就是時(shí)空權(quán)衡了。
最后在dbDelete的調(diào)用者里還發(fā)現(xiàn)這樣一個(gè)函數(shù)：
/* This function gets called when 'maxmemory' is set on the config file to limit
* the max memory used by the server, and we are out of memory.
* This function will try to, in order:
*
* - Free objects from the free list
* - Try to remove keys with an EXPIRE set
*
* It is not possible to free enough memory to reach used-memory < maxmemory
* the server will start refusing commands that will enlarge even more the
* memory usage.
*/
void freeMemoryIfNeeded(void)
這個(gè)函數(shù)太長(zhǎng)就不再詳述了，注釋部分說(shuō)明只有在配置文件中設(shè)置了最大內(nèi)存時(shí)候才會(huì)調(diào)用這個(gè)函數(shù)，而設(shè)置這個(gè)參數(shù)的意義是，你把redis當(dāng)做一個(gè)內(nèi)存cache而不是key-value數(shù)據(jù)庫(kù)。
以上3種刪除過(guò)期key的途徑，第二種定期刪除一定比例的key是主要的刪除途徑，第一種“讀時(shí)刪除”保證過(guò)期key不會(huì)被訪問(wèn)到，第三種是一個(gè)當(dāng)內(nèi)存超出設(shè)定時(shí)的暴力手段。由此也能看出redis設(shè)計(jì)的巧妙之處，