前言
在Redis中的LRU算法文中說到,LRU有一個缺陷,在如下情況下:
~~~~~A~~~~~A~~~~~A~~~~A~~~~~A~~~~~A~~|
~~B~~B~~B~~B~~B~~B~~B~~B~~B~~B~~B~~B~|
~~~~~~~~~~C~~~~~~~~~C~~~~~~~~~C~~~~~~|
~~~~~D~~~~~~~~~~D~~~~~~~~~D~~~~~~~~~D|
會將數(shù)據(jù)D誤認(rèn)為將來最有可能被訪問到的數(shù)據(jù)。
Redis作者曾想改進(jìn)LRU算法,但發(fā)現(xiàn)Redis的LRU算法受制于隨機(jī)采樣數(shù)maxmemory_samples,在maxmemory_samples等于10的情況下已經(jīng)很接近于理想的LRU算法性能,也就是說,LRU算法本身已經(jīng)很難再進(jìn)一步了。
于是,將思路回到原點,淘汰算法的本意是保留那些將來最有可能被再次訪問的數(shù)據(jù),而LRU算法只是預(yù)測最近被訪問的數(shù)據(jù)將來最有可能被訪問到。我們可以轉(zhuǎn)變思路,采用一種LFU(Least Frequently Used)算法,也就是最頻繁被訪問的數(shù)據(jù)將來最有可能被訪問到。在上面的情況中,根據(jù)訪問頻繁情況,可以確定保留優(yōu)先級:B>A>C=D。
Redis中的LFU思路
在LFU算法中,可以為每個key維護(hù)一個計數(shù)器。每次key被訪問的時候,計數(shù)器增大。計數(shù)器越大,可以約等于訪問越頻繁。
上述簡單算法存在兩個問題:
- 在LRU算法中可以維護(hù)一個雙向鏈表,然后簡單的把被訪問的節(jié)點移至鏈表開頭,但在LFU中是不可行的,節(jié)點要嚴(yán)格按照計數(shù)器進(jìn)行排序,新增節(jié)點或者更新節(jié)點位置時,時間復(fù)雜度可能達(dá)到O(N)。
- 只是簡單的增加計數(shù)器的方法并不完美。訪問模式是會頻繁變化的,一段時間內(nèi)頻繁訪問的key一段時間之后可能會很少被訪問到,只增加計數(shù)器并不能體現(xiàn)這種趨勢。
第一個問題很好解決,可以借鑒LRU實現(xiàn)的經(jīng)驗,維護(hù)一個待淘汰key的pool。第二個問題的解決辦法是,記錄key最后一個被訪問的時間,然后隨著時間推移,降低計數(shù)器。
Redis對象的結(jié)構(gòu)如下:
typedef struct redisObject {
unsigned type:4;
unsigned encoding:4;
unsigned lru:LRU_BITS; /* LRU time (relative to global lru_clock) or
* LFU data (least significant 8 bits frequency
* and most significant 16 bits access time). */
int refcount;
void *ptr;
} robj;
在LRU算法中,24 bits的lru是用來記錄LRU time的,在LFU中也可以使用這個字段,不過是分成16 bits與8 bits使用:
16 bits 8 bits
+----------------+--------+
+ Last decr time | LOG_C |
+----------------+--------+
高16 bits用來記錄最近一次計數(shù)器降低的時間ldt,單位是分鐘,低8 bits記錄計數(shù)器數(shù)值counter。
LFU配置
Redis4.0之后為maxmemory_policy淘汰策略添加了兩個LFU模式:
- volatile-lfu:對有過期時間的key采用LFU淘汰算法
- allkeys-lfu:對全部key采用LFU淘汰算法
還有2個配置可以調(diào)整LFU算法:
lfu-log-factor 10
lfu-decay-time 1
lfu-log-factor可以調(diào)整計數(shù)器counter的增長速度,lfu-log-factor越大,counter增長的越慢。
lfu-decay-time是一個以分鐘為單位的數(shù)值,可以調(diào)整counter的減少速度
源碼實現(xiàn)
在lookupKey中:
robj *lookupKey(redisDb *db, robj *key, int flags) {
dictEntry *de = dictFind(db->dict,key->ptr);
if (de) {
robj *val = dictGetVal(de);
/* Update the access time for the ageing algorithm.
* Don't do it if we have a saving child, as this will trigger
* a copy on write madness. */
if (server.rdb_child_pid == -1
server.aof_child_pid == -1
!(flags LOOKUP_NOTOUCH))
{
if (server.maxmemory_policy MAXMEMORY_FLAG_LFU) {
updateLFU(val);
} else {
val->lru = LRU_CLOCK();
}
}
return val;
} else {
return NULL;
}
}
當(dāng)采用LFU策略時,updateLFU更新lru:
/* Update LFU when an object is accessed.
* Firstly, decrement the counter if the decrement time is reached.
* Then logarithmically increment the counter, and update the access time. */
void updateLFU(robj *val) {
unsigned long counter = LFUDecrAndReturn(val);
counter = LFULogIncr(counter);
val->lru = (LFUGetTimeInMinutes()8) | counter;
}
降低LFUDecrAndReturn
首先,LFUDecrAndReturn對counter進(jìn)行減少操作:
/* If the object decrement time is reached decrement the LFU counter but
* do not update LFU fields of the object, we update the access time
* and counter in an explicit way when the object is really accessed.
* And we will times halve the counter according to the times of
* elapsed time than server.lfu_decay_time.
* Return the object frequency counter.
*
* This function is used in order to scan the dataset for the best object
* to fit: as we check for the candidate, we incrementally decrement the
* counter of the scanned objects if needed. */
unsigned long LFUDecrAndReturn(robj *o) {
unsigned long ldt = o->lru >> 8;
unsigned long counter = o->lru 255;
unsigned long num_periods = server.lfu_decay_time ? LFUTimeElapsed(ldt) / server.lfu_decay_time : 0;
if (num_periods)
counter = (num_periods > counter) ? 0 : counter - num_periods;
return counter;
}
函數(shù)首先取得高16 bits的最近降低時間ldt與低8 bits的計數(shù)器counter,然后根據(jù)配置的lfu_decay_time計算應(yīng)該降低多少。
LFUTimeElapsed用來計算當(dāng)前時間與ldt的差值:
/* Return the current time in minutes, just taking the least significant
* 16 bits. The returned time is suitable to be stored as LDT (last decrement
* time) for the LFU implementation. */
unsigned long LFUGetTimeInMinutes(void) {
return (server.unixtime/60) 65535;
}
/* Given an object last access time, compute the minimum number of minutes
* that elapsed since the last access. Handle overflow (ldt greater than
* the current 16 bits minutes time) considering the time as wrapping
* exactly once. */
unsigned long LFUTimeElapsed(unsigned long ldt) {
unsigned long now = LFUGetTimeInMinutes();
if (now >= ldt) return now-ldt;
return 65535-ldt+now;
}
具體是當(dāng)前時間轉(zhuǎn)化成分鐘數(shù)后取低16 bits,然后計算與ldt的差值now-ldt。當(dāng)ldt > now時,默認(rèn)為過了一個周期(16 bits,最大65535),取值65535-ldt+now。
然后用差值與配置lfu_decay_time相除,LFUTimeElapsed(ldt) / server.lfu_decay_time,已過去n個lfu_decay_time,則將counter減少n,counter - num_periods。
增長LFULogIncr
增長函數(shù)LFULogIncr如下:
/* Logarithmically increment a counter. The greater is the current counter value
* the less likely is that it gets really implemented. Saturate it at 255. */
uint8_t LFULogIncr(uint8_t counter) {
if (counter == 255) return 255;
double r = (double)rand()/RAND_MAX;
double baseval = counter - LFU_INIT_VAL;
if (baseval 0) baseval = 0;
double p = 1.0/(baseval*server.lfu_log_factor+1);
if (r p) counter++;
return counter;
}
counter并不是簡單的訪問一次就+1,而是采用了一個0-1之間的p因子控制增長。counter最大值為255。取一個0-1之間的隨機(jī)數(shù)r與p比較,當(dāng)rp時,才增加counter,這和比特幣中控制產(chǎn)出的策略類似。p取決于當(dāng)前counter值與lfu_log_factor因子,counter值與lfu_log_factor因子越大,p越小,rp的概率也越小,counter增長的概率也就越小。增長情況如下:
+--------+------------+------------+------------+------------+------------+
| factor | 100 hits | 1000 hits | 100K hits | 1M hits | 10M hits |
+--------+------------+------------+------------+------------+------------+
| 0 | 104 | 255 | 255 | 255 | 255 |
+--------+------------+------------+------------+------------+------------+
| 1 | 18 | 49 | 255 | 255 | 255 |
+--------+------------+------------+------------+------------+------------+
| 10 | 10 | 18 | 142 | 255 | 255 |
+--------+------------+------------+------------+------------+------------+
| 100 | 8 | 11 | 49 | 143 | 255 |
+--------+------------+------------+------------+------------+------------+
可見counter增長與訪問次數(shù)呈現(xiàn)對數(shù)增長的趨勢,隨著訪問次數(shù)越來越大,counter增長的越來越慢。
新生key策略
另外一個問題是,當(dāng)創(chuàng)建新對象的時候,對象的counter如果為0,很容易就會被淘汰掉,還需要為新生key設(shè)置一個初始counter,createObject:
robj *createObject(int type, void *ptr) {
robj *o = zmalloc(sizeof(*o));
o->type = type;
o->encoding = OBJ_ENCODING_RAW;
o->ptr = ptr;
o->refcount = 1;
/* Set the LRU to the current lruclock (minutes resolution), or
* alternatively the LFU counter. */
if (server.maxmemory_policy MAXMEMORY_FLAG_LFU) {
o->lru = (LFUGetTimeInMinutes()8) | LFU_INIT_VAL;
} else {
o->lru = LRU_CLOCK();
}
return o;
}
counter會被初始化為LFU_INIT_VAL,默認(rèn)5。
pool
pool算法就與LRU算法一致了:
if (server.maxmemory_policy (MAXMEMORY_FLAG_LRU|MAXMEMORY_FLAG_LFU) ||
server.maxmemory_policy == MAXMEMORY_VOLATILE_TTL)
計算idle時有所不同:
} else if (server.maxmemory_policy MAXMEMORY_FLAG_LFU) {
/* When we use an LRU policy, we sort the keys by idle time
* so that we expire keys starting from greater idle time.
* However when the policy is an LFU one, we have a frequency
* estimation, and we want to evict keys with lower frequency
* first. So inside the pool we put objects using the inverted
* frequency subtracting the actual frequency to the maximum
* frequency of 255. */
idle = 255-LFUDecrAndReturn(o);
使用了255-LFUDecrAndReturn(o)當(dāng)做排序的依據(jù)。
參考鏈接
- Random notes on improving the Redis LRU algorithm
- Using Redis as an LRU cache
總結(jié)
以上就是這篇文章的全部內(nèi)容了,希望本文的內(nèi)容對大家的學(xué)習(xí)或者工作具有一定的參考學(xué)習(xí)價值,謝謝大家對腳本之家的支持。
您可能感興趣的文章:- C++ 實現(xiàn)LRU 與 LFU 的緩存算法
- C++泛型編程基本概念詳解
- C++算法與泛型算法(algorithm、numeric)
- C++ 泛型編程詳解
- C++實現(xiàn)支持泛型的LFU詳解