曹工说Redis源码（8）--面试时，redis 内存淘汰总被问，但是总答不好

Copy
# Don't use more memory than the specified amount of bytes.
# When the memory limit is reached Redis will try to remove keys
# according to the eviction policy selected (see maxmemory-policy).
#
# If Redis can't remove keys according to the policy, or if the policy is
# set to 'noeviction', Redis will start to reply with errors to commands
# that would use more memory, like SET, LPUSH, and so on, and will continue
# to reply to read-only commands like GET.
#
# This option is usually useful when using Redis as an LRU cache, or to set
# a hard memory limit for an instance (using the 'noeviction' policy).
#
# WARNING: If you have slaves attached to an instance with maxmemory on,
# the size of the output buffers needed to feed the slaves are subtracted
# from the used memory count, so that network problems / resyncs will
# not trigger a loop where keys are evicted, and in turn the output
# buffer of slaves is full with DELs of keys evicted triggering the deletion
# of more keys, and so forth until the database is completely emptied.
#
# In short... if you have slaves attached it is suggested that you set a lower
# limit for maxmemory so that there is some free RAM on the system for slave
# output buffers (but this is not needed if the policy is 'noeviction').
#
# maxmemory <bytes>

渣翻译如下：

不能使用超过指定数量bytes的内存。当该内存限制被达到时，redis会根据过期策略（eviction policy，通过参数 maxmemory-policy来指定）来驱逐key。

如果redis根据指定的策略，或者策略被设置为“noeviction”，redis会开始针对如下这种命令，回复错误。什么命令呢？会使用更多内存的那类命令，比如set、lpush；只读命令还是不受影响，可以正常响应。

该选项通常在redis使用LRU缓存时有用，或者在使用noeviction策略时，设置一个进程级别的内存limit。

内存淘汰策略#

所谓策略，意思是，当我们要删除部分key的时候，删哪些，不删哪些？是不是需要一个策略？比如是随机删，就像灭霸一样？还是按照lru时间来删，lru的策略意思就是，最近最少使用的key，将被优先删除。

总之，我们需要定一个规则。

redis默认支持以下策略：

Copy
# MAXMEMORY POLICY: how Redis will select what to remove when maxmemory
# is reached. You can select among five behaviors:
# 
# volatile-lru -> remove the key with an expire set using an LRU algorithm
# allkeys-lru -> remove any key accordingly to the LRU algorithm
# volatile-random -> remove a random key with an expire set
# allkeys-random -> remove a random key, any key
# volatile-ttl -> remove the key with the nearest expire time (minor TTL)
# noeviction -> don't expire at all, just return an error on write operations
# 
# Note: with any of the above policies, Redis will return an error on write
#       operations, when there are not suitable keys for eviction.
#
#       At the date of writing this commands are: set setnx setex append
#       incr decr rpush lpush rpushx lpushx linsert lset rpoplpush sadd
#       sinter sinterstore sunion sunionstore sdiff sdiffstore zadd zincrby
#       zunionstore zinterstore hset hsetnx hmset hincrby incrby decrby
#       getset mset msetnx exec sort
#
# The default is:
#
# maxmemory-policy noeviction
maxmemory-policy allkeys-lru

Copy
针对设置了过期时间的，使用lru算法
# volatile-lru -> remove the key with an expire set using an LRU algorithm

针对全部key，使用lru算法
# allkeys-lru -> remove any key accordingly to the LRU algorithm

针对设置了过期时间的，随机删
# volatile-random -> remove a random key with an expire set

针对全部key，随机删
# allkeys-random -> remove a random key, any key

针对设置了过期时间的，马上要过期的，删掉
# volatile-ttl -> remove the key with the nearest expire time (minor TTL)

不过期，不能写了，就报错
# noeviction -> don't expire at all, just return an error on write operations

一般呢，我们会设置为：

allkeys-lru，即，针对全部key，进行lru。

源码实现#

配置读取#

在如下结构体中，定义了如下字段：

Copy
struct redisServer {
	...
	unsigned long long maxmemory;   /* Max number of memory bytes to use */
    int maxmemory_policy;           /* Policy for key eviction */
    int maxmemory_samples;          /* Pricision of random sampling */
    ...
}

当我们在配置文件中，进入如下配置时，该结构体中几个字段的值如下：

Copy
maxmemory 3mb
maxmemory-policy allkeys-lru
# maxmemory-samples 5  这个取了默认值

maxmemory_policy为3，是因为枚举值为3：

Copy
#define REDIS_MAXMEMORY_VOLATILE_LRU 0
#define REDIS_MAXMEMORY_VOLATILE_TTL 1
#define REDIS_MAXMEMORY_VOLATILE_RANDOM 2
#define REDIS_MAXMEMORY_ALLKEYS_LRU 3
#define REDIS_MAXMEMORY_ALLKEYS_RANDOM 4
#define REDIS_MAXMEMORY_NO_EVICTION 5
#define REDIS_DEFAULT_MAXMEMORY_POLICY REDIS_MAXMEMORY_NO_EVICTION

处理命令时，判断是否进行内存淘汰#

在处理命令的时候，会调用中的

Copy
redis.c  processCommand
    
int processCommand(redisClient *c) {
    /* The QUIT command is handled separately. Normal command procs will
     * go through checking for replication and QUIT will cause trouble
     * when FORCE_REPLICATION is enabled and would be implemented in
     * a regular command proc. */
    // 特别处理 quit 命令
    void *commandName = c->argv[0]->ptr;
    redisLog(REDIS_NOTICE, "The server is now processing %s", commandName);

    if (!strcasecmp(c->argv[0]->ptr, "quit")) {
        addReply(c, shared.ok);
        c->flags |= REDIS_CLOSE_AFTER_REPLY;
        return REDIS_ERR;
    }

    /* Now lookup the command and check ASAP about trivial error conditions
     * such as wrong arity, bad command name and so forth. */
    // 1 查找命令，并进行命令合法性检查，以及命令参数个数检查
    c->cmd = c->lastcmd = lookupCommand(c->argv[0]->ptr);
    if (!c->cmd) {
        // 没找到指定的命令
        flagTransaction(c);
        addReplyErrorFormat(c, "unknown command '%s'",
                            (char *) c->argv[0]->ptr);
        return REDIS_OK;
    }

    /* Check if the user is authenticated */
    //2 检查认证信息
    if (server.requirepass && !c->authenticated && c->cmd->proc != authCommand) {
        flagTransaction(c);
        addReply(c, shared.noautherr);
        return REDIS_OK;
    }

    /* If cluster is enabled perform the cluster redirection here.
     *
     * 3 如果开启了集群模式，那么在这里进行转向操作。
     *
     * However we don't perform the redirection if:
     *
     * 不过，如果有以下情况出现，那么节点不进行转向：
     *
     * 1) The sender of this command is our master.
     *    命令的发送者是本节点的主节点
     *
     * 2) The command has no key arguments. 
     *    命令没有 key 参数
     */
    if (server.cluster_enabled &&
        !(c->flags & REDIS_MASTER) &&
        !(c->cmd->getkeys_proc == NULL && c->cmd->firstkey == 0)) {
        int hashslot;

        // 集群已下线
        if (server.cluster->state != REDIS_CLUSTER_OK) {
            flagTransaction(c);
            addReplySds(c, sdsnew("-CLUSTERDOWN The cluster is down. Use CLUSTER INFO for more information\r\n"));
            return REDIS_OK;

            // 集群运作正常
        } else {
            int error_code;
            clusterNode *n = getNodeByQuery(c, c->cmd, c->argv, c->argc, &hashslot, &error_code);
            // 不能执行多键处理命令
            if (n == NULL) {
                flagTransaction(c);
                if (error_code == REDIS_CLUSTER_REDIR_CROSS_SLOT) {
                    addReplySds(c, sdsnew("-CROSSSLOT Keys in request don't hash to the same slot\r\n"));
                } else if (error_code == REDIS_CLUSTER_REDIR_UNSTABLE) {
                    /* The request spawns mutliple keys in the same slot,
                     * but the slot is not "stable" currently as there is
                     * a migration or import in progress. */
                    addReplySds(c, sdsnew("-TRYAGAIN Multiple keys request during rehashing of slot\r\n"));
                } else {
                    redisPanic("getNodeByQuery() unknown error.");
                }
                return REDIS_OK;

                //3.1 命令针对的槽和键不是本节点处理的，进行转向
            } else if (n != server.cluster->myself) {
                flagTransaction(c);
                // -<ASK or MOVED> <slot> <ip>:<port>
                // 例如 -ASK 10086 127.0.0.1:12345
                addReplySds(c, sdscatprintf(sdsempty(),
                                            "-%s %d %s:%d\r\n",
                                            (error_code == REDIS_CLUSTER_REDIR_ASK) ? "ASK" : "MOVED",
                                            hashslot, n->ip, n->port));

                return REDIS_OK;
            }

            // 如果执行到这里，说明键 key 所在的槽由本节点处理
            // 或者客户端执行的是无参数命令
        }
    }

    /* Handle the maxmemory directive.
     *
     * First we try to free some memory if possible (if there are volatile
     * keys in the dataset). If there are not the only thing we can do
     * is returning an error. */
    //4 如果设置了最大内存，那么检查内存是否超过限制，并做相应的操作
    if (server.maxmemory) {
        //4.1 如果内存已超过限制，那么尝试通过删除过期键来释放内存
        int retval = freeMemoryIfNeeded();
        // 如果即将要执行的命令可能占用大量内存（REDIS_CMD_DENYOOM）
        // 并且前面的内存释放失败的话
        // 那么向客户端返回内存错误
        if ((c->cmd->flags & REDIS_CMD_DENYOOM) && retval == REDIS_ERR) {
            flagTransaction(c);
            addReply(c, shared.oomerr);
            return REDIS_OK;
        }
    }    
    ....

1处，查找命令，对应的函数指针（类似于java里的策略模式，根据命令，找对应的策略）
2处，检查，是否密码正确
3处，集群相关操作；
3.1处，不是本节点处理，直接返回ask，指示客户端转向
4处，判断是否设置了maxMemory，这里就是本文重点：设置了maxMemory时，内存淘汰策略
4.1处，调用了下方的 freeMemoryIfNeeded

接下来，深入4.1处：

Copy

int freeMemoryIfNeeded(void) {
    size_t mem_used, mem_tofree, mem_freed;
    int slaves = listLength(server.slaves);

    /* Remove the size of slaves output buffers and AOF buffer from the
     * count of used memory. */
    // 计算出 Redis 目前占用的内存总数，但有两个方面的内存不会计算在内：
    // 1）从服务器的输出缓冲区的内存
    // 2）AOF 缓冲区的内存
    mem_used = zmalloc_used_memory();
    if (slaves) {
		...
    }
    if (server.aof_state != REDIS_AOF_OFF) {
        mem_used -= sdslen(server.aof_buf);
        mem_used -= aofRewriteBufferSize();
    }

    /* Check if we are over the memory limit. */
    //1 如果目前使用的内存大小比设置的 maxmemory 要小，那么无须执行进一步操作
    if (mem_used <= server.maxmemory) return REDIS_OK;

    //2 如果占用内存比 maxmemory 要大，但是 maxmemory 策略为不淘汰，那么直接返回
    if (server.maxmemory_policy == REDIS_MAXMEMORY_NO_EVICTION)
        return REDIS_ERR; /* We need to free memory, but policy forbids. */

    /* Compute how much memory we need to free. */
    // 3 计算需要释放多少字节的内存
    mem_tofree = mem_used - server.maxmemory;

    // 初始化已释放内存的字节数为 0
    mem_freed = 0;

    // 根据 maxmemory 策略，
    //4 遍历字典，释放内存并记录被释放内存的字节数
    while (mem_freed < mem_tofree) {
        int j, k, keys_freed = 0;

        // 遍历所有字典
        for (j = 0; j < server.dbnum; j++) {
            long bestval = 0; /* just to prevent warning */
            sds bestkey = NULL;
            dictEntry *de;
            redisDb *db = server.db + j;
            dict *dict;

            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM) {
                // 如果策略是 allkeys-lru 或者 allkeys-random 
                //5 那么淘汰的目标为所有数据库键
                dict = server.db[j].dict;
            } else {
                // 如果策略是 volatile-lru 、 volatile-random 或者 volatile-ttl 
                //6 那么淘汰的目标为带过期时间的数据库键
                dict = server.db[j].expires;
            }


            /* volatile-random and allkeys-random policy */
            // 如果使用的是随机策略，那么从目标字典中随机选出键
            if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_RANDOM ||
                server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_RANDOM) {
                de = dictGetRandomKey(dict);
                bestkey = dictGetKey(de);
            }
            /* volatile-lru and allkeys-lru policy */
            //7 
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_ALLKEYS_LRU ||
                     server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_LRU) {
                struct evictionPoolEntry *pool = db->eviction_pool;

                while (bestkey == NULL) {
                    // 8 
                    evictionPoolPopulate(dict, db->dict, db->eviction_pool);
                    /* Go backward from best to worst element to evict. */
                    for (k = REDIS_EVICTION_POOL_SIZE - 1; k >= 0; k--) {
                        if (pool[k].key == NULL) continue;
                        // 8.1
                        de = dictFind(dict, pool[k].key);

                        /* 8.2 Remove the entry from the pool. */
                        sdsfree(pool[k].key);
                        /* Shift all elements on its right to left. */
                        memmove(pool + k, pool + k + 1,
                                sizeof(pool[0]) * (REDIS_EVICTION_POOL_SIZE - k - 1));
                        /* Clear the element on the right which is empty
                         * since we shifted one position to the left.  */
                        pool[REDIS_EVICTION_POOL_SIZE - 1].key = NULL;
                        pool[REDIS_EVICTION_POOL_SIZE - 1].idle = 0;

                        /* If the key exists, is our pick. Otherwise it is
                         * a ghost and we need to try the next element. */
                        // 8.3
                        if (de) {
                            bestkey = dictGetKey(de);
                            break;
                        } else {
                            /* Ghost... */
                            continue;
                        }
                    }
                }
            }

                /* volatile-ttl */
                // 策略为 volatile-ttl ，从一集 sample 键中选出过期时间距离当前时间最接近的键
            else if (server.maxmemory_policy == REDIS_MAXMEMORY_VOLATILE_TTL) {
                ...
            }

            /* Finally remove the selected key. */
            // 8.4 删除被选中的键
            if (bestkey) {
                long long delta;

                robj *keyobj = createStringObject(bestkey, sdslen(bestkey));
                propagateExpire(db, keyobj);
                /* We compute the amount of memory freed by dbDelete() alone.
                 * It is possible that actually the memory needed to propagate
                 * the DEL in AOF and replication link is greater than the one
                 * we are freeing removing the key, but we can't account for
                 * that otherwise we would never exit the loop.
                 *
                 * AOF and Output buffer memory will be freed eventually so
                 * we only care about memory used by the key space. */
                // 计算删除键所释放的内存数量
                delta = (long long) zmalloc_used_memory();
                dbDelete(db, keyobj);
                delta -= (long long) zmalloc_used_memory();
                mem_freed += delta;

                // 对淘汰键的计数器增一
                server.stat_evictedkeys++;

                notifyKeyspaceEvent(REDIS_NOTIFY_EVICTED, "evicted",
                                    keyobj, db->id);
                decrRefCount(keyobj);
                keys_freed++;
				...
            }
        }

        if (!keys_freed) return REDIS_ERR; /* nothing to free... */
    }

    return REDIS_OK;
}

1处，如果目前使用的内存大小比设置的 maxmemory 要小，那么无须执行进一步操作
2处，如果占用内存比 maxmemory 要大，但是 maxmemory 策略为不淘汰，那么直接返回
3处，计算需要释放多少字节的内存
4处，遍历字典，释放内存并记录被释放内存的字节数
5处，如果策略是 allkeys-lru 或者 allkeys-random 那么淘汰的目标为所有数据库键
6处，如果策略是 volatile-lru 、 volatile-random 或者 volatile-ttl ，那么淘汰的目标为带过期时间的数据库键
7处，如果使用的是 LRU 策略，那么从 sample 键中选出 IDLE 时间最长的那个键
8处，调用evictionPoolPopulate，该函数在下面讲解,该函数的功能是，传入一个链表，即这里的db->eviction_pool，然后在函数内部，随机找出n个key，放入传入的链表中，并按照空闲时间排序，空闲最久的，放到最后。

当该函数，返回后，db->eviction_pool这个链表里就存放了我们要淘汰的key。
8.1处，找到这个key，这个key，在后边会被删除
8.2处，下面这一段，从db->eviction_pool将这个已经处理了的key删掉
8.3处，如果这个key，是存在的，则跳出循环，在后面8.4处，会被删除
8.4处，删除这个key

选择哪些key作为被淘汰的key#

前面我们看到，在7处，如果为lru策略，则会进入8处的函数：

evictionPoolPopulate。

该函数的名称为：填充(populate)驱逐(eviction)对象池(pool)。驱逐的意思，就是现在达到了maxmemory，没办法，只能开始删除掉一部分元素，来腾空间了，不然新的put类型的命令，根本没办法执行。

该方法的大概思路是，使用lru的时候，随机找n个key，类似于抽样，然后放到一个链表，根据空闲时间排序。

具体看看该方法的实现：

Copy
void evictionPoolPopulate(dict *sampledict, dict *keydict, struct evictionPoolEntry *pool) {

其中，传入的第三个参数，是要被填充的对象，在c语言中，习惯传入一个入参，然后在函数内部填充或者修改入参对象的属性。

该属性，就是前面说的那个链表，用来存放收集的随机的元素，该链表中节点的结构如下：

Copy
struct evictionPoolEntry {
    unsigned long long idle;    /* Object idle time. */
    sds key;                    /* Key name. */
};

该结构共2个字段，一个存储key，一个存储空闲时间。

该链表中，共maxmemory-samples个元素，会按照idle时间长短排序，idle时间长的在链表尾部，（假设头在左，尾在右）。

Copy
void evictionPoolPopulate(dict *sampledict, dict *keydict, struct evictionPoolEntry *pool) {
    int j, k, count;
    dictEntry *_samples[EVICTION_SAMPLES_ARRAY_SIZE];
    dictEntry **samples;

    /* Try to use a static buffer: this function is a big hit...
     * Note: it was actually measured that this helps. */
    if (server.maxmemory_samples <= EVICTION_SAMPLES_ARRAY_SIZE) {
        samples = _samples;
    } else {
        samples = zmalloc(sizeof(samples[0]) * server.maxmemory_samples);
    }

    /* 1 Use bulk get by default. */
    count = dictGetRandomKeys(sampledict, samples, server.maxmemory_samples);

	// 2
    for (j = 0; j < count; j++) {
        unsigned long long idle;
        sds key;
        robj *o;
        dictEntry *de;

        de = samples[j];
        key = dictGetKey(de);
        /* If the dictionary we are sampling from is not the main
         * dictionary (but the expires one) we need to lookup the key
         * again in the key dictionary to obtain the value object. */
        if (sampledict != keydict) de = dictFind(keydict, key);
        // 3
        o = dictGetVal(de);
        // 4
        idle = estimateObjectIdleTime(o);

        /* 5  Insert the element inside the pool.
         * First, find the first empty bucket or the first populated
         * bucket that has an idle time smaller than our idle time. */
        k = 0;
        while (k < REDIS_EVICTION_POOL_SIZE &&
               pool[k].key &&
               pool[k].idle < idle)
            k++;
        
		...
            
        // 6
        pool[k].key = sdsdup(key);
        pool[k].idle = idle;
    }
    if (samples != _samples) zfree(samples);
}

1处，获取 server.maxmemory_samples个key，这里是随机获取的，（dictGetRandomKeys），这个值，默认值为5，放到samples中
2处，遍历返回来的samples

3处，调用如下宏，获取val

he的类型为dictEntry：

Copy
/*
 * 哈希表节点
 */
typedef struct dictEntry {
    
    // 键
    void *key;

    // 值
    union {
        // 1
        void *val;
        uint64_t u64;
        int64_t s64;
    } v;

    // 指向下个哈希表节点，形成链表
    struct dictEntry *next;

} dictEntry;

所以，这里去

Copy
robj *o; 

o = dictGetVal(de);

实际就是获取其v属性中的val,(1处)：

Copy
#define dictGetVal(he) ((he)->v.val)

4处，准备计算该val的空闲时间

我们上面3处，看到，获取的o的类型为robj。我们现在看看怎么计算对象的空闲时长：

Copy
/* Given an object returns the min number of milliseconds the object was never
 * requested, using an approximated LRU algorithm. */
unsigned long long estimateObjectIdleTime(robj *o) {
    //4.1 获取系统的当前时间
    unsigned long long lruclock = LRU_CLOCK();
    // 4.2
    if (lruclock >= o->lru) {
        // 4.3
        return (lruclock - o->lru) * REDIS_LRU_CLOCK_RESOLUTION;
    } else {
        return (lruclock + (REDIS_LRU_CLOCK_MAX - o->lru)) *
                    REDIS_LRU_CLOCK_RESOLUTION;
    }
}

这里，4.1处，获取系统的当前时间；

4.2处，如果系统时间，大于对象的lru时间

4.3处，则用系统时间减去对象的lru时间，再乘以单位,换算为毫秒，最终返回的单位，为毫秒(可以看注释。)

Copy
#define REDIS_LRU_CLOCK_RESOLUTION 1000 /* LRU clock resolution in ms */

5处，这里拿当前元素，和pool中已经放进去的元素，从第0个开始比较，如果当前元素的idle时长，大于pool中指针0指向的元素，则和pool中索引1的元素比较；直到条件不满足为止。

这句话意思就是，类似于冒泡，把当前元素一直往后冒，直到idle时长小于被比较的元素为止。
6处，把当前元素放进pool中。

经过上面的处理后，链表中存放了全部的抽样元素，且ide时间最长的，在最右边。

对象还有字段存储空闲时间？#

前面4处，说到，用系统的当前时间，减去对象的lru时间。

大家看看对象的结构体

Copy
typedef struct redisObject {

    // 类型
    unsigned type:4;

    // 编码
    unsigned encoding:4;

    //1 对象最后一次被访问的时间
    unsigned lru:REDIS_LRU_BITS; /* lru time (relative to server.lruclock) */

    // 引用计数
    int refcount;

    // 指向实际值的指针
    void *ptr;

} robj;

上面1处，lru属性，就是用来存储这个。

创建对象时，直接使用当前系统时间创建#

Copy
robj *createObject(int type, void *ptr) {

    robj *o = zmalloc(sizeof(*o));

    o->type = type;
    o->encoding = REDIS_ENCODING_RAW;
    o->ptr = ptr;
    o->refcount = 1;

    /*1 Set the LRU to the current lruclock (minutes resolution). */
    o->lru = LRU_CLOCK();
    return o;
}

1处即是。

Copy
robj *createEmbeddedStringObject(char *ptr, size_t len) {
    robj *o = zmalloc(sizeof(robj)+sizeof(struct sdshdr)+len+1);
    struct sdshdr *sh = (void*)(o+1);

    o->type = REDIS_STRING;
    o->encoding = REDIS_ENCODING_EMBSTR;
    o->ptr = sh+1;
    o->refcount = 1;
    // 1
    o->lru = LRU_CLOCK();

    sh->len = len;
    sh->free = 0;
    if (ptr) {
        memcpy(sh->buf,ptr,len);
        sh->buf[len] = '\0';
    } else {
        memset(sh->buf,0,len+1);
    }
    return o;
}

1处即是。

每次查找该key时，刷新时间#

Copy
robj *lookupKey(redisDb *db, robj *key) {

    // 查找键空间
    dictEntry *de = dictFind(db->dict,key->ptr);

    // 节点存在
    if (de) {
        

        // 取出值
        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. */
        // 更新时间信息（只在不存在子进程时执行，防止破坏 copy-on-write 机制）
        if (server.rdb_child_pid == -1 && server.aof_child_pid == -1)
            // 1
            val->lru = LRU_CLOCK();

        // 返回值
        return val;
    } else {

        // 节点不存在

        return NULL;
    }
}