您的位置:首页 > 数据库 > Memcache

深入理解Memcached原理

2015-07-24 08:59 603 查看

1.为什么要使用memcache

由于网站的高并发读写需求,传统的关系型数据库开始出现瓶颈,例如:

1)对数据库的高并发读写:

关系型数据库本身就是个庞然大物,处理过程非常耗时(如解析SQL语句,事务处理等)。如果对关系型数据库进行高并发读写(每秒上万次的访问),那么它是无法承受的。

2)对海量数据的处理:

对于大型的SNS网站,每天有上千万次的数据产生(如twitter, 新浪微博)。对于关系型数据库,如果在一个有上亿条数据的数据表种查找某条记录,效率将非常低。


使用memcache能很好的解决以上问题。

在实际使用中,通常把数据库查询的结果保存到Memcache中,下次访问时直接从memcache中读取,而不再进行数据库查询操作,这样就在很大程度上减少了数据库的负担。

保存在memcache中的对象实际放置在内存中,这也是memcache如此高效的原因。


2.memcache的安装和使用

这个网上有太多教程了,不做赘言。

3.基于libevent的事件处理


libevent是个程序库,它将Linux的epoll、BSD类操作系统的kqueue等事件处理功能 封装成统一的接口。即使对服务器的连接数增加,也能发挥O(1)的性能。

memcached使用这个libevent库,因此能在Linux、BSD、Solaris等操作系统上发挥其高性能。

参考:


4.memcache使用实例:

[php] view plaincopyprint?
  1. <?php
  2. $mc=newMemcache();
  3. $mc->connect('127.0.0.1',11211);
  4. $uid=(int)$_GET['uid'];
  5. $sql="select*fromuserswhereuid='uid'";
  6. $key=md5($sql);
  7. if(!($data=$mc->get($key))){
  8. $conn=mysql_connect('localhost','test','test');
  9. mysql_select_db('test');
  10. $result=mysql_fetch_object($result);
  11. while($row=mysql_fetch_object($result)){
  12. $data[]=$row;
  13. }
  14. $mc->add($key,$datas);
  15. }
  16. var_dump($datas);
  17. ?>
<?php
$mc = new Memcache();
$mc->connect('127.0.0.1', 11211);

$uid = (int)$_GET['uid'];
$sql = "select * from users where uid='uid' ";
$key = md5($sql);
if(!($data = $mc->get($key))) {
$conn = mysql_connect('localhost', 'test', 'test');
mysql_select_db('test');
$result = mysql_fetch_object($result);
while($row = mysql_fetch_object($result)) {
$data[] = $row;
}
$mc->add($key, $datas);
}

var_dump($datas);
?>


5.memcache如何支持高并发(此处还需深入研究)

memcache使用多路复用I/O模型,如(epoll, select等),传统I/O中,系统可能会因为某个用户连接还没做好I/O准备而一直等待,知道这个连接做好I/O准备。这时如果有其他用户连接到服务器,很可能会因为系统阻塞而得不到响应。

而多路复用I/O是一种消息通知模式,用户连接做好I/O准备后,系统会通知我们这个连接可以进行I/O操作,这样就不会阻塞在某个用户连接。因此,memcache才能支持高并发。

此外,memcache使用了多线程机制。可以同时处理多个请求。线程数一般设置为CPU核数,这研报告效率最高。


6.使用Slab分配算法保存数据

slab分配算法的原理是:把固定大小(1MB)的内存分为n小块,如下图所示:



slab分配算法把每1MB大小的内存称为一个slab页,每次向系统申请一个slab页,然后再通过分隔算法把这个slab页分割成若干个小块的chunk(如上图所示),然后把这些chunk分配给用户使用,分割算法如下(在slabs.c文件中):

(注:memcache的github项目地址:https://github.com/wusuopubupt/memcached)

[cpp] view plaincopyprint?
  1. /**
  2. *Determinesthechunksizesandinitializestheslabclassdescriptors
  3. *accordingly.
  4. */
  5. voidslabs_init(constsize_tlimit,constdoublefactor,constboolprealloc){
  6. inti=POWER_SMALLEST-1;
  7. unsignedintsize=sizeof(item)+settings.chunk_size;
  8. mem_limit=limit;
  9. if(prealloc){
  10. /*Allocateeverythinginabigchunkwithmalloc通过malloc的方式申请内存*/
  11. mem_base=malloc(mem_limit);
  12. if(mem_base!=NULL){
  13. mem_current=mem_base;
  14. mem_avail=mem_limit;
  15. }else{
  16. fprintf(stderr,"Warning:Failedtoallocaterequestedmemoryin"
  17. "onelargechunk.\nWillallocateinsmallerchunks\n");
  18. }
  19. }
  20. memset(slabclass,0,sizeof(slabclass));
  21. while(++i<POWER_LARGEST&&size<=settings.item_size_max/factor){
  22. /*Makesureitemsarealwaysn-bytealigned注意这里的字节对齐*/
  23. if(size%CHUNK_ALIGN_BYTES)
  24. size+=CHUNK_ALIGN_BYTES-(size%CHUNK_ALIGN_BYTES);
  25. slabclass[i].size=size;
  26. slabclass[i].perslab=settings.item_size_max/slabclass[i].size;
  27. size*=factor;//以1.25为倍数增大chunk
  28. if(settings.verbose>1){
  29. fprintf(stderr,"slabclass%3d:chunksize%9uperslab%7u\n",
  30. i,slabclass[i].size,slabclass[i].perslab);
  31. }
  32. }
  33. power_largest=i;
  34. slabclass[power_largest].size=settings.item_size_max;
  35. slabclass[power_largest].perslab=1;
  36. if(settings.verbose>1){
  37. fprintf(stderr,"slabclass%3d:chunksize%9uperslab%7u\n",
  38. i,slabclass[i].size,slabclass[i].perslab);
  39. }
  40. /*forthetestsuite:fakingofhowmuchwe'vealreadymalloc'd*/
  41. {
  42. char*t_initial_malloc=getenv("T_MEMD_INITIAL_MALLOC");
  43. if(t_initial_malloc){
  44. mem_malloced=(size_t)atol(t_initial_malloc);
  45. }
  46. }
  47. if(prealloc){
  48. slabs_preallocate(power_largest);
  49. }
  50. }
/**
* Determines the chunk sizes and initializes the slab class descriptors
* accordingly.
*/
void slabs_init(const size_t limit, const double factor, const bool prealloc) {
int i = POWER_SMALLEST - 1;
unsigned int size = sizeof(item) + settings.chunk_size;

mem_limit = limit;

if (prealloc) {
/* Allocate everything in a big chunk with malloc 通过malloc的方式申请内存*/
mem_base = malloc(mem_limit);
if (mem_base != NULL) {
mem_current = mem_base;
mem_avail = mem_limit;
} else {
fprintf(stderr, "Warning: Failed to allocate requested memory in"
" one large chunk.\nWill allocate in smaller chunks\n");
}
}

memset(slabclass, 0, sizeof(slabclass));

while (++i < POWER_LARGEST && size <= settings.item_size_max / factor) {
/* Make sure items are always n-byte aligned  注意这里的字节对齐*/
if (size % CHUNK_ALIGN_BYTES)
size += CHUNK_ALIGN_BYTES - (size % CHUNK_ALIGN_BYTES);

slabclass[i].size = size;
slabclass[i].perslab = settings.item_size_max / slabclass[i].size;
size *= factor;//以1.25为倍数增大chunk
if (settings.verbose > 1) {
fprintf(stderr, "slab class %3d: chunk size %9u perslab %7u\n",
i, slabclass[i].size, slabclass[i].perslab);
}
}

power_largest = i;
slabclass[power_largest].size = settings.item_size_max;
slabclass[power_largest].perslab = 1;
if (settings.verbose > 1) {
fprintf(stderr, "slab class %3d: chunk size %9u perslab %7u\n",
i, slabclass[i].size, slabclass[i].perslab);
}

/* for the test suite:  faking of how much we've already malloc'd */
{
char *t_initial_malloc = getenv("T_MEMD_INITIAL_MALLOC");
if (t_initial_malloc) {
mem_malloced = (size_t)atol(t_initial_malloc);
}

}

if (prealloc) {
slabs_preallocate(power_largest);
}
}



上面代码中的slabclass是一个类型为slabclass_t结构的数组,其定义如下:

[cpp] view plaincopyprint?
  1. typedefstruct{
  2. unsignedintsize;/*sizesofitems*/
  3. unsignedintperslab;/*howmanyitemsperslab*/
  4. void**slots;/*listofitemptrs*/
  5. unsignedintsl_total;/*sizeofpreviousarray*/
  6. unsignedintsl_curr;/*firstfreeslot*/
  7. void*end_page_ptr;/*pointertonextfreeitematendofpage,or0*/
  8. unsignedintend_page_free;/*numberofitemsremainingatendoflastallocedpage*/
  9. unsignedintslabs;/*howmanyslabswereallocatedforthisclass*/
  10. void**slab_list;/*arrayofslabpointers*/
  11. unsignedintlist_size;/*sizeofprevarray*/
  12. unsignedintkilling;/*index+1ofdyingslab,orzeroifnone*/
  13. size_trequested;/*Thenumberofrequestedbytes*/
  14. }slabclass_t;
typedef struct {
unsigned int size;      /* sizes of items */
unsigned int perslab;   /* how many items per slab */
void **slots;           /* list of item ptrs */
unsigned int sl_total;  /* size of previous array */
unsigned int sl_curr;   /* first free slot */
void *end_page_ptr;         /* pointer to next free item at end of page, or 0 */
unsigned int end_page_free; /* number of items remaining at end of last alloced page */
unsigned int slabs;     /* how many slabs were allocated for this class */
void **slab_list;       /* array of slab pointers */
unsigned int list_size; /* size of prev array */
unsigned int killing;  /* index+1 of dying slab, or zero if none */
size_t requested; /* The number of requested bytes */
} slabclass_t;

借用别人的一张图说明slabclass_t结构:



由分割算法的源代码可知,slab算法按照不同大小的chunk分割slab页,而不同大小的chunk以factor(默认是1.25)倍增大。

使用memcache -u root -vv 命令查看内存分配情况(8字节对齐):




找到大小最合适的chunk分配给请求缓存的数据:

[cpp] view plaincopyprint?
  1. /*
  2. *Figuresoutwhichslabclass(chunksize)isrequiredtostoreanitemof
  3. *agivensize.
  4. *
  5. *Givenobjectsize,returnidtousewhenallocating/freeingmemoryforobject
  6. *0meanserror:can'tstoresuchalargeobject
  7. */
  8. unsignedintslabs_clsid(constsize_tsize){
  9. intres=POWER_SMALLEST;//初始化为最小的chunk
  10. if(size==0)
  11. return0;
  12. while(size>slabclass[res].size)//逐渐增大chunksize,直到找到第一个比申请的size大的chunk
  13. if(res++==power_largest)/*won'tfitinthebiggestslab*/
  14. return0;
  15. returnres;
  16. }
/*
* Figures out which slab class (chunk size) is required to store an item of
* a given size.
*
* Given object size, return id to use when allocating/freeing memory for object
* 0 means error: can't store such a large object
*/

unsigned int slabs_clsid(const size_t size) {
int res = POWER_SMALLEST;// 初始化为最小的chunk

if (size == 0)
return 0;
while (size > slabclass[res].size) //逐渐增大chunk size,直到找到第一个比申请的size大的chunk
if (res++ == power_largest)     /* won't fit in the biggest slab */
return 0;
return res;
}


内存分配:

(此处参考:http://slowsnail.com.cn/?p=20

[cpp] view plaincopyprint?
  1. staticvoid*do_slabs_alloc(constsize_tsize,unsignedintid){
  2. slabclass_t*p;
  3. void*ret=NULL;
  4. item*it=NULL;
  5. if(id<POWER_SMALLEST||id>power_largest){//判断id是否会导致slabclass[]数组越界
  6. MEMCACHED_SLABS_ALLOCATE_FAILED(size,0);
  7. returnNULL;
  8. }
  9. p=&slabclass[id];//获取slabclass[id]的引用
  10. assert(p->sl_curr==0||((item*)p->slots)->slabs_clsid==0);//判断slabclass[id]是否有剩余的chunk
  11. if(!(p->sl_curr!=0||do_slabs_newslab(id)!=0)){//如果slabclass[id]中已经没有空余chunk并且试图向系统申请一个“页”(slab)的chunk失败,则返回NULL
  12. /*Wedon'thavemorememoryavailable*/
  13. ret=NULL;
  14. }elseif(p->sl_curr!=0){//slabclass[id]的空闲链表中还有chunk,则直接将其分配出去
  15. it=(item*)p->slots;//获取空闲链表的头指针
  16. p->slots=it->next;//将头结点指向下一个结点(取下头结点)
  17. if(it->next)it->next->prev=0;//将新头结点的prev指针置空
  18. p->sl_curr--;//减少slabclass[id]空闲链表中的chunk计数
  19. ret=(void*)it;//将头结点赋给ret指针
  20. }
  21. if(ret){//请求成功
  22. p->requested+=size;//更新slabclass[id]所分配的内存总数
  23. MEMCACHED_SLABS_ALLOCATE(size,id,p->size,ret);
  24. }else{
  25. MEMCACHED_SLABS_ALLOCATE_FAILED(size,id);
  26. }
  27. returnret;
  28. }
static void *do_slabs_alloc(const size_t size, unsigned int id) {
slabclass_t *p;
void *ret = NULL;
item *it = NULL;

if (id < POWER_SMALLEST || id > power_largest) {//判断id是否会导致slabclass[]数组越界
MEMCACHED_SLABS_ALLOCATE_FAILED(size, 0);
return NULL;
}

p = &slabclass[id];//获取slabclass[id]的引用
assert(p->sl_curr == 0 || ((item *)p->slots)->slabs_clsid == 0);//判断slabclass[id]是否有剩余的chunk

if (! (p->sl_curr != 0 || do_slabs_newslab(id) != 0)) {//如果slabclass[id]中已经没有空余chunk并且试图向系统申请一个“页”(slab)的chunk失败,则返回NULL
/* We don't have more memory available */
ret = NULL;
} else if (p->sl_curr != 0) {//slabclass[id]的空闲链表中还有chunk,则直接将其分配出去
it = (item *)p->slots;//获取空闲链表的头指针
p->slots = it->next;//将头结点指向下一个结点(取下头结点)
if (it->next) it->next->prev = 0;//将新头结点的prev指针置空
p->sl_curr--;//减少slabclass[id]空闲链表中的chunk计数
ret = (void *)it;//将头结点赋给ret指针
}

if (ret) {//请求成功
p->requested += size;//更新slabclass[id]所分配的内存总数
MEMCACHED_SLABS_ALLOCATE(size, id, p->size, ret);
} else {
MEMCACHED_SLABS_ALLOCATE_FAILED(size, id);
}

return ret;
}

do_slabs_allc()函数首先尝试从slot列表(被回收的chunk)中获取可用的chunk,如果有可用的就返回,否则从空闲的chunk列表中获取可用的chunk并返回。

删除过期item:

延迟删除过期item到查找时进行,可以提高memcache的效率,因为不必每时每刻检查过期item,从而提高CPU工作效率


使用LRU(last recently used)算法淘汰数据:

[cpp] view plaincopyprint?
  1. /*
  2. *trytogetoneofftherightLRU
  3. *don'tnecessariulyunlinkthetailbecauseitmaybelocked:refcount>0
  4. *searchupfromtailanitemwithrefcount==0andunlinkit;giveupafter50
  5. *tries
  6. */
  7. if(tails[id]==0){
  8. itemstats[id].outofmemory++;
  9. returnNULL;
  10. }
  11. for(search=tails[id];tries>0&&search!=NULL;tries--,search=search->prev){
  12. if(search->refcount==0){//refount==0的情况,释放掉
  13. if(search->exptime==0||search->exptime>current_time){
  14. itemstats[id].evicted++;
  15. itemstats[id].evicted_time=current_time-search->time;
  16. STATS_LOCK();
  17. stats.evictions++;
  18. STATS_UNLOCK();
  19. }
  20. do_item_unlink(search);
  21. break;
  22. }
  23. }
  24. it=slabs_alloc(ntotal,id);
  25. if(it==0){
  26. itemstats[id].outofmemory++;
  27. /*Lastditcheffort.Thereisaveryrarebugwhichcauses
  28. *refcountleaks.We'vefixedmostofthem,butitstillhappens,
  29. *anditmayhappeninthefuture.
  30. *Wecanreasonablyassumenoitemcanstaylockedformorethan
  31. *threehours,soifwefindoneinthetailwhichisthatold,
  32. *freeitanyway.
  33. */
  34. tries=50;
  35. for(search=tails[id];tries>0&&search!=NULL;tries--,search=search->prev){
  36. if(search->refcount!=0&&search->time+10800<current_time){//最近3小时没有被访问到的情况,释放掉
  37. itemstats[id].tailrepairs++;
  38. search->refcount=0;
  39. do_item_unlink(search);
  40. break;
  41. }
  42. }
  43. it=slabs_alloc(ntotal,id);
  44. if(it==0){
  45. returnNULL;
  46. }
  47. }
/*
* try to get one off the right LRU
* don't necessariuly unlink the tail because it may be locked: refcount>0
* search up from tail an item with refcount==0 and unlink it; give up after 50
* tries
*/

if (tails[id] == 0) {
itemstats[id].outofmemory++;
return NULL;
}

for (search = tails[id]; tries > 0 && search != NULL; tries--, search=search->prev) {
if (search->refcount == 0) { //refount==0的情况,释放掉
if (search->exptime == 0 || search->exptime > current_time) {
itemstats[id].evicted++;
itemstats[id].evicted_time = current_time - search->time;
STATS_LOCK();
stats.evictions++;
STATS_UNLOCK();
}
do_item_unlink(search);
break;
}
}
it = slabs_alloc(ntotal, id);
if (it == 0) {
itemstats[id].outofmemory++;
/* Last ditch effort. There is a very rare bug which causes
* refcount leaks. We've fixed most of them, but it still happens,
* and it may happen in the future.
* We can reasonably assume no item can stay locked for more than
* three hours, so if we find one in the tail which is that old,
* free it anyway.
*/
tries = 50;
for (search = tails[id]; tries > 0 && search != NULL; tries--, search=search->prev) {
if (search->refcount != 0 && search->time + 10800 < current_time) { //最近3小时没有被访问到的情况,释放掉
itemstats[id].tailrepairs++;
search->refcount = 0;
do_item_unlink(search);
break;
}
}
it = slabs_alloc(ntotal, id);
if (it == 0) {
return NULL;
}
}

从item列表的尾部开始遍历,找到refcount==0的chunk,调用do_item_unlink()函数释放掉,另外,search->time+10800<current_time(即最近3小时没有被访问过的item),也释放掉--这就是LRU算法的原理。


附:阿里2014笔试题一道:

某缓存系统采用LRU淘汰算法,假定缓存容量为4,并且初始为空,那么在顺序访问一下数据项的时候:1,5,1,3,5,2,4,1,2出现缓存直接命中的次数是?,最后缓存中即将准备淘汰的数据项是?
答案:3, 5 解答:
  1. 1调入内存 1
  2. 5调入内存 1 5
  3. 1调入内存 5 1(命中 1,更新次序)
  4. 3调入内存 5 1 3
  5. 5调入内存 1 3 5 (命中5)
  6. 2调入内存 1 3 5 2
  7. 4调入内存(1最久未使用,淘汰1) 3 5 2 4
  8. 1调入内存(3最久未使用,淘汰3) 5 2 4 1
  9. 2调入内存 5 4 1 2(命中2)
因此,直接命中次数是3,最后缓存即将准备淘汰的数据项是5 以上是转自https://www.geek-share.com/detail/2600627961.html

阅读更多
内容来自用户分享和网络整理,不保证内容的准确性,如有侵权内容,可联系管理员处理 点击这里给我发消息
标签: