buffer 与cache 的理解

buffer 与cache :

通过 $ man free

Buffers Memory used by kernel buffers (Buffers in /proc/meminfo)

cache  Memory used by the page cache and slabs (Cached and SReclaimable in /proc/meminfo)

buff/cache     Sum of buffers and cache

可以看到，buffers是内核缓冲区用到的内存；cache是内核页缓存和Slab用到的内存，cache大小等于Cached +SReclaimable。

$man proc

Buffers %lu
    Relatively temporary storage for raw disk blocks that shouldn't get tremendously large (20MB or so).

Cached %lu
   In-memory cache for files read from the disk (the page cache).  Doesn't include SwapCached.
...
SReclaimable %lu (since Linux 2.6.19)
    Part of Slab, that might be reclaimed, such as caches.
    
SUnreclaim %lu (since Linux 2.6.19)
    Part of Slab, that cannot be reclaimed on memory pressure.
上面这段可以简单理解为：

CPU(写) ------> buffer ------> disk（解决CPU与硬盘性能差异，待写数据放入buffer池）

CPU（读）<------ cache <------ disk（命中缓存，不用每次去disk获取数据，命中率越高，IO读（bi）越小）

A buffer is something that has yet to be "written" to disk.
A cache is something that has been "read" from the disk and stored for later use.
 

$ free -m
             total       used       free     shared    buffers     cached
Mem:          2006       1735        270          1         39        728
-/+ buffers/cache:        967       1038
Swap:            0          0          0 

 

# total = used + free + shared

'-buffers/cache used' = 第一行used - buffers - cached - shared;

'+buffers/cache free' = 第一行free + buffers + cached + shared;

 

#内存不足，会触发缓存回收，swap机制和OOM。Swap 越高，说明内核置换cache越频繁，内存资源越紧张。

 

$  vmstat 1 3 #查看bi，bi:磁盘读，bo：磁盘写
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
 0  0      0 277136  40616 745644    0    0    88   237 1188  526  7  2 91  0  0
 0  0      0 277136  40632 745636    0    0     0    64 1264  628  0  0 100  0  0
 0  0      0 277136  40632 745644    0    0     0     4 1258  607  0  0 100  0  0

#cache越大，bi越小