Pod中进程内存缓存问题

#进入cgroup工作目录
cd /sys/fs/cgroup/memory/

#查找容器id
docker ps | grep {name}

#通过容器id查找Pod slice目录，如下
systemd-cgls | grep -C3 contianerId
│   └─kubepods-burstable-pod2df5422b_fed3_11e9_be3c_5254008ade6c.slice
│     ├─docker-b9f2ce762cdc8435167bff21eb2cd31d9e1214ad75f2326bffba5e2f1d46422d.scope #业务容器
│     │ ├─32014 tini -- /run.sh
│     │ └─32584 java -jar /deployments/test.jar
│     └─docker-77fde00b840ba03d4d2d52dd0e08ff2a2db7c85e0233bf2847aaa8fa2250b657.scope #根容器
│       └─28520 /usr/bin/pod

#查看内存信息相关目录如下
cat /sys/fs/cgroup/memory/kubepods.slice/kubepods-burstable.slice/kubepods-burstable-pod2df5422b_fed3_11e9_be3c_5254008ade6c.slice/docker-b9f2ce762cdc8435167bff21eb2cd31d9e1214ad75f2326bffba5e2f1d46422d.scope/memory.usage_in_bytes

cat /sys/fs/cgroup/memory/kubepods.slice/kubepods-burstable.slice/kubepods-burstable-pod2df5422b_fed3_11e9_be3c_5254008ade6c.slice/docker-b9f2ce762cdc8435167bff21eb2cd31d9e1214ad75f2326bffba5e2f1d46422d.scope/memory.stat

memory.usage_in_bytes文件中的值即为metrics监控中的值

memory.stat文件中cahce字段的值就是这个容器用于cache的内存

#查看java应用的log4j2配置，发现其中配置了最大保留两个归档文件，每个归档文件大小为2GB，也就是说日志量最大为6GB左右，查看pod中的日志也确实达到了最大的量。做以下步骤验证
1.取消日志输出到文件，metrics监控到得值与dokcer stats或top中的值相近
2.配置最大保留两个归档文件，每个归档文件大小为100M，查看memory.stat中的cache值为350M左右

也就是说log42配置的日志输出到文件会缓存到内存

#是没有任何影响的，java进程只消耗内存1.7G左右，剩余的内存是被log4j用于缓存以充分利用内存提高读写效率，当java进程实例消耗内存增加，cache也会相应的释放，这应该是linux的内存机制决定的
Linux has this basic rule: a page of free RAM is wasted RAM. RAM is used for a lot more than just user application data. It also stores data for the kernel itself and, most importantly, can mirror data stored on the disk for super-fast access, this is reported usually as “buffers/cache”, “disk cache” or “cached” by top. Cached memory is essentially free, in that it can be replaced quickly if a running (or newly starting) program needs the memory.