How to get the page cache for a containerized process?

[grissomsh]

If I ran a pod in a Kubernetes environment, I could get the pid by using ps -ef of the process running inside the container.

When I run
./pgcacher -pid=49573

I got a lot of skipping messages
2023/04/26 17:10:08 skipping "/usr/lib/xxx/lib/gson-2.2.4.jar": could not open file for read: open /usr/lib/inceptor/lib/gson-2.2.4.jar: no such file or directory

And the result is quite fewer.

If I copy the pgcacher into the container by using kubectl cp command, I will get the result without any error and with more records.

./pgcacher -pid=1

Is that the right behavior?

[grissomsh]

结合 nsenter 的解决方法

docker ps | grep # 获取容器 ID

docker inspect -f {{.State.Pid}} # 获取 pid

#进入容器的 process 空间获取进程 id 信息
nsenter --target 3550072 -p -r ps -ef

#进入容器的 process 空间获取 cache 信息
nsenter --target 3550072 -p -r ./pgcacher -pid=119