docker troubleshoot

large disk usage

1. find large log and remove

   $ for name in $(docker ps -a  | awk '{print $1}' | grep -v CONTAINER); do docker inspect $name | grep LogPath | awk '{print $NF}' | tr -d '",' |xargs du -sh;done
   

2. tune log config at /etc/docker/daemon.json

3. remove unused images, like intermediate images, unused and not reference as well

   for images_id_1 in `docker images  | awk '$2 ~ "<none>"{print $3}'`
   do
    docker rmi $images_id_1
   done
   

4. namespace

   lsns -p <pid>
   unshare
   nsenter -t <pid> -u command
   

5. cgroup

   /sys/fs/cgroup/memory/foo/memory.limit_in_bytes
   echo 2428 > /sys/fs/cgroup/memory/foo/cgroup.procs
   ps -o cgroup 2428