docker exec 无法进入容器

OCI runtime exec failed: exec failed: container_linux.go:348: starting container process caused "process_linux.go:78: starting setns process caused \"fork/exec /proc/self/exe: no such file or directory\"": unknown

使用nsenter方式进入容器

# 获取进程pid
# docker inspect -f {{.State.Pid}} 44fc0f0582d9  
# nsenter命令访问该容器 pid 3326
# nsenter --target 3326 --mount --uts --ipc --net --pid

# 命令学习
# nsenter --help

用法：
 nsenter [options] <program> [<argument>...]

Run a program with namespaces of other processes.

选项：
 -t, --target <pid>     要获取名字空间的目标进程
 -m, --mount[=<file>]   enter mount namespace
 -u, --uts[=<file>]     enter UTS namespace (hostname etc)
 -i, --ipc[=<file>]     enter System V IPC namespace
 -n, --net[=<file>]     enter network namespace
 -p, --pid[=<file>]     enter pid namespace
 -U, --user[=<file>]    enter user namespace
 -S, --setuid <uid>     set uid in entered namespace
 -G, --setgid <gid>     set gid in entered namespace
     --preserve-credentials do not touch uids or gids
 -r, --root[=<dir>]     set the root directory
 -w, --wd[=<dir>]       set the working directory
 -F, --no-fork          执行 <程序> 前不 fork
 -Z, --follow-context   set SELinux context according to --target PID

 -h, --help     显示此帮助并退出
 -V, --version  输出版本信息并退出



# man nsenter

NSENTER(1)                                                                   User Commands                                                                  NSENTER(1)



NAME
       nsenter - run program with namespaces of other processes

SYNOPSIS
       nsenter [options] [program [arguments]]

DESCRIPTION
       Enters the namespaces of one or more other processes and then executes the specified program.  Enterable namespaces are:

       mount namespace
              Mounting  and  unmounting  filesystems  will not affect the rest of the system (CLONE_NEWNS flag), except for filesystems which are explicitly marked as
              shared (with mount --make-shared; see /proc/self/mountinfo for the shared flag).

       UTS namespace
              Setting hostname or domainname will not affect the rest of the system.  (CLONE_NEWUTS flag)

       IPC namespace
              The process will have an independent namespace for System V message queues, semaphore sets and shared memory segments.  (CLONE_NEWIPC flag)

       network namespace
              The process will have independent IPv4 and IPv6 stacks, IP routing tables, firewall rules, the /proc/net and /sys/class/net  directory  trees,  sockets,
              etc.  (CLONE_NEWNET flag)

       PID namespace
              Children will have a set of PID to process mappings separate from the nsenter process (CLONE_NEWPID flag).  nsenter will fork by default if changing the
              PID namespace, so that the new program and its children share the same PID namespace and are visible to each other.  If --no-fork is used, the new  pro‐
              gram will be exec'ed without forking.

       user namespace
              The process will have a distinct set of UIDs, GIDs and capabilities.  (CLONE_NEWUSER flag)

       See clone(2) for the exact semantics of the flags.

       If program is not given, then ``${SHELL}'' is run (default: /bin/sh).


OPTIONS
       -t, --target pid
              Specify a target process to get contexts from.  The paths to the contexts specified by pid are:

              /proc/pid/ns/mnt    the mount namespace
              /proc/pid/ns/uts    the UTS namespace
              /proc/pid/ns/ipc    the IPC namespace
              /proc/pid/ns/net    the network namespace
              /proc/pid/ns/pid    the PID namespace
              /proc/pid/ns/user   the user namespace
              /proc/pid/root      the root directory
              /proc/pid/cwd       the working directory respectively

       -m, --mount[=file]
              Enter  the  mount namespace.  If no file is specified, enter the mount namespace of the target process.  If file is specified, enter the mount namespace
              specified by file.

       -u, --uts[=file]
              Enter the UTS namespace.  If no file is specified, enter the UTS namespace of the target process.  If file is specified, enter the UTS namespace  speci‐
              fied by file.

       -i, --ipc[=file]
              Enter  the IPC namespace.  If no file is specified, enter the IPC namespace of the target process.  If file is specified, enter the IPC namespace speci‐
              fied by file.

       -n, --net[=file]
              Enter the network namespace.  If no file is specified, enter the network namespace of the target process.  If  file  is  specified,  enter  the  network
              namespace specified by file.

       -p, --pid[=file]
              Enter  the PID namespace.  If no file is specified, enter the PID namespace of the target process.  If file is specified, enter the PID namespace speci‐
              fied by file.

       -U, --user[=file]
              Enter the user namespace.  If no file is specified, enter the user namespace of the target process.  If file is  specified,  enter  the  user  namespace
              specified by file.  See also the --setuid and --setgid options.

       -G, --setgid gid
              Set the group ID which will be used in the entered namespace and drop supplementary groups.  nsenter(1) always sets GID for user namespaces, the default
              is 0.

       -S, --setuid uid
              Set the user ID which will be used in the entered namespace.  nsenter(1) always sets UID for user namespaces, the default is 0.

       --preserve-credentials
              Don't modify UID and GID when enter user namespace. The default is to drops supplementary groups and sets GID and UID to 0.

       -r, --root[=directory]
              Set the root directory.  If no directory is specified, set the root directory to the root directory of the target process.  If directory  is  specified,
              set the root directory to the specified directory.

       -w, --wd[=directory]
              Set  the  working  directory.   If no directory is specified, set the working directory to the working directory of the target process.  If directory is
              specified, set the working directory to the specified directory.

       -F, --no-fork
              Do not fork before exec'ing the specified program.  By default, when entering a PID namespace, nsenter calls fork before calling exec so that any  chil‐
              dren will also be in the newly entered PID namespace.

       -Z, --follow-context
              Set the SELinux security context used for executing a new process according to already running process specified by --target PID. (The util-linux has to
              be compiled with SELinux support otherwise the option is unavailable.)

       -V, --version
              Display version information and exit.

       -h, --help
              Display help text and exit.

SEE ALSO
       setns(2), clone(2)

AUTHORS
       Eric Biederman ⟨biederm@xmission.com⟩
       Karel Zak ⟨kzak@redhat.com⟩

AVAILABILITY
       The nsenter command is part of the util-linux package and is available from Linux Kernel Archive ⟨ftp://ftp.kernel.org/pub/linux/utils/util-linux/⟩.



util-linux                                                                     June 2013                                                                    NSENTER(1)