
Currently kubelet volume management works on the concept of desired and actual world of states. The volume manager periodically compares the two worlds and perform volume mount/unmount and/or attach/detach operations. When kubelet restarts, the cache of those two worlds are gone. Although desired world can be recovered through apiserver, actual world can not be recovered which may cause some volumes cannot be cleaned up if their information is deleted by apiserver. This change adds the reconstruction of the actual world by reading the pod directories from disk. The reconstructed volume information is added to both desired world and actual world if it cannot be found in either world. The rest logic would be as same as before, desired world populator may clean up the volume entry if it is no longer in apiserver, and then volume manager should invoke unmount to clean it up.
232 lines
8.0 KiB
Go
232 lines
8.0 KiB
Go
// +build linux
|
|
|
|
/*
|
|
Copyright 2014 The Kubernetes Authors.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
*/
|
|
|
|
package mount
|
|
|
|
import (
|
|
"os"
|
|
"path/filepath"
|
|
"strings"
|
|
|
|
"github.com/golang/glog"
|
|
"k8s.io/kubernetes/pkg/util/exec"
|
|
)
|
|
|
|
// NsenterMounter is part of experimental support for running the kubelet
|
|
// in a container. Currently, all docker containers receive their own mount
|
|
// namespaces. NsenterMounter works by executing nsenter to run commands in
|
|
// the host's mount namespace.
|
|
//
|
|
// NsenterMounter requires:
|
|
//
|
|
// 1. Docker >= 1.6 due to the dependency on the slave propagation mode
|
|
// of the bind-mount of the kubelet root directory in the container.
|
|
// Docker 1.5 used a private propagation mode for bind-mounts, so mounts
|
|
// performed in the host's mount namespace do not propagate out to the
|
|
// bind-mount in this docker version.
|
|
// 2. The host's root filesystem must be available at /rootfs
|
|
// 3. The nsenter binary must be on the Kubelet process' PATH in the container's
|
|
// filesystem.
|
|
// 4. The Kubelet process must have CAP_SYS_ADMIN (required by nsenter); at
|
|
// the present, this effectively means that the kubelet is running in a
|
|
// privileged container.
|
|
// 5. The volume path used by the Kubelet must be the same inside and outside
|
|
// the container and be writable by the container (to initialize volume)
|
|
// contents. TODO: remove this requirement.
|
|
// 6. The host image must have mount, findmnt, and umount binaries in /bin,
|
|
// /usr/sbin, or /usr/bin
|
|
//
|
|
// For more information about mount propagation modes, see:
|
|
// https://www.kernel.org/doc/Documentation/filesystems/sharedsubtree.txt
|
|
type NsenterMounter struct {
|
|
// a map of commands to their paths on the host filesystem
|
|
paths map[string]string
|
|
}
|
|
|
|
func NewNsenterMounter() *NsenterMounter {
|
|
m := &NsenterMounter{
|
|
paths: map[string]string{
|
|
"mount": "",
|
|
"findmnt": "",
|
|
"umount": "",
|
|
},
|
|
}
|
|
// search for the mount command in other locations besides /usr/bin
|
|
for binary := range m.paths {
|
|
// default to root
|
|
m.paths[binary] = filepath.Join("/", binary)
|
|
for _, path := range []string{"/bin", "/usr/sbin", "/usr/bin"} {
|
|
binPath := filepath.Join(path, binary)
|
|
if _, err := os.Stat(filepath.Join(hostRootFsPath, binPath)); err != nil {
|
|
continue
|
|
}
|
|
m.paths[binary] = binPath
|
|
break
|
|
}
|
|
// TODO: error, so that the kubelet can stop if the mounts don't exist
|
|
}
|
|
return m
|
|
}
|
|
|
|
// NsenterMounter implements mount.Interface
|
|
var _ = Interface(&NsenterMounter{})
|
|
|
|
const (
|
|
hostRootFsPath = "/rootfs"
|
|
hostProcMountsPath = "/rootfs/proc/1/mounts"
|
|
nsenterPath = "nsenter"
|
|
)
|
|
|
|
// Mount runs mount(8) in the host's root mount namespace. Aside from this
|
|
// aspect, Mount has the same semantics as the mounter returned by mount.New()
|
|
func (n *NsenterMounter) Mount(source string, target string, fstype string, options []string) error {
|
|
bind, bindRemountOpts := isBind(options)
|
|
|
|
if bind {
|
|
err := n.doNsenterMount(source, target, fstype, []string{"bind"})
|
|
if err != nil {
|
|
return err
|
|
}
|
|
return n.doNsenterMount(source, target, fstype, bindRemountOpts)
|
|
}
|
|
|
|
return n.doNsenterMount(source, target, fstype, options)
|
|
}
|
|
|
|
// doNsenterMount nsenters the host's mount namespace and performs the
|
|
// requested mount.
|
|
func (n *NsenterMounter) doNsenterMount(source, target, fstype string, options []string) error {
|
|
glog.V(5).Infof("nsenter Mounting %s %s %s %v", source, target, fstype, options)
|
|
args := n.makeNsenterArgs(source, target, fstype, options)
|
|
|
|
glog.V(5).Infof("Mount command: %v %v", nsenterPath, args)
|
|
exec := exec.New()
|
|
outputBytes, err := exec.Command(nsenterPath, args...).CombinedOutput()
|
|
if len(outputBytes) != 0 {
|
|
glog.V(5).Infof("Output of mounting %s to %s: %v", source, target, string(outputBytes))
|
|
}
|
|
|
|
return err
|
|
}
|
|
|
|
// makeNsenterArgs makes a list of argument to nsenter in order to do the
|
|
// requested mount.
|
|
func (n *NsenterMounter) makeNsenterArgs(source, target, fstype string, options []string) []string {
|
|
nsenterArgs := []string{
|
|
"--mount=/rootfs/proc/1/ns/mnt",
|
|
"--",
|
|
n.absHostPath("mount"),
|
|
}
|
|
|
|
args := makeMountArgs(source, target, fstype, options)
|
|
|
|
return append(nsenterArgs, args...)
|
|
}
|
|
|
|
// Unmount runs umount(8) in the host's mount namespace.
|
|
func (n *NsenterMounter) Unmount(target string) error {
|
|
args := []string{
|
|
"--mount=/rootfs/proc/1/ns/mnt",
|
|
"--",
|
|
n.absHostPath("umount"),
|
|
target,
|
|
}
|
|
|
|
glog.V(5).Infof("Unmount command: %v %v", nsenterPath, args)
|
|
exec := exec.New()
|
|
outputBytes, err := exec.Command(nsenterPath, args...).CombinedOutput()
|
|
if len(outputBytes) != 0 {
|
|
glog.V(5).Infof("Output of unmounting %s: %v", target, string(outputBytes))
|
|
}
|
|
|
|
return err
|
|
}
|
|
|
|
// List returns a list of all mounted filesystems in the host's mount namespace.
|
|
func (*NsenterMounter) List() ([]MountPoint, error) {
|
|
return listProcMounts(hostProcMountsPath)
|
|
}
|
|
|
|
// IsLikelyNotMountPoint determines whether a path is a mountpoint by calling findmnt
|
|
// in the host's root mount namespace.
|
|
func (n *NsenterMounter) IsLikelyNotMountPoint(file string) (bool, error) {
|
|
file, err := filepath.Abs(file)
|
|
if err != nil {
|
|
return true, err
|
|
}
|
|
|
|
// Check the directory exists
|
|
if _, err = os.Stat(file); os.IsNotExist(err) {
|
|
glog.V(5).Infof("findmnt: directory %s does not exist", file)
|
|
return true, err
|
|
}
|
|
// Add --first-only option: since we are testing for the absense of a mountpoint, it is sufficient to get only
|
|
// the first of multiple possible mountpoints using --first-only.
|
|
// Also add fstype output to make sure that the output of target file will give the full path
|
|
// TODO: Need more refactoring for this function. Track the solution with issue #26996
|
|
args := []string{"--mount=/rootfs/proc/1/ns/mnt", "--", n.absHostPath("findmnt"), "-o", "target,fstype", "--noheadings", "--first-only", "--target", file}
|
|
glog.V(5).Infof("findmnt command: %v %v", nsenterPath, args)
|
|
|
|
exec := exec.New()
|
|
out, err := exec.Command(nsenterPath, args...).CombinedOutput()
|
|
if err != nil {
|
|
glog.V(2).Infof("Failed findmnt command for path %s: %v", file, err)
|
|
// Different operating systems behave differently for paths which are not mount points.
|
|
// On older versions (e.g. 2.20.1) we'd get error, on newer ones (e.g. 2.26.2) we'd get "/".
|
|
// It's safer to assume that it's not a mount point.
|
|
return true, nil
|
|
}
|
|
mountTarget := strings.Split(string(out), " ")[0]
|
|
mountTarget = strings.TrimSuffix(mountTarget, "\n")
|
|
glog.V(5).Infof("IsLikelyNotMountPoint findmnt output for path %s: %v:", file, mountTarget)
|
|
|
|
if mountTarget == file {
|
|
glog.V(5).Infof("IsLikelyNotMountPoint: %s is a mount point", file)
|
|
return false, nil
|
|
}
|
|
glog.V(5).Infof("IsLikelyNotMountPoint: %s is not a mount point", file)
|
|
return true, nil
|
|
}
|
|
|
|
// DeviceOpened checks if block device in use by calling Open with O_EXCL flag.
|
|
// Returns true if open returns errno EBUSY, and false if errno is nil.
|
|
// Returns an error if errno is any error other than EBUSY.
|
|
// Returns with error if pathname is not a device.
|
|
func (n *NsenterMounter) DeviceOpened(pathname string) (bool, error) {
|
|
return exclusiveOpenFailsOnDevice(pathname)
|
|
}
|
|
|
|
// PathIsDevice uses FileInfo returned from os.Stat to check if path refers
|
|
// to a device.
|
|
func (n *NsenterMounter) PathIsDevice(pathname string) (bool, error) {
|
|
return pathIsDevice(pathname)
|
|
}
|
|
|
|
//GetDeviceNameFromMount given a mount point, find the volume id from checking /proc/mounts
|
|
func (n *NsenterMounter) GetDeviceNameFromMount(mountPath, pluginDir string) (string, error) {
|
|
return getDeviceNameFromMount(n, mountPath, pluginDir)
|
|
}
|
|
|
|
func (n *NsenterMounter) absHostPath(command string) string {
|
|
path, ok := n.paths[command]
|
|
if !ok {
|
|
return command
|
|
}
|
|
return path
|
|
}
|