This moves the bind mounts like /.dockerinit, /etc/hostname, volumes,
etc into the container namespace, by setting them up using lxc.
This is useful to avoid littering the global namespace with a lot of
mounts that are internal to each container and are not generally
needed on the outside. In particular, it seems that having a lot of
mounts is problematic wrt scaling to a lot of containers on systems
where the root filesystem is mounted --rshared.
Note that the "private" option is only supported by the native driver, as
lxc doesn't support setting this. This is not a huge problem, but it does
mean that some mounts are unnecessarily shared inside the container if you're
using the lxc driver.
Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
Upstream-commit: 6c266c4b42eeabe2d433a994753d86637fe52a0b
Component: engine
Use the DOCKER_RAMDISK env var to tell the native driver not to use
a pivot root when setting up the rootfs of a container.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
Upstream-commit: 36dd124b16a76704a88142fa96bb4bb6260dd821
Component: engine
Now that we unmount all the mounts from the global namespace we can
use a private namespace rather than a slave one (as we have no need
for unmounts of inherited global mounts to propagate into the
container).
Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
Upstream-commit: 757b5775725fb90262cee1fa6068fa9dcbbff59f
Component: engine
Instead of keeping all the old mounts in the container namespace and
just using subtree as root we pivot_root so that the actual root in
the namespace is the root we want, and then we unmount the previous
mounts.
This has multiple advantages:
* The namespace mount tree is smaller (in the kernel)
* If you break out of the chroot you could previously access the host
filesystem. Now the host filesystem is fully invisible to the namespace.
* We get rid of all unrelated mounts from the parent namespace, which means
we don't hog these. This is important if we later switch to MS_PRIVATE instead
of MS_SLAVE as otherwise these mounts would be impossible to unmount from the
parent namespace.
Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
Upstream-commit: 5b5c884cc8266d0c2a56da0bc2df14cc9d5d85e8
Component: engine
Michael Crosby
Michael Crosby
Alexander Larsson
Guillaume J. Charmes
Guillaume J. Charmes