When manipulating container and VM images we need efficient and atomic
directory snapshots and file copies, as well as disk quota. btrfs
provides this, legacy file systems do not. Hence, implicitly create a
loopback file system in /var/lib/machines.raw and mount it to
/var/lib/machines, if that directory is not on btrfs anyway.
This is done implicitly and transparently the first time the user
invokes "machinectl import-xyz".
This allows us to take benefit of btrfs features for container
management without actually having the rest of the system use btrfs.
The loopback is sized 500M initially. Patches to grow it dynamically are
to follow.
The call iterates through cmsg list and closes all fds passed via
SCM_RIGHTS.
This patch also ensures the call is used wherever appropriate, where we
might get spurious fds sent and we should better close them, then leave
them lying around.
Also, allow clients to alter their own objects without any further
priviliges. i.e. this allows clients to kill and lock their own sessions
without involving PK.
We will be woken up on rtnl or dbus activity, so let's just quit if some time has passed and that is the only thing that can happen.
Note that we will always stay around if we expect network activity (e.g. DHCP is enabled), as we are not restarted on that.
After all it is now much more like strjoin() than strappend(). At the
same time, add support for NULL sentinels, even if they are normally not
necessary.
With this change the pull protocol implementation processes will pass
progress data to importd which then passes this information on via the
bus. We use sd_notify() as generic transport for this communication,
making importd listen to them, while matching the incoming messages to
the right transfer.
The old "systemd-import" binary is now an internal tool. We still use it
as asynchronous backend for systemd-importd. Since the import tool might
require some IO and CPU resources (due to qcow2 explosion, and
decompression), and because we might want to run it with more minimal
priviliges we still keep it around as the worker binary to execute as
child process of importd.
machinectl now has verbs for pulling down images, cancelling them and
listing them.
With this change the import tool will now unpack qcow2 images into
normal raw disk images, suitable for usage with nspawn.
This allows has the benefit of also allowing importing Ubuntu Cloud
images for usage with nspawn.
Given that this is also the place to store raw disk images which are
very much bootable with qemu/kvm it sounds like a misnomer to call the
directory "container". Hence, let's change this sooner rather than
later, and use the generic name, in particular since we otherwise try to
use the generic "machine" preferably over the more specific "container"
or "vm".
After all, nspawn can now dissect MBR partition levels, too, hence
".gpt" appears a misnomer. Moreover, the the .raw suffix for these files
is already pretty popular (the Fedora disk images use it for example),
hence sounds like an OK scheme to adopt.
Given the write patterns on disk images, we better should turn COW off
for them. In particular as the file systems used inside the disk images
should do their own data integrity checks anyway and we don't need
multiple layers of it.
I figure "pull-dck" is not a good name, given that one could certainly
read the verb in a way that might be funny for 16year-olds. ;-)
Also, don't hardcode the index URL to use, make it runtime and configure
time configurable instead.
This adds a simply but powerful tool for downloading container images
from the most popular container solution used today. Use it like
this:
# systemd-import pull-dck mattdm/fedora
# systemd-nspawn -M fedora
This will donwload the layers for "mattdm/fedora", and make them
available locally as /var/lib/container/fedora.
The tool is pretty complete, as long as it's only about pulling down
images, or updating them. Pushing or searching is not supported yet.
Lennart Poettering