Perform an early validation check on both container checkpoint and
restore paths to fail-fast if the CRIU binary is missing or is older
than the minimum required version (3.16.0).
To support runtime-configured environments, the validation respects the
custom PATH from the shim manager environment if configured, skipping
any relative paths to avoid incorrect daemon-relative resolution. If not
configured, it falls back to a standard system PATH lookup. The check
result is cached using sync.Once to prevent redundant process spawning.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
(cherry picked from commit 186397511b)
Signed-off-by: Samuel Karp <samuelkarp@google.com>
The gRPC sandbox controller service only forwarded the `options` field
when calling the local controller. The `netns_path`, `rootfs`, and
`annotations` fields were silently dropped, causing clients using the
gRPC proxy path to receive incomplete sandbox configurations.
Event topics were missing the leading `/` prefix ("sandboxes/create"
instead of "/sandboxes/create"), causing the event exchange to reject
the publish and return an error to the caller.
Add unit tests for the controller service that exercise all RPC
methods.
Signed-off-by: William Myers <willmyrs@amazon.com>
(cherry picked from commit 557c54b116)
Signed-off-by: Phil Estes <estesp@gmail.com>
- Create /var/lib/containerd with 0o700 (was: 0o711).
- Create config.TempDir with 0o700 (was: 0o711).
- Create /run/containerd/io.containerd.grpc.v1.cri with 0o700 (was: 0o755).
- Create /run/containerd/io.containerd.sandbox.controller.v1.shim with 0o700 (was: 0o711).
- Leave /run/containerd and /run/containerd/io.containerd.runtime.v2.task created with 0o711,
as required by userns-remapped containers.
/run/containerd/io.containerd.runtime.v2.task/<NS>/<ID> is created with:
- 0o700 for non-userns-remapped containers
- 0o710 for userns-remapped containers with the remapped root group as the owner group.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
(cherry picked from commit 51b0cf11dc5af7ed1919beba259e644138b28d96)
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
Enabling the IMMUTABLE_FL file attribute causes dirty data to be
flushed synchronously at least on EXT4, which can greatly impact
container launch performance. In contrast, the overlayfs snapshotter
does not use syncfs by default.
Most users may not need IMMUTABLE_FL, let's make IMMUTABLE_FL optional
to align with the behavior of the overlayfs snapshotter and recover the
original performance.
1. tensorflow
Test commands:
$ nerdctl image pull --snapshotter=X --unpack="false" tensorflow/tensorflow:2.19.0
$ time nerdctl container --snapshotter=X run -d tensorflow/tensorflow:2.19.0 /bin/sh
Results:
overlayfs | 0m18.748s
erofs (no IMMUTABLE_FL) | 0m10.090s
erofs (with IMMUTABLE_FL) | 0m21.074s
2. ubuntu 22.04
Test commands:
$ nerdctl image pull --snapshotter=X --unpack="false" ubuntu:22.04
$ time nerdctl container --snapshotter=X run -d ubuntu:22.04 /bin/sh
Results:
overlayfs | 0m1.147s
erofs (no IMMUTABLE_FL) | 0m0.795s
erofs (with IMMUTABLE_FL) | 0m1.094s
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Derive filesystem UUIDs (`lsblk -o +UUID`) from the OCI layer digests
(although diffIDs are better in principle, but they're unavailable by
differs in advance) rather than generating a random one. This allows
EROFS to uniquely identify each layer using the content-addressable
filesystem UUID.
It can also be used for reproducible builds. To achieve this, configure
`mkfs_options` with `-T0 --mkfs-time` (However, `--mkfs-time` requires
erofs-utils 1.8+; Otherwise, all inode timestamps will be reset w/o it):
``` toml
[plugins."io.containerd.differ.v1.erofs"]
mkfs_options = ["-T0 --mkfs-time"]
```
Fixes: c73c8e5d52 ("Introduce EROFS differ")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Currently the error details are not included in the output error and
there is no log. One of the reasons a the transferer was skipped could
be do to a specific component which is not implemented (such as trying
to use erofs differ) or unsupported image (pulling schema1). This
information is useful to find a bad configuration.
Signed-off-by: Derek McGowan <derek@mcg.dev>
- adds a transfer service progress reporter to handle timeouts. Also other test fixes
- fallback to local image pull when configuration conflict
Signed-off-by: Tony Fang <nhfang@amazon.com>
Co-authored-by: Swagat Bora <sbora@amazon.com>
If the layer media type is expected as an EROFS native layer (ending
with `.erofs`), copy the content as the layer blob.
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Avoid introducing utils package outside of internal. This package
should not be imported by other modules.
Signed-off-by: Derek McGowan <derek@mcg.dev>
follow-up to 09f34d18b7
- Use strings.Cut instead of trimming prefixes and strings.Split
to reduce allocations
- Use a switch for mount-type for slightly better readability
than if / else if / else.
- Fix GoDoc to start with the function name.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Unlike the walking differ, which implements a generic method to
accommodate all kinds of snapshotters, the EROFS differ is just
implemented for EROFS and EROFS snapshotter so it can utilize the
recent DiffDirChanges() [1] to avoid traversing the entire rootfs
directory in order to improve `nerdctl commit` performance.
Additionally, I think `baseDir` is unnecessary too (in principle,
only `upperdir` is useful for OCI format convention). However,
addressing this requires more work, so left as is for now.
It's also useful to implement a customized Compare() method for
EROFS differ so that we can dump the native EROFS-formatted blob
to the content store later.
[1] https://github.com/containerd/continuity/pull/145
Signed-off-by: Gao Xiang <xiang@kernel.org>
Schema 1 (`application/vnd.docker.distribution.manifest.v1+prettyjws`) has been
officially deprecated since containerd v1.7 (PR 6884), and disabled since v2.0 (PR 9765).
Users who have been seeing warnings like `conversion from schema 1 images is deprecated`
now have to rebuild the image with Schema 2 or OCI.
Schema 2 was introduced in Docker 1.10 (Feb 2016), so most users should have been already
using Schema 2 or OCI.
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
To make it as DEPRECATED, this PR does the following:
1. Changes config default to use `NetworkPluginBinDirs`;
2. Mark `NetworkPluginBinDir` as deprecated (in config version 3);
3. Add config migration from 2 to 3, which migrates `bin_dir`
in version 2 to `bin_dirs` in version 3.
Signed-off-by: Jin Dong <djdongjin95@gmail.com>
[wip] add deprecation warning
Signed-off-by: Jin Dong <djdongjin95@gmail.com>
This implements container restore as described in:
https://kubernetes.io/blog/2022/12/05/forensic-container-checkpointing-alpha/#restore-checkpointed-container-standalone
For detailed step by step instruction also see contrib/checkpoint/checkpoint-restore-cri-test.sh
The code changes are based on changes I have done in Podman around 2018
and CRI-O around 2020.
The history behind restoring container via CRI/Kubernetes probably
requires some explanation. The initial proposal to bring
checkpoint/restore to Kubernetes was looking at pod checkpoint and
restoring and the corresponding CRI changes.
https://github.com/kubernetes-sigs/cri-tools/pull/662https://github.com/kubernetes/kubernetes/pull/97194
After discussing this topic for about two years another approach was
implemented as described in KEP-2008:
https://github.com/kubernetes/enhancements/issues/2008
"Forensic Container Checkpointing" allowed us to separate checkpointing
from restoring. For the "Forensic Container Checkpointing" it is enough
to create a checkpoint of the container. Restoring is not necessary as
the analysis of the checkpoint archive can happen without restoring the
container.
While thinking about a way to restore a container it was by coincidence
that we started to look into restoring containers in Kubernetes via
Create and Start. The way it was done in CRI-O is to figure out during
Create if the container image is a checkpoint image and if that is true
we are using another code path. The same was implemented now with this
change in containerd.
With this change it is possible to restore the container from a
checkpoint tar archive that is created during checkpointing via CRI.
To restore a container via Kubernetes we convert the tar archive to an
OCI image as described in the kubernetes.io blog post from above. Using
this OCI image it is possible to restore a container in Kubernetes.
At this point I think it should be doable to restore containers in
CRI-O and containerd no matter if they have been created by containerd or
CRI-O. The biggest difference is the container metadata and that can
be adapted during restore.
Open items:
* It is not clear to me why restoring a container in containerd goes
through task/Create(). But as the restore code already exists this
change extended the existing code path to restore a container in
task/Create() to also restore a container through the CRI via
Create and Start.
* Automatic image pulling. containerd does not pull images
automatically if created via the CRI. There is an option in
crictl to pull images before starting, but that uses the CRI
image pull interface. It is still a separate pull and create
operation. Restoring containers from an OCI image is a bit
different. The checkpoint OCI image does not include the base
image, but just a reference to the image (NAME@DIGEST).
Using crictl with pulling will enable the pulling of the
checkpoint image, but not of the base image the checkpoint is
based on. So during preparation of the checkpoint containerd
will automatically pull the base image, but I was not able how
to pull an image blockingly in containerd. So there is a for
loop waiting for the container image to appear in the internal
store. I think this probably can be implemented better.
Anyway, this is a first step towards container restored in Kubernetes
when using containerd.
Signed-off-by: Adrian Reber <areber@redhat.com>
Otherwise, the following error avoids snapshot GC:
level=warning msg="snapshot garbage collection failed" error="failed
to clear IMMUTABLE_FL: failed to open: open /var/lib/containerd/io.
containerd.snapshotter.v1.erofs/snapshots/2/layer.erofs: no such file
or directory
Fixes: b477cf8e97 ("erofs-snapshotter: protect layer blobs with FS_IMMUTABLE_FL")
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Currently, containerd cannot dynamically select between EROFS block
or file-based mounting approaches based on the specific runtime (or
the Linux kernel version of the runtime) due to its static mount
structure.
For example, the EROFS snapshotter fails on Linux 5.4 (Ubuntu 20.04)
with `bin/nerdctl run --net=host --snapshotter=erofs busybox:latest`:
FATA[0005] failed to mount {Type:erofs Source:/var/lib/containerd/
io.containerd.snapshotter.v1.erofs/snapshots/1/layer.erofs Target:
Options:[ro]} on "/tmp/initialC1374142795": block device required
Temporarily fix this by appending `-oloop` for single-layer images.
The upcoming mount manager will make it better [1].
[1] https://github.com/containerd/containerd/issues/11303
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
As documented in ioctl_iflags(2):
```
FS_IMMUTABLE_FL
The file is immutable: no changes are permitted to the file contents
or metadata (permissions, timestamps, ownership, link count, and so
on). (This restriction applies even to the superuser.)
```
For example, any user cannot delete/move layer blobs when
FS_IMMUTABLE_FL is set:
``` sh
# cd /var/lib/containerd/io.containerd.snapshotter.v1.erofs/snapshots/4
# mv layer{,1}.erofs
mv: cannot move 'layer.erofs' to 'layer1.erofs': Operation not permitted
# rm layer.erofs
rm: cannot remove 'layer.erofs': Operation not permitted
```
Note that it's a best-effort approach for data loss prevention. IOWs,
just warn out if FS_IMMUTABLE_FL cannot be set anyway (e.g., due to lack
of support in the underlying filesystem.)
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
