Derek McGowan
f75817eb3b
A layer served from the layer content cache was symlinked into the snapshot as its layer.erofs. That put a path inside the snapshot directory aliasing a blob owned by the operator and shared with every other snapshot of that layer. Everything written to that path went straight into the cache entry. Each write site then needed its own guard: the differ had to refuse to apply, Commit had to refuse to convert through a dangling link, and Remove had to avoid clearing IMMUTABLE_FL on a file it did not own. A missed guard corrupted content shared across images. A pruned entry left a dangling link that read as an absent blob on one path and as populated content on another. Record where the blob is on the snapshot and mount it from there. Nothing in the snapshot directory refers to content it does not own. A stray write creates a local file, conversion cannot reach the cache, and removal has nothing to follow. The guards Commit and Remove needed for a symlink are gone. The differ still refuses, on the read-only mounts the snapshotter hands out. resolveBlob is the one place a blob is located. It reports whether the blob is already populated, meaning there is nothing to apply or convert, and whether it belongs to this snapshot, meaning it may be written. Mounts are then assembled without consulting the filesystem to find out what a layer is. A source that does not resolve is an error on every path, including the parentless path in mounts(). A flattened parent's device= is resolved the same way. fsverity is only measured for a blob the snapshot owns. Measuring a cache entry would require the operator to enable fsverity on content this snapshotter never wrote. NewSnapshotter already rejects fsverity together with the layer content cache. The record is kept in labels outside the containerd.io/snapshot/ namespace. Labels in that namespace are inherited from image annotations, which would let an image name any path on the host. The same filter runs on everything the metadata snapshotter passes down. A client cannot set these labels, and cannot send them back after reading them through Stat. Commit and Update carry them across. The dm-verity sidecar is read from beside the cache entry. Nothing is copied into the snapshot. The sidecar cannot fall out of step with the blob it pins. An entry required to have one is rejected at Prepare. Signed-off-by: Derek McGowan <derek@mcg.dev>
containerd is an industry-standard container runtime with an emphasis on simplicity, robustness, and portability. It is available as a daemon for Linux and Windows, which can manage the complete container lifecycle of its host system: image transfer and storage, container execution and supervision, low-level storage and network attachments, etc.
containerd is a member of CNCF with 'graduated' status.
containerd is designed to be embedded into a larger system, rather than being used directly by developers or end-users.
Announcements
Now Recruiting
We are a large inclusive OSS project that is welcoming help of any kind shape or form:
- Documentation help is needed to make the product easier to consume and extend.
- We need OSS community outreach/organizing help to get the word out; manage and create messaging and educational content; and help with social media, community forums/groups, and google groups.
- We are actively inviting new security advisors to join the team.
- New subprojects are being created, core and non-core that could use additional development help.
- Each of the containerd projects has a list of issues currently being worked on or that need help resolving.
- If the issue has not already been assigned to someone or has not made recent progress, and you are interested, please inquire.
- If you are interested in starting with a smaller/beginner-level issue, look for issues with an
exp/beginnertag, for example containerd/containerd beginner issues.
Getting Started
See our documentation on containerd.io:
To get started contributing to containerd, see CONTRIBUTING.
If you are interested in trying out containerd see our example at Getting Started.
Nightly builds
There are nightly builds available for download here.
Binaries are generated from main branch every night for Linux and Windows.
Please be aware: nightly builds might have critical bugs, it's not recommended for use in production and no support provided.
Kubernetes (k8s) CI Dashboard Group
The k8s CI dashboard group for containerd contains test results regarding the health of kubernetes when run against main and a number of containerd release branches.
Runtime Requirements
Runtime requirements for containerd are very minimal. Most interactions with
the Linux and Windows container feature sets are handled via runc and/or
OS-specific libraries (e.g. hcsshim for Microsoft).
The current required version of runc is described in RUNC.md.
There are specific features used by containerd core code and snapshotters that will require a minimum kernel version on Linux. With the understood caveat of distro kernel versioning, a reasonable starting point for Linux is a minimum 4.x kernel version.
The overlay filesystem snapshotter, used by default, uses features that were finalized in the 4.x kernel series. If you choose to use btrfs, there may be more flexibility in kernel version (minimum recommended is 3.18), but will require the btrfs kernel module and btrfs tools to be installed on your Linux distribution.
To use Linux checkpoint and restore features, you will need criu installed on
your system. See more details in Checkpoint and Restore.
Build requirements for developers are listed in BUILDING.
Supported Registries
Any registry which is compliant with the OCI Distribution Specification is supported by containerd.
For configuring registries, see registry host configuration documentation
Features
For a detailed overview of containerd's core concepts and the features it supports, please refer to the FEATURES.MD document.
Releases and API Stability
Please see RELEASES.md for details on versioning and stability of containerd components.
Downloadable 64-bit Intel/AMD binaries of all official releases are available on our releases page. Binaries are published for additional platforms as well; see Platform Support for the full list of platforms and the level of support for each.
For other architectures and distribution support, you will find that many Linux distributions package their own containerd and provide it across several architectures, such as Canonical's Ubuntu packaging.
Enabling command auto-completion
Starting with containerd 1.4, the urfave client feature for auto-creation of bash and zsh
autocompletion data is enabled. To use the autocomplete feature in a bash shell for example, source
the autocomplete/ctr file in your .bashrc, or manually like:
$ source ./contrib/autocomplete/ctr
Distribution of ctr autocomplete for bash and zsh
For bash, copy the contrib/autocomplete/ctr script into
/etc/bash_completion.d/ and rename it to ctr. The zsh_autocomplete
file is also available and can be used similarly for zsh users.
Provide documentation to users to source this file into their shell if
you don't place the autocomplete file in a location where it is automatically
loaded for the user's shell environment.
CRI
cri is a containerd plugin implementation of the Kubernetes container runtime interface (CRI). With it, you are able to use containerd as the container runtime for a Kubernetes cluster.
CRI Status
cri is a native plugin of containerd. Since containerd 1.1, the cri plugin is built into the release binaries and enabled by default.
The cri plugin has reached GA status, representing that it is:
- Feature complete
- Works with Kubernetes 1.10 and above
- Passes all CRI validation tests.
- Passes all node e2e tests.
- Passes all e2e tests.
See results on the containerd k8s test dashboard
Validating Your cri Setup
A Kubernetes incubator project, cri-tools, includes programs for exercising CRI implementations. More importantly, cri-tools includes the program critest which is used for running CRI Validation Testing.
CRI Guides
- Installing with Ansible and Kubeadm
- For Non-Ansible Users, Performing a Custom Installation Using the Release Tarball and Kubeadm
- CRI Plugin Testing Guide
- Debugging Pods, Containers, and Images with
crictl - Configuring
criPlugins - Configuring containerd
Communication
For async communication and long-running discussions please use issues and pull requests on the GitHub repo. This will be the best place to discuss design and implementation.
For sync communication catch us in the #containerd and #containerd-dev Slack channels on Cloud Native Computing Foundation's (CNCF) Slack - cloud-native.slack.com. Everyone is welcome to join and chat. Get Invite to CNCF Slack.
Join our next community meeting hosted on Zoom. The schedule is posted on the CNCF Calendar (search 'containerd' to filter).
Security audit
Security audits for the containerd project are hosted on our website. Please see the security page at containerd.io for more information.
Reporting security issues
Please follow the instructions at containerd/project
Licenses
The containerd codebase is released under the Apache 2.0 license. The README.md file and files in the "docs" folder are licensed under the Creative Commons Attribution 4.0 International License. You may obtain a copy of the license, titled CC-BY-4.0, at http://creativecommons.org/licenses/by/4.0/.
Project details
containerd is the primary open source project within the broader containerd GitHub organization.
However, all projects within the repo have common maintainership, governance, and contributing
guidelines which are stored in a project repository commonly for all containerd projects.
Please find all these core project documents, including the:
information in our containerd/project repository.
Adoption
Interested to see who is using containerd? Are you using containerd in a project? Please add yourself via pull request to our ADOPTERS.md file.