Starting with kernel v6.12, kernel memory TCP accounting is deprecated for cgroups v1.
Note: kernel memory TCP accounting is not supported by cgroups v2.
See d046ff46ee
Signed-off-by: Austin Vazquez <austin.vazquez@docker.com>
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
kernel-memory limits are not supported in cgroups v2, and were obsoleted in
[kernel v5.4], producing a `ENOTSUP` in kernel v5.16. Support for this option
was removed in runc and other runtimes, as various LTS kernels contained a
broken implementation, resulting in unpredictable behavior.
We deprecated this option in [moby@b8ca7de], producing a warning when used,
and actively ignore the option since [moby@0798f5f].
Given that setting this option had no effect in most situations, we should
just remove this option instead of continuing to handle it with the expectation
that a runtime may still support it.
Note that we still support RHEL 8 (kernel 4.18) and RHEL 9 (kernel 5.14). We
no longer build packages for Ubuntu 20.04 (kernel 5.4) and Debian Bullseye 11
(kernel 5.10), which still have an LTS / ESM programme, but for those it would
only impact situations where a runtime is used that still supports it, and
an old API version was used.
[kernel v5.4]: https://github.com/torvalds/linux/commit/0158115f702b0ba208ab0
[moby@b8ca7de]: b8ca7de823
[moby@0798f5f]: 0798f5f5cf
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This package is used internally by the daemon, and was only used out
side of the daemon by pkg/plugins (for which we still need to look
where it should be kept).
Making it internal because it's trivial to implement if needed by
anyone. The only reason it's a package is to keep it central, and
to make it easier to discover where we have rootlesskit-specific
codepaths in our codebase.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
These comments were added to enforce using the correct import path for
our packages ("github.com/docker/docker", not "github.com/moby/moby").
However, when working in go module mode (not GOPATH / vendor), they have
no effect, so their impact is limited.
Remove these imports in preparation of migrating our code to become an
actual go module.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
For kernels that don't have CONFIG_IP_NF_RAW, if the env
var DOCKER_INSECURE_NO_IPTABLES_RAW is set to "1", don't
try to create raw rules.
This means direct routing to published ports is possible
from other hosts on the local network, even if the port
is published to a loopback address.
Signed-off-by: Rob Murray <rob.murray@docker.com>
When a NAT-based port mapping is created with a HostIP specified, we
insert a DNAT rule in nat-DOCKER to replace the dest addr with the
container IP. Then, in filter chains, we allow access to the container
port for any packet not coming from the container's network itself (if
hairpinning is disabled), nor from another host bridge.
However we don't set any rule that prevents a rogue neighbor that shares
a L2 segment with the host, but not the one where the port binding is
expected to be published, from sending packets destined to that HostIP.
For instance, if a port binding is created with HostIP == '127.0.0.1',
this port should not be accessible from anything but the lo interface.
That's currently not the case and this provides a false sense of
security.
Since nat-DOCKER mangles the dest addr, and the nat table rejects DROP
rules, this change adds rules into raw-PREROUTING to filter ingress
packets destined to mapped ports based on the input interface, the dest
addr and the dest port.
Interfaces are dynamically resolved when packets hit the host, thanks
to iptables' addrtype extension. This extension does a fib lookup of the
dest addr and checks that it's associated with the interface reached.
Also, when a proxy-based port mapping is created, as is the case when an
IPv6 HostIP is specified but the container is only IPv4-capable, we
don't set any sort of filtering. So the same issue might happen. The
reason is a bit different - in that case, that's just how the kernel
works. But, in order to stay consistent with NAT-based mappings, these
rules are also applied.
The env var `DOCKER_DISABLE_INPUT_IFACE_FILTERING` can be set to any
true-ish value to globally disable this behavior.
Signed-off-by: Albin Kerouanton <albinker@gmail.com>
Historically, the `bridge` network-driver would detect whether netfiltering
was enabled in the kernel or, if disabled, try to do a `modprobe` when
initializing the driver. This approach became problematic, as loading the
module was not always performed at startup depending on daemon configuration,
or the daemon may have failed to load the module. The `/info` response
would include a warning to inform the user that some functionality may not
be available;
WARNING: bridge-nf-call-iptables is disabled
WARNING: bridge-nf-call-ip6tables is disabled
Starting with db25b0dcd0, detecting whether
netfiltering is enabled now [happens when needed][1], which was further improved
on to not depend on `modprobe` in 264c15bfc4 and
4740820716.
Because of the above, the `/info` output would now return warnings in any
situation where netfiltering was not enabled on the host before the daemon
started, which may be either _incorrect_ (i.e., the module may have been
loaded afterwards), or irrelevant, because netfiltering is not needed in
all situations.
This patch removes the warnings from the `/info` response,
[1]: 944e403502/libnetwork/drivers/bridge/setup_bridgenetfiltering.go (L16-L77)
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The `Commit` type was introduced in 2790ac68b3,
to assist triaging issues that were reported with an incorrect version of
runc or containerd. At the time, both `runc` and `containerd` were not yet
stable, and had to be built from a specific commit to guarantee compatibility.
We encountered various situations where unexpected (and incompatible) versions
of those binaries were packaged, resulting in hard to trace bug-reports.
For those situations, a "expected" version was set at compile time, to
indicate if the version installed was different from the expected version;
docker info
...
runc version: a592beb5bc4c4092b1b1bac971afed27687340c5 (expected: 69663f0bd4b60df09991c08812a60108003fa340)
Both `runc` and `containerd` are stable now, and docker 19.03 and up set the
expected version to the actual version since c65f0bd13c
and 23.0 did the same for the `init` binary b585c64e2b,
to prevent the CLI from reporting "unexpected version".
In short; the `Expected` fields no longer serves a real purpose.
In future, we can even consider deprecating the `ContainerdCommit`, `RuncCommit`
and `InitCommit` fields on the `/info` response (as we also include this
information as part of the components returned in `/version`), but those
can still be useful currently for situations where a user only provides
`docker info` output.
This patch starts with deprecating the `Expected` field.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
I was trying to find out why `docker info` was sometimes slow so
plumbing a context through to propagate trace data through.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
The github.com/containerd/containerd/log package was moved to a separate
module, which will also be used by upcoming (patch) releases of containerd.
This patch moves our own uses of the package to use the new module.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This utility was only used in a single location (as part of `docker info`),
but the `pkg/rootless` package is imported in various locations, causing
rootlesskit to be a dependency for consumers of that package.
Move GetRootlessKitClient to the daemon code, which is the only location
it was used.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The daemon has made a habit of mutating the DefaultRuntime and Runtimes
values in the Config struct to merge defaults. This would be fine if it
was a part of the regular configuration loading and merging process,
as is done with other config options. The trouble is it does so in
surprising places, such as in functions with 'verify' or 'validate' in
their name. It has been necessary in order to validate that the user has
not defined a custom runtime named "runc" which would shadow the
built-in runtime of the same name. Other daemon code depends on the
runtime named "runc" always being defined in the config, but merging it
with the user config at the same time as the other defaults are merged
would trip the validation. The root of the issue is that the daemon has
used the same config values for both validating the daemon runtime
configuration as supplied by the user and for keeping track of which
runtimes have been set up by the daemon. Now that a completely separate
value is used for the latter purpose, surprising contortions are no
longer required to make the validation work as intended.
Consolidate the validation of the runtimes config and merging of the
built-in runtimes into the daemon.setupRuntimes() function. Set the
result of merging the built-in runtimes config and default default
runtime on the returned runtimes struct, without back-propagating it
onto the config.Config argument.
Signed-off-by: Cory Snider <csnider@mirantis.com>
The existing runtimes reload logic went to great lengths to replace the
directory containing runtime wrapper scripts as atomically as possible
within the limitations of the Linux filesystem ABI. Trouble is,
atomically swapping the wrapper scripts directory solves the wrong
problem! The runtime configuration is "locked in" when a container is
started, including the path to the runC binary. If a container is
started with a runtime which requires a daemon-managed wrapper script
and then the daemon is reloaded with a config which no longer requires
the wrapper script (i.e. some args -> no args, or the runtime is dropped
from the config), that container would become unmanageable. Any attempts
to stop, exec or otherwise perform lifecycle management operations on
the container are likely to fail due to the wrapper script no longer
existing at its original path.
Atomically swapping the wrapper scripts is also incompatible with the
read-copy-update paradigm for reloading configuration. A handler in the
daemon could retain a reference to the pre-reload configuration for an
indeterminate amount of time after the daemon configuration has been
reloaded and updated. It is possible for the daemon to attempt to start
a container using a deleted wrapper script if a request to run a
container races a reload.
Solve the problem of deleting referenced wrapper scripts by ensuring
that all wrapper scripts are *immutable* for the lifetime of the daemon
process. Any given runtime wrapper script must always exist with the
same contents, no matter how many times the daemon config is reloaded,
or what changes are made to the config. This is accomplished by using
everyone's favourite design pattern: content-addressable storage. Each
wrapper script file name is suffixed with the SHA-256 digest of its
contents to (probabilistically) guarantee immutability without needing
any concurrency control. Stale runtime wrapper scripts are only cleaned
up on the next daemon restart.
Split the derived runtimes configuration from the user-supplied
configuration to have a place to store derived state without mutating
the user-supplied configuration or exposing daemon internals in API
struct types. Hold the derived state and the user-supplied configuration
in a single struct value so that they can be updated as an atomic unit.
Signed-off-by: Cory Snider <csnider@mirantis.com>
Ensure data-race-free access to the daemon configuration without
locking by mutating a deep copy of the config and atomically storing
a pointer to the copy into the daemon-wide configStore value. Any
operations which need to read from the daemon config must capture the
configStore value only once and pass it around to guarantee a consistent
view of the config.
Signed-off-by: Cory Snider <csnider@mirantis.com>
commit dc11d2a2d8 removed the devicemapper
storage-driver, so these warnings are no longer relevant.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
commit 0abb8dec3f removed support for
running overlay/overlay2 on top of a backing filesystem without d_type
support, and turned it into a fatal error when starting the daemon,
so there's no need to generate warnings for this situation.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Extended attributes are set on files in container images for a reason.
Fail to unpack if extended attributes are present in a layer and setting
the attributes on the unpacked files fails for any reason.
Add an option to the vfs graph driver to opt into the old behaviour
where ENOTSUPP and EPERM errors encountered when setting extended
attributes are ignored. Make it abundantly clear to users and anyone
triaging their bug reports that they are shooting themselves in the
foot by enabling this option.
Signed-off-by: Cory Snider <csnider@mirantis.com>
This option was deprecated in 5a922dc162, which
is part of the v24.0.0 release, so we can remove it from master.
This patch;
- adds a check to ValidatePlatformConfig, and produces a fatal error
if oom-score-adjust is set
- removes the deprecated libcontainerd/supervisor.WithOOMScore
- removes the warning from docker info
With this patch:
dockerd --oom-score-adjust=-500 --validate
Flag --oom-score-adjust has been deprecated, and will be removed in the next release.
unable to configure the Docker daemon with file /etc/docker/daemon.json: merged configuration validation from file and command line flags failed: DEPRECATED: The "oom-score-adjust" config parameter and the dockerd "--oom-score-adjust" options have been removed.
And when using `daemon.json`:
dockerd --validate
unable to configure the Docker daemon with file /etc/docker/daemon.json: merged configuration validation from file and command line flags failed: DEPRECATED: The "oom-score-adjust" config parameter and the dockerd "--oom-score-adjust" options have been removed.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The `oom-score-adjust` option was added in a894aec8d8,
to prevent the daemon from being OOM-killed before other processes. This
option was mostly added as a "convenience", as running the daemon as a
systemd unit was not yet common.
Having the daemon set its own limits is not best-practice, and something
better handled by the process-manager starting the daemon.
Commit cf7a5be0f2 fixed this option to allow
disabling it, and 2b8e68ef06 removed the default
score adjust.
This patch deprecates the option altogether, recommending users to set these
limits through the process manager used, such as the "OOMScoreAdjust" option
in systemd units.
With this patch:
dockerd --oom-score-adjust=-500 --validate
Flag --oom-score-adjust has been deprecated, and will be removed in the next release.
configuration OK
echo '{"oom-score-adjust":-500}' > /etc/docker/daemon.json
dockerd
INFO[2023-04-12T21:34:51.133389627Z] Starting up
INFO[2023-04-12T21:34:51.135607544Z] containerd not running, starting managed containerd
WARN[2023-04-12T21:34:51.135629086Z] DEPRECATED: The "oom-score-adjust" config parameter and the dockerd "--oom-score-adjust" option will be removed in the next release.
docker info
Client:
Context: default
Debug Mode: false
...
DEPRECATED: The "oom-score-adjust" config parameter and the dockerd "--oom-score-adjust" option will be removed in the next release
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The `docker-init` binary is not intended to be a user-facing command, and as such it is more appropriate for it to be found in `/usr/libexec` (or similar) than in `PATH` (see the FHS, especially https://refspecs.linuxfoundation.org/FHS_3.0/fhs/ch04s07.html and https://refspecs.linuxfoundation.org/FHS_2.3/fhs-2.3.html#USRLIBLIBRARIESFORPROGRAMMINGANDPA).
This adjusts the logic for using that configuration option to take this into account and appropriately search for `docker-init` (or the user's configured alternative) in these directories before falling back to the existing `PATH` lookup behavior.
This behavior _used_ to exist for the old `dockerinit` binary (of a similar name and used in a similar way but for an alternative purpose), but that behavior was removed in 4357ed4a73 when that older `dockerinit` was also removed.
Most of this reasoning _also_ applies to `docker-proxy` (and various `containerd-xxx` binaries such as the shims), but this change does not affect those. It would be relatively straightforward to adapt `LookupInitPath` to be a more generic function such as `libexecLookupPath` or similar if we wanted to explore that.
See 14482589df/cli-plugins/manager/manager_unix.go for the related path list in the CLI which loads CLI plugins from a similar set of paths (with a similar rationale - plugin binaries are not typically intended to be run directly by users but rather invoked _via_ the CLI binary).
Signed-off-by: Tianon Gravi <admwiggin@gmail.com>
Since runtimes can now just be containerd shims, we need to check if the
reference is possibly a containerd shim.
Signed-off-by: Brian Goff <cpuguy83@gmail.com>
Older versions of Go don't format comments, so committing this as
a separate commit, so that we can already make these changes before
we upgrade to Go 1.19.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
These checks were added when we required a specific version of containerd
and runc (different versions were known to be incompatible). I don't think
we had a similar requirement for tini, so this check was redundant. Let's
remove the check altogether.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
We have upgraded runc to rc93 and added CI for cgroup 2.
So we can move cgroup v2 out of experimental.
Fix issue 41916
Signed-off-by: Akihiro Suda <akihiro.suda.cz@hco.ntt.co.jp>
