Reusing k8s existing implementation. This could be re-written
to do port-forwarding natively rather than relying on socat/nsenter.
Signed-off-by: Mrunal Patel <mrunalp@gmail.com>
When powering off the system, we want the ocid service, to shutdown
all containers running on the system so they can cleanup properly
This patch will cleanup all pods on poweroff.
The ocid-shutdown.service drops a file /var/run/ocid.shutdown when the system
is shutting down. The ocid-shutdown.service should only be executed at system
shutdown.
On bootup sequence should be
start ocid.service
start ocid-shutdown.service (This is a NO-OP)
On system shutdown
stop ocid-shutdown.service (Creates /var/run/ocid.shutdown)
stop ocid.service (Notices /var/run/ocid.service and stops all pods before exiting.)
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
Fix the following upstream k8s's e2e-node test:
```
should be able to pull from private registry with secret [Conformance]
```
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
When RunPodSandbox fails after calling s.addSandbox(sb),
we're left with a sandbox in s.state.sandboxes while the
sandbox is not created.
We fix that by adding removeSandbox() to the deferred cleanup
call
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
This patch prevents k8s's e2e_node tests from killing CRI-O (because of
a panic in marshaling nil responses). This will ensure tests keep
running and just logging the failure.
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
Because kubelet will create broken symlinks for logPath it is necessary
to remove those symlinks before we attempt to write to them. This is a
temporary workaround while the issue is fixed upstream.
Ref: https://issues.k8s.io/44043
Signed-off-by: Aleksa Sarai <asarai@suse.de>
The main purpose of these tests is to make sure that the log actually
contains output from the container. We don't test the timestamps or the
stream that's stated at the moment.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
This adds a very simple implementation of logging within conmon, where
every buffer read from the masterfd of the container is also written to
the log file (with errors during writing to the log file ignored).
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Interleaving asynchronous updates with pod or container creations can
lead to unrecoverable races and corruptions of the pod or container hash
tables. This is fixed by serializing update against pod or container
creation operations, while pod and container creation operations can
run in parallel.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Now that the image package has fixes to support docker images v2s1,
we can remove our buildOCIProcessARgs() hack for empty image configs
and simplify this routine.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
We want new sandboxes to be added to the sandbox hash table before
adding their ID to the pod Index registrar, in order to avoid potential
Update() races.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
After change in `Makefile` at #304 where `PREFIX`
has changed from `/usr` to `/usr/local` these changes
has to be reflected in default `conmonPath`.
Signed-off-by: Suraj Deshmukh <surajssd009005@gmail.com>
When a pod sandbox comes with DNS settings, the resulting resolv.conf
file needs to be bind mounted in all pod containers under
/etc/resolv.conf.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
When we get a pod with DNS settings, we need to build
a resolv.conf file and mount it in all pod containers.
In order to do that, we have to track the built resolv.conf
file and store/load it.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
containers/storage is defaulting to /var/lib/containers/storage
for image and containers storage. It is also defaulting to
/var/run/containers/storage for all runtime. The defaults
for CRI-O should match so that lots of other tools that use
containers/storage can share the same storage.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
We have moved selinux support out of opencontainers/runc into its
own package. This patch moves to using the new selinux go bindings.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
We need to support a 2x2 matrix of use cases with both
kubelet giving us (command, args) slices and the OCI
image config file giving us (ENTRYPOINT, CMD) slices.
Here we always prioritize the kubelet information over
the OCI image one, and use the latter when the former
is incomplete.
Not that this routine will be slightly simpler when
issue #395 is fixed.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
When starting pods or containers, we create the mount points
first. It seems natural to do something symetrical when stopping
pods or containers, i.e. removing the mount point at last.
Also, the current logic may not work with VM based containers as the
hypervisor may hold a reference on the mount point while we're trying to
remove them.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
The way we build the OCI Process Args slice is incorrect.
With the current implementation we may for example end up building this
slice with only the entry point arguments, if the kubelet passed
information is missing the Command slice.
We also will end up building the Args slice with the Image config
process arguments, without the defined entry point, if kubelet does not
tell us anything about the container process command to be run.
This patch fixes that by favoring the kubelet ContainerConfig
information. If that is missing, we try to complete it with the
container image information. We always use ContainerConfig.Command[] or
ImageConfig.EntryPoint[] as the first OCI Process Args slice entries.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
The pause container is creating an AVC since the /dev/null device
is not labeled correctly. Looks like we are only setting the label of
the process not the label of the content inside of the container.
This change will label content in the pause container correctly and
eliminate the AVC.
Signed-off-by: Daniel J Walsh <dwalsh@redhat.com>
The sandbox privileged flag is set to true only if either the
pod configuration privileged flag is set to true or when any
of the pod namespaces are the host ones.
A container inherit its privileged flag from its sandbox, and
will be run by the privileged runtime only if it's set to true.
In other words, the privileged runtime (when defined) will be
when one of the below conditions is true:
- The sandbox will be asked to run at least one privileged container.
- The sandbox requires access to either the host IPC or networking
namespaces.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
We add a privileged flag to the container and sandbox structures
and can now select the appropriate runtime path for any container
operations depending on that flag.
Here again, the default runtime will be used for non privileged
containers and for privileged ones in case there are no privileged
runtime defined.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Not all runtimes are able to handle some of the kubelet
security context options, in particular the ones granting
host privileges to containers.
By adding a host privileged runtime path configuration, we
allow ocid to use a different runtime for host privileged
operations like e.g. host namespaces access.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
kubelet sends a request to create a container with an image ID (as
opposed as an image name). That ID comes from the ImageStatus response.
This patch fixes that by setting the image ID as well as the image name
and fix the login to lookup for image ID as well.
Found while running `make test-e2e-node`.
Signed-off-by: Antonio Murdaca <runcom@redhat.com>
Previously ocicni did not have support for setting the plugin directory.
Now that it has grown support for it, use it to actually respect the
setting a user has provided for ocid.network.* options.
Signed-off-by: Aleksa Sarai <asarai@suse.de>
The CRI doesn't expect us to implicitly pull an image if it isn't
already present before we're asked to use it to create a container, and
the tests no longer depend on us doing so, either.
Limit the logic which attempts to pull an image, if it isn't present, to
only pulling the configured "pause" image, since our use of that image
for running pod sandboxes is an implementation detail that our clients
can't be expected to know or care about. Include the name of the image
that we didn't pull in the error we return when we don't pull one.
Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
Use containers/storage to store images, pod sandboxes, and containers.
A pod sandbox's infrastructure container has the same ID as the pod to
which it belongs, and all containers also keep track of their pod's ID.
The container configuration that we build using the data in a
CreateContainerRequest is stored in the container's ContainerDirectory
and ContainerRunDirectory.
We catch SIGTERM and SIGINT, and when we receive either, we gracefully
exit the grpc loop. If we also think that there aren't any container
filesystems in use, we attempt to do a clean shutdown of the storage
driver.
The test harness now waits for ocid to exit before attempting to delete
the storage root directory.
Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
Vendor updated containers/image and containers/storage, along
with any new dependencies they drag in, and updated versions of other
dependencies that happen to get pulled in.
github.com/coreos/go-systemd/daemon/SdNotify() now takes a boolean to
control whether or not it unsets the NOTIFY_SOCKET variable from the
calling process's environment. Adapt.
github.com/opencontainers/runtime-tools/generate/Generator.AddProcessEnv()
now takes the environment variable name and value as two arguments, not
one. Adapt.
Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
The default configuration can only be accessed from the cmd/server
package, which cannot be imported (since it's a "package main").
This change promotes DefaultConfig() to the "server" package.
Closes: #315
Signed-off-by: Jonathan Yu <jawnsy@redhat.com>
When removing a pod sandbox or container, remove the ID of the item from
the corresponding ID index, so that we can correctly determine if it was
us or another actor that cleaned them up.
Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
We add 2 ocid options for choosing the CNI configuration and plugin
binaries directories: --cni-config-dir and --cni-plugin-dir.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Update the versions of containers/storage and containers/image, and add
new dependencies that they pull in.
Signed-off-by: Nalin Dahyabhai <nalin@redhat.com>
ns.Close() will not remove and unmount the networking namespace
if it's not currently marked as mounted.
When we restore a sandbox, we generate the sandbox netns from
ns.GetNS() which does not mark the sandbox as mounted.
There currently is a PR open to fix that in the ns package:
https://github.com/containernetworking/cni/pull/342
but meanwhile this patch fixes a netns leak when restoring a pod.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>