Make the DiscoverNew switch only responsible for asserting the correct
data type, and push the conversion logic into the setKeys and updateKeys
methods.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This test was only testing behavior of the CLI itself (or even basic
functionality provided by Cobra).
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
- close idle connections after test
- don't discard failures to write upgrade response
- ignore errors in defer to make the linters happy
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Also:
- remove the hostConfig param from Daemon.createContainerVolumesOS.
- rename var container -> ctr
Signed-off-by: Rob Murray <rob.murray@docker.com>
The container's constructor, Daemon.newContainer, already has
hostConfig and can just assign it directly.
Signed-off-by: Rob Murray <rob.murray@docker.com>
Call registerMountPoints after the rest of the container's
configuration has been set up.
This will make it possible to call an NRI plugin with the
container's config, allowing it to adjust the mounts in that
config, before it's used to find volumes etc.
Signed-off-by: Rob Murray <rob.murray@docker.com>
Daemon.createContainerOSSpecificSettings adds container config for
the OS, and creates volumes. Split those two things.
This will make it possible to call an NRI plugin after the config
is complete, before volumes are created - so the NRI plugin can
adjust a full set of config, including volumes.
Signed-off-by: Rob Murray <rob.murray@docker.com>
The call from Daemon.create -> Daemon.setHostConfig acquired
container.Lock, but didn't need to because the container is
newly created and solely owned by the caller. The call from
Daemon.restore did not acquire the lock.
Signed-off-by: Rob Murray <rob.murray@docker.com>
- add early returns for `nil` body, `http.NoBody`, and `json.RawMessage`
- use `http.NoBody` instead of `nil` for empty bodies; it's more clear
on intent.
- use json.Encode instead of json.Encoder.Encode(), as we're marshaling
a single JSON document; this also avoid adding a trailing newline.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
This function was setting `text/plain` as default content-type for any
request that had a non-nil body.
However, this would also set the content-type if (e.g.) `http.NoBody` was set,
or if an empty reader was used, which would result in the daemon potentialy
rejecting the request, as it validates request to be using `application/json`;
d9ee22d1ab/daemon/server/httputils/httputils.go (L47-L58)
=== RUN TestCommitInheritsEnv
commit_test.go:30: assertion failed: error is not nil: Error response from daemon: unsupported Content-Type header (text/plain): must be 'application/json'
--- FAIL: TestCommitInheritsEnv (0.02s)
This patch removes setting the default content-type.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
When manually setting the API version to use, automatic API version
negotiation should no longer be performed. Instead of keeping track
of these options individually, we can mark negotiation to have happend
if either the version was set manually, or if API version negotiation
took place.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
If a container is using a docker_gwbridge endpoint as its gateway,
when it's connected to another network that provides a gateway, the
docker_gwbridge endpoint is removed when that endpoint is added (in
a recursive nightmare).
So, the "before" gateway for the container has been removed
before the new gateway is updateExternalConnectivity'd.
Don't pass the old gateway to updateExternalConnectivity in that
case, because the network driver's already forgotten about it.
Signed-off-by: Rob Murray <rob.murray@docker.com>
While a manual overridden version shouldn't perform automatic version
negotiation, the "ForceNegotiate" option could still be used to (re)
negotiate a version. This allows a client to be configured with an
initial API version, then triggered to perform API-version negotiation.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Migrated TestAPIImagesDelete from the legacy integration-cli suite
(docker_api_images_test.go) to the new integration test framework under
integration/image/remove_test.go.
This update:
- Fixes ENV instruction syntax to use "ENV FOO=bar"
- Adds error type check using errdefs.IsNotFound for cleaner assertions
- Ensures consistent cleanup handling
Signed-off-by: Aditya Mishra <mishraaditya675@gmail.com>
We've seen various failures recently where GitHub actions runners are
running out of space. Skip this test for now.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Environment-variables are expected to override config / defaults, so
make sure that the DOCKER_API_VERSION env-var always takes priority.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Make these options more strict to not allow arbitrary values. Historically,
the `DOCKER_API_VERSION` env-var did not perform any validation as it was
intended for testing-purposes, but given the wider use of this env-var,
we should perform some amount of validation.
Both `WithAPIVersion` and `WithAPIVersionFromEnv` still allow specifying
API versions that are not supported by the client for testing purposes
(e.g. to use API versions beyond `MinAPIVersion` and `MaxAPIVersion`),
but must be well-formed.
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Export trace spans from the github.com/microsoft/hcsshim module, which
is instrumented with OpenCensus, to the daemon's OpenTelemetry exporter
to provide more visibility into Windows container lifecycle operations.
Signed-off-by: Cory Snider <csnider@mirantis.com>
If the DNS name still resolves to an IP address, and that address is
assigned to a running container, the ping command will run indefinitely
and the test suite will time out for 10 mins.
This is confusing, as it looks like a daemon hang, or a test suite hang,
whereas it's just a test failure. Add '-c1' to ping to make it return
immediately.
Signed-off-by: Albin Kerouanton <albin.kerouanton@docker.com>
Previous commit reverted a faulty change that broke DNS resolution for
non swarm-scoped networks once a node has joined a Swarm cluster.
This commit adds an integration test to verify that we don't break DNS
resolution again.
Signed-off-by: Albin Kerouanton <albin.kerouanton@docker.com>
Commit a8b9eff90 removed a call to Network.updateSvcRecord from
Network.createEndpoint on the grounds that:
> all callers of Network.createEndpoint follow up with an Endpoint.Join,
> which also sets up the DNS entry.
However, the original call in Network.createEndpoint was gated by:
```
if !n.getController().isSwarmNode() || n.Scope() != scope.Swarm || !n.driverIsMultihost() {
n.updateSvcRecord(context.WithoutCancel(ctx), ep, true)
}
```
whereas the call in Endpoint.sbJoin() (invoked by Endpoint.Join()) is
gated by:
```
if !n.getController().isAgent() {
if !n.getController().isSwarmNode() || n.Scope() != scope.Swarm || !n.driverIsMultihost() {
n.updateSvcRecord(context.WithoutCancel(ctx), ep, true)
}
}
```
As a result, once a node has joined a Swarm cluster, no DNS entries are
created for non swarm-scoped networks.
Change the condition used by `sbJoin` to match the original condition
used in `createEndpoint`.
Signed-off-by: Albin Kerouanton <albin.kerouanton@docker.com>
