This fix tries to address the request in 31325 by adding
`--filter mode=global|replicated` to `docker service ls`.
As `docker service ls` has a `MODE` column by default, it is natural
to support `--filter mode=global|replicated` for `docker service ls`.
There are multiple ways to address the issue. One way is to pass
the filter of mode to SwarmKit, another way is to process the filter
of mode in the daemon.
This fix process the filter in the daemon.
Related docs has been updated.
An integration test has been added.
This fix fixes 31325.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
This change adds the ability to do --tail and --since on docker service
logs. It wires up the API endpoints to each other and fixes some older
bugs. It adds integration tests for these new features.
Signed-off-by: Drew Erny <drew.erny@docker.com>
Right now getting logs from a service with an attached TTY does not
work. The behavior was undefined and caused the command to hang and
strange messages to occur in the daemon logs.
This returns errors, both deep in the swarmkit adapter (to guard against
undefined behavior, which is Bad) and in the daemon (to tell users that
the thing they're asking for is not possible).
Signed-off-by: Drew Erny <drew.erny@docker.com>
The response would never reach the client because it was being
redeclared in the current scope.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
The current behavior is for dockerd to fail to start if the swarm
component can't be started for some reason. This can be difficult to
debug remotely because the daemon won't be running at all, so it's not
possible to hit endpoints like /info to see what's going on. It's also
very difficult to recover from the situation, since commands like
"docker swarm leave" are unavailable.
Change the behavior to allow startup to proceed.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
When pinning by digest, the registry might be slow or unresponsive. This
could cause the context to already be expired by the time UpdateService
or CreateService is called. We want digest pinning to be a best-effort
operation, so it's problematic if a slow or misbehaving registry
prevents the service operation from completing. Replace the context
after communicating with the registry, so we have a fresh timeout for
the gRPC call.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
Server-side rollback can take advantage of the rollback-specific update
parameters, instead of being treated as a normal update that happens to
go back to a previous version of the spec.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
… in order to remove duplication.
Each time we update a cluster object, we do some common
operations (lock, verify it's on a manager, get the request context,
and the update). This introduce a method and refactor few
update/remove method that allows to duplicate less code.
Signed-off-by: Vincent Demeester <vincent@sbr.pm>
endpointSpecFromGRPC and endpointFromGRPC do the exact same thing for
endpoint{,Spec}.Ports, let's extract that to a method.
Signed-off-by: Vincent Demeester <vincent@sbr.pm>
This adds support for placement preferences in Swarm services.
- Convert PlacementPreferences between GRPC API and HTTP API
- Add --placement-pref, --placement-pref-add and --placement-pref-rm to CLI
- Add support for placement preferences in service inspect --pretty
- Add integration test
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
Before this change, doing service logs was just tossing the stream
selectors and always using the default (both streams). This change adds
a check for which streams the user wants and only includes those.
Fixes#31306
Signed-off-by: Drew Erny <drew.erny@docker.com>
I found that sometimes tasks would end up in a rejected state when
trying to update them quickly. The problem was that Shutdown could fail
if called before the container was started. Instead of returning an
error in this case, Shutdown should succeed. This allows tasks to
progress to the "shutdown" state as expected.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
This fix tries to address the issue raised in 25696 where
it was not possible to specify `--stop-signal` for `docker service create`
and `docker service update`, in order to use special signal to stop
the container.
This fix adds `--stop-signal` and update the `StopSignal` in `Config`
through `service create` and `service update`.
Related docs has been updated.
Integration test has been added.
This fix fixes 25696.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
- otherwise the attachment task will stay in store and
consume IP addresses and there is no way to remove it.
Signed-off-by: Alessandro Boch <aboch@docker.com>
Remove forked reference package. Use normalized named values
everywhere and familiar functions to convert back to familiar
strings for UX and storage compatibility.
Enforce that the source repository in the distribution metadata
is always a normalized string, ignore invalid values which are not.
Update distribution tests to use normalized values.
Signed-off-by: Derek McGowan <derek@mcgstyle.net> (github: dmcgowan)
While it is important to not create controllers for an invalid task,
certain properties should only be checked immediately before use. Early
host validation of mounts prevents resolution of the task Executor when
the mounts are not relevant to execution flow. In this case, we have a
check for the existence of a bind mount path in a creation function that
prevents a task controller from being resolved. Such early validation
prevents one from interacting directly with a controller and result in
unnecessary error reporting.
In accordance with the above, we move the validation of the existence of
host bind mount paths to the `Controller.Start` phase. We also call
these "checks", as they are valid mounts but reference non-existent
paths.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
