Shorthand method for service.restart()

[rycus86]

This PR would add a new shorthand method to restart a service, similarly to the recently added scale method.

Usage:

service = client.services.create(..)
service.restart()

Thanks!

[rycus86]

I think I've got some unrelated test failures on Jenkins:

22:45:48 [py3.6_17.12.0-ce] tests/integration/api_swarm_test.py::SwarmTest::test_init_swarm_with_ca_config FAILED

[shin-]

Yeah, sadly that test is hit or miss on 17.06 - don't worry about it.

[shin-]

Code LGTM, but not sure about the method's name.

[shin-]

I think restart doesn't accurately represent the operation that's being performed here. In the Docker ecosystem, restart usually means "take the existing containers with their configuration and stop them then start them again". The same idea exists in other common software like nginx with the reload vs restart distinction.

To my understanding. in this case, the containers for the service will actually be recreated with the same configuration - maybe force_update or reload is a more accurate name?

[rycus86]

Thanks again, @shin- !
I have renamed the method to force_update and changed the docstring to be the same (or very similar to) what the engine uses as help strings.

I've also snuck in another change: the force_update parameter on the service model's update method now accepts a bool, and if it is True then it does the counter increment logic. I thought it makes more sense to have it there, so you can just say service.update(force_update=True, **other_kwargs) and it's more inline with the docker service update --force flag being a boolean flag.

It would be great if you could give the change another look because of these extras. Thank you!

[shin-]

LGTM, thank you!

[shin-]

If you could just squash those commits into one, I can go ahead and merge it.

[rycus86]

Sure, it's done now, thanks a lot!