Provide support for the terminationGracePeriodSeconds field

[ErikMogensen]

Proposal

I want to be able to set the terminationGracePeriodSeconds for the deployment resource.

Use-Case

I want it be able to set it above the default 30 seconds. For instance if a backend service takes three minutes to respond (a bit crazy but that happens) and the pod is sent a kill signal, responses may be lost. Therefore we want to be able to set it.

Also, to maintain compatibility with the yaml generated in the portal, which sets this value to 60.

Anything else?

I am able to provide a PR for this.