DR: document process for planned failover

We're making it easier to promote a secondary to primary in 2-node (1 primary, 1 secondary) configurations https://gitlab.com/gitlab-org/gitlab-ee/issues/1921 and for n-node (1 primary, n-1 secondary) configurations https://gitlab.com/gitlab-org/gitlab-ee/issues/4130. This might be done in unplanned disaster scenarios, but can also be used for scheduled migrations from one location to another.

In unplanned situations it is possible for some data to be lost because people µay be using the primary up until and even during the disaster. However, for planned migrations, we should make recommendations about this.