Updates UCP release notes and Upgrade docs

Fixes release notes to mention need to force reset browser after upgrade, as
well as still requiring --swarm-experimental flag for rescheduling. Fixes
upgrade docs with browser refresh. This adds a note for users to flush browser
cache after an upgrade to avoid API/UI mismatch.

installation/upgrade.md

@@ -152,6 +152,7 @@ replica nodes):
     2. Make sure that node is healthy, using the instructions
     in [step 4](#no-ha-step-4).
 
+6. Make sure to flush your browser's cache prior to to running the upgraded UCP in the UI.
 
 ### My cluster is set for high-availability
 
@@ -269,6 +270,7 @@ If your cluster is set up for high-availability (has several replica nodes):
     2. Make sure the node is healthy, using the instructions
     in [step 7](#ha-step-7).
 
+11. Make sure to flush your browser's cache prior to to running the upgraded UCP in the UI.
 
 ## Where to go next
 

release_notes.md

@@ -62,8 +62,6 @@ weight=110
 
 **Misc**
 
-* You no longer need to install UCP with --swarm-experimental to use swarm
-container rescheduling,
 * All UCP containers now have the 'com.docker.ucp.version' label with their
 upstream version or UCP version,
 * When running docker/ucp in interactive mode, the parameters and environment
@@ -77,6 +75,8 @@ The first is deprecated but still available.
 a success message is displayed, but that node will not be part of the
 cluster. As a workaround, join new controller nodes before upgrading, or
 perform a fresh installation of UCP 1.1.0.
+* If you have an active login session in UCP and do an upgrade, you should force
+refresh the browser or you may run into UI errors.
 * When joining replicas to the cluster, you may be prompted to restart the
 Docker daemon on that node. For a faster installation, only restart the Docker
 daemon after joining all replicas.