Chronograf 1.3.3 'There was a problem loading hosts' error when upgrading from 1.3.1

Is there a known problem with Chronograf where it says There was a problem loading hosts, and won’t connect to Kapacitor?

The Chronograf 1.3.1-alpine container I was using worked just fine with the same configuration.

Although, 1.3.1 throws a Local setting failed: TypeError: Cannot read property 'VERSION' of null, but it at least connects and sees hosts and alerts / alerting endpoints correctly.

Previous versions of Chronograf generated none of these errors.