Problem has been solved and everything is back to functioning as expected.

The problem was caused by the deployment of a new feature which needed more time to deploy then expected.

We're seeing increased error rates when accessing the assets view in Analytics and Health, and accessing the instrumentation endpoint on the API.