The report should specify both whether agent-side failover occurred and if it did, which master was successfully used.
The information that needs to be conveyed:
- Did failover occur?
- If so, which master was successfully used?
- Which masters were provided to use? (the server_urls list)
We're adding a 'producer' key to the report, to be synthesized in the puppetdb terminus, so we can tell if failover happened in a Simple HA setup. But the agent has a much better, more consistent perspective on this; something like 'failover' or 'the master I actually used' or similar could be useful.