When I got early access to pgrust my first question was whether it had been tested for building read-only standby analytics questions. The problem has goals beyond that, but having a real analytics node available just by playing WAL data from your primary makes pgrust a low risk eval for your data. Just have to build confidence the results are connection, which is being applied by all the PG infrastructure regression tests the code has been trained on,
reply