Since song history data is in-memory, it does not fully function properly when Cadence scales beyond one replica. Each replica maintains its own copy of song history.
If we moved this data into one of the databases, we would be properly separate API and data, and only keep one master copy of it instead of having each replica duplicate it.
Depends on #224.
Since song history data is in-memory, it does not fully function properly when Cadence scales beyond one replica. Each replica maintains its own copy of song history.
If we moved this data into one of the databases, we would be properly separate API and data, and only keep one master copy of it instead of having each replica duplicate it.
Depends on #224.