I used it to MVP a site that didn’t gain traction so never achieved “high volume”; single process with multiple readers and single writer. My tests showed it be

I used it to MVP a site that didn’t gain traction so never achieved “high volume”; single process with multiple readers and single writer. My tests showed it beat Postgres by a large margin on mixed workload (reads and writes 50/50). The use case was “single db per account/user”. I’d use it again.