I'm thrilled to see this, because the work Salvatore was putting into Redis-on-disk functionality had so little perceived payoff for us. This portends Redis not trying to create an idiosyncratic replacement for SQL, and focusing instead on its unique strengths.
If you want the cost/performance characteristics of disk storage, use MySQL or Postgres. If you want them and the speed and flexibility of the Redis data structure server, use Redis + SQL as a one-two punch.
We use Redis for "live" data, and SQL for archival/report/versioned data.
Comments
I'm thrilled to see this, because the work Salvatore was putting into Redis-on-disk functionality had so little perceived payoff for us. This portends Redis not trying to create an idiosyncratic replacement for SQL, and focusing instead on its unique strengths.
If you want the cost/performance characteristics of disk storage, use MySQL or Postgres. If you want them and the speed and flexibility of the Redis data structure server, use Redis + SQL as a one-two punch.
We use Redis for "live" data, and SQL for archival/report/versioned data.