What do you like best about the product?

My best thing about ScyllaDB is the way it performs under heavy load. We've got a live analytics service we're utilizing, and ScyllaDB doesn't even break a sweat at a million hits per minute. The fact that it's C++-based means that it's more system-efficient than some NoSQL databases we've worked with in the past. We also save on hardware since we don't need as many servers to get the same results. Scaling also couldn't be simpler—horizontal scaling with ScyllaDB simply happens with no extra effort.

What do you dislike about the product?

My worst fear was how hard it is to learn. If you don't have knowledge of distributed systems or Cassandra-type databases, it's really intimidating to begin. Some of the advanced features are great, but not necessarily easy to grasp for a new user. And also, the tutorials or guides that do exist are generally based on a lot of pre-existing knowledge, so it's difficult for new users. I think more beginner tutorials or step-by-step guides would get a lot more people started.

What problems is the product solving and how is that benefiting you?

Our biggest challenge was scaling our real-time application without having to constantly think about downtime or system lag. ScyllaDB fixes this by offering a high-performance, horizontally scalable backend that can scale with our requirements. It supports high volumes of reads and writes with very minimal latency, which is precisely what we were looking for. Prior to ScyllaDB, we had slow query issues and nodes that were overloaded, but now both of these issues are largely a thing of the past. The efficiency of running on fewer resources also saves us money on cloud infrastructure, which allows us to put more money into optimizing other areas of our stack.