ScyllaDB is one of the best of its kind when it comes to its performance and efficient resource utilization. Its ability to use modern hardware and deliver very high throughput with extremely low latency is very impressive. It seamlessly handles large volumes of data and concurrent requests while performing better than other traditional NoSQL databases. Plus, it is easy to integrate with the existing workflows. Review collected by and hosted on G2.com.

While it is such a great database for handling large chunks of data, the operational challenge can be a bottleneck for newly distributed teams. Review collected by and hosted on G2.com.

I use ScyllaDB for efficient storage. I appreciate its tail latency guarantees and the fact that it scales well beyond my needs, so I don't need to worry about migrating anywhere. I like its compatibility with CQL and the custom extensions that are performance-oriented, such as bypassing the cache for certain queries. I find their Rust driver to be top class for both ScyllaDB and Cassandra, and it's important because latency-sensitive systems are often built in Rust. I also enjoy that the whole ecosystem of tools that works with Cassandra works with ScyllaDB too. The initial setup was easy to follow from the docs. Review collected by and hosted on G2.com.

I sometimes wish for more SQL compatibility, like being able to send a complex query that joins multiple tables. Review collected by and hosted on G2.com.

I like ScyllaDB for its high performance and the ability to handle massive amounts of security data without slowing down. Its low latency querying is valuable because it allows me to retrieve information almost instantly during investigations. I also appreciate its speed, scalability, and reliability, which are critical for my cybersecurity projects. Review collected by and hosted on G2.com.

Configuring nodes, replication, and monitoring in ScyllaDB was a bit intimidating, especially under the hackathons' time frame. Setting up isn't always straightforward and it presents a big learning curve. Initial setup was not easy. Simplifying cluster wizards for quickly setting up smaller clusters and pre-configured templates for common scenarios would help. Review collected by and hosted on G2.com.

I am currently learning about ScyllaDB for use in a server environment for data storage. I like that it's easy to manage and that it delivers low latency and high performance even under heavy workloads. ScyllaDB stands out because it is built to stay fast and predictable at scale, not only in ideal conditions but also under real pressure. The initial setup of ScyllaDB was absolutely perfect. Review collected by and hosted on G2.com.

Since I'm still early in learning ScyllaDB, there are a few areas where I would definitely want more depth and clarity as I go further. Like understanding how to properly design tables for high performance workload. Review collected by and hosted on G2.com.

I like that ScyllaDB is more flexible for me compared to MongoDB, and the terminal structure and commands are better to use. I also find the teaching at the summit very good because it helps me see my mistakes and learn from them. Using the cqlsh terminal to interact with the cluster efficiently through CQL queries is great, and ScyllaDB's use of shard per core architecture maximizes performance and reduces latency in a distributed database. Review collected by and hosted on G2.com.

I think improving the setup guide and helping new users understand the distributed database concept. A clear dashboard-like insight to help developers in observability tools and expanding integration with cloud native ecosystems can make it more powerful for developers. Review collected by and hosted on G2.com.

I use ScyllaDB as a database store for internal management services like ticketing management. One thing I like most about ScyllaDB is the low latency. I had tried Redis; it was low but expensive, so I had to switch. Then I saw the Last Monster Summit and decided to try ScyllaDB, and till now, I have no regrets. Users see updates on the leaderboard instantly instead of waiting 400ms with the database. This low latency was the main driver for switching from Postgres and Redis to ScyllaDB. Initially, the setup was good. Review collected by and hosted on G2.com.

One thing I felt that could be improved is a bit on documentation and some examples for already setup systems would be appreciated. Review collected by and hosted on G2.com.

It is a reliable and cost-effective NoSQL solution that effectively eliminates performance bottlenecks for today’s heavy-duty applications. Review collected by and hosted on G2.com.

Cost of DB for running my applications compared to a self hosted solution. Review collected by and hosted on G2.com.

ScyllaDB is open source and can be deployed independently without any license. Its cluster can also be managed with Kubernetes. Review collected by and hosted on G2.com.

ScyllaDB doesn’t offer a self-managed service on any of the major cloud providers. Review collected by and hosted on G2.com.

ScyllaDB gives very high performance for our ML pipeline. It uses shard-per-core design, so each CPU core handles its own data. Because of this, read and write speed is very high. It supports high throughput and keeps latency very low even when we store and fetch large training data. Review collected by and hosted on G2.com.

Data modeling needs to be done carefully. We must design partition keys properly to get best speed. Monitoring and tuning need some effort when data grows very large. Review collected by and hosted on G2.com.

I like ScyllaDB for its shard per core design because it's NUMA friendly and eliminates contentions and locks in network event streaming. It provides linear scalability per core, which I find great since there are no GC pauses, helping to eliminate tail latency spikes like we experienced with Cassandra. I also appreciate the massive throughput, which helps with high volume ingestion. Review collected by and hosted on G2.com.

I find ScyllaDB's data modeling challenging and the learning curve quite strict. The strict data modeling makes schema changes later difficult, especially if the event starts to have additional data. Review collected by and hosted on G2.com.