I like how ScyllaDB makes large-scale distributed data feel simple and allows me to work with its design patterns effortlessly. I really appreciate the automatic distribution and the hot reload performance of 1 million operations per second. One thing I especially like about ScyllaDB is its high performance and predictable low latency, even under heavy workloads. I also value its compatibility with the Cassandra ecosystem, which makes migration and integration straightforward. Features like shard-aware drivers, horizontal scalability, and efficient resource utilization work particularly well for large-scale, high-throughput applications like ours. Review collected by and hosted on G2.com.

Overall, my experience with ScyllaDB has been positive, but there are a few areas that could be improved. Initial cluster setup and tuning can be complex, especially for production deployments. Troubleshooting shard-awareness and networking issues also required a good understanding of the architecture. I also think the documentation could include more end-to-end examples and performance tuning guides for large-scale deployments. I would like to see more built-in observability and performance diagnostics. Additionally, although I use ScyllaDB with GUI tools like DataGrip and DBeaver for querying data, browsing tables, and schema management, I feel these tools are often paid and limit users for a better experience interacting with ScyllaDB. Review collected by and hosted on G2.com.

ScyllaDB provides exceptional throughput and sub-millisecond, predictable latency. Because it is written in C++ rather than Java, we completely eliminated garbage collection pause issues and saw a massive performance boost over Cassandra with far fewer nodes. Review collected by and hosted on G2.com.

ScyllaDB delivers high performance, but its ecosystem of third-party monitoring plugins and integrations is still a bit smaller than what you’ll find with older databases such as Apache Cassandra. Review collected by and hosted on G2.com.

The performance is really impressive - we're seeing much lower latency and way better throughput compared to what we had before. The fact that it's compatible with Cassandra made our migration pretty smooth, and honestly, the speed improvements alone made it worth the switch. Review collected by and hosted on G2.com.

It took us a good bit of time to get comfortable with it, especially coming from a simpler database setup. The docs are okay but sometimes feel a bit sparse - we spent way too much time googling things during our initial deployment. Would love to see more practical examples and troubleshooting tips for when things don't go as planned. Review collected by and hosted on G2.com.

I like ScyllaDB's performance, especially because it's impacted by its indexing and the whole database being written in C++, which serves as the building block for a faster NoSQL DB. It's great for storing our ERP system, especially as we introduce new features and are unsure about the schema size. This NoSQL wide column structure allows us to store a wide range of data. Performance is great compared to MongoDB, and it has very low latency. We experienced a decrease in latency while debugging through the log. Review collected by and hosted on G2.com.

I find ScyllaDB's UI to be very minimal. It doesn't have native integration for cloud functions like MongoDB or Firebase. Also, live streaming requires extra setup compared to Firebase. Review collected by and hosted on G2.com.

I like ScyllaDB for its impressive performance and speed, offering 10 times more performance compared to Cassandra. It's rewritten in C++ instead of Java, which provides predictable low latency, like p99 latencies under heavy load, and overcomes JVM limitations. I also appreciate its ability to tune settings, reducing administrative overheads and making it cost-efficient as well. Review collected by and hosted on G2.com.

It could be the steep learning curve. You know, optimizing and, managing the ScyllaDB at scale, it requires deep knowledge of distributed systems. Especially around the horizontal scaling using the shard key selections and data modeling. Sometimes it warrants a complex data modeling, because this is crucial. To avoid those expensive mistakes at a later stage. The third one would be a, smaller community. So even though it is growing, but there is still a smaller community, compared to the Postgres and other databases available. Review collected by and hosted on G2.com.

In Blockchain, we deal with a massive amount of transaction data. ScyllaDB handles high-speed writes and reads without slow down for storing our transactional data fron ethereum(other Blockchains) chain to our local storage just to make it more acessible and faster. Review collected by and hosted on G2.com.

To get the absolute best results, you need to understand how to model your data correctly (partition keys and clustering keys). Review collected by and hosted on G2.com.

I'm extremely happy about ScyllaDB's low latency and seamless scalability, especially when compared to the delayed processes in traditional databases. Its ability to handle high write throughput with less latency really stood out. We also realized somewhat better analytics when compared to traditional options, and I experienced more reliable monitoring with it. Review collected by and hosted on G2.com.

Cluster setup took us some time, so that's one area I would want it to simplify. Like a one-click setup for POC purposes with defaults. Also, auto-recommendations like optimizing, cost savings, etc., could be on par with others. The initial setup was a bit complex considering my novice experience with ScyllaDB. Review collected by and hosted on G2.com.

I like ScyllaDB for its significant cost savings and compatibility with Cassandra. The operational simplicity and robust ecosystem make it efficient to use. I value the high performance on fewer nodes, which reduces infrastructure costs due to CPU optimization and the ability to work with smaller clusters. The auto-tuning and self-optimization features help cut down on manual DBA overhead. I also appreciate the strong community and OEM support, which reduce downtime risks and facilitate adoption across production environments. Additionally, I found the initial setup to be very easy, especially with managed services, and the process is familiar for Cassandra users but simplified for new users, particularly with automated tuning scripts. Review collected by and hosted on G2.com.

The learning curve around the shard-per-core architecture and tuning is steep for new team members who are not familiar with it. Monitoring integration and developer SDK still lag behind other databases like Cassandra. Although ScyllaDB's native monitoring stack with Prometheus and Grafana works well, it requires manual setup and configuration compared to managed DB services like Azure and AWS, and using popular observability tools like New Relic or DataDog with pre-built alerting templates for common failures would be much better than having to write numerous queries in Prometheus. Review collected by and hosted on G2.com.

I love how the ScyllaDB community is rising right now. I also love how they have a lower cost, especially as a startup or small organization where we are very cost stringent and expect a lot more features from database systems. The initial setup of ScyllaDB was very easy for me. The documentation is great to refer to, and I just created a cluster and deployed it on ScyllaDB Cloud easily. Review collected by and hosted on G2.com.

Right now, it's a simple database implementation I have done, and, we still have to migrate. We have the backup data from Supabase, and, we have to implement it yet on ScyllaDB. So I think I would need time to give a better answer for that. Review collected by and hosted on G2.com.

I like ScyllaDB's performance-oriented architecture and its compatibility with existing Cassandra tooling, which makes it efficient and scalable. Its focus on efficiency and scalability stands out as key strengths for me. I appreciate how it allows for familiar APIs and workflows to be reused without major changes during evaluations. It's great for evaluating system disparities, ensuring low latency and predictable behavior. Its efficient use of underlying hardware and ease of integration into distributed data architectures needing to grow while maintaining consistent performance makes it valuable. Review collected by and hosted on G2.com.

One idea could be, the initial learning curve, especially for people who are new to the distributed database or performance tuning concepts. Some configurations could use some work on that. And establish clearer guidance. Review collected by and hosted on G2.com.