ScyllaDB Reviews & Product Details

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.

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.

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.

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.

It handles big read and write loads without slowing down. The data is divided into shards, so each node takes care of its part. This makes the system fast and stable. The data moves easily between nodes, and the jobs finish faster. Review collected by and hosted on G2.com.

There is nothing much to dislike. But it would be nice if ScyllaDB can have full user management and authentication built in. That way, we would not need to use other tools for access control. Review collected by and hosted on G2.com.

Below are things which I like the most about our ScyllaDB usecase:

- Works really well for our online shopping site

- Able Handle lots of customers at once without slowing down the performance

- Easy to add more servers when we get busy during sales time

- Product catalog loads super fast for customers

- Shopping cart data never gets lost

- Works well with our existing tools

- Customer orders process quickly even during peak hours Review collected by and hosted on G2.com.

- Setting it up the first time was tricky

- Not many online tutorials for beginners

- Some features we used in our old database aren't available Review collected by and hosted on G2.com.

I use ScyllaDB as a NoSQL database for appending message data before transitioning it to downstream services, where its efficiency stands out significantly. ScyllaDB is cheaper than DynamoDB, yet it offers comparable or sometimes even superior performance, which is highly beneficial for cost-effective scalability. I appreciate the SQL-like syntax that ScyllaDB provides, which enhances the usability by making it easier to query and manage data. The shard-per-core architecture is particularly impressive because it allows us to maximize CPU utilization, squeezing out extraordinary levels of performance, which is critical for our intensive processing needs. Additionally, the support team at ScyllaDB has been incredibly active and helpful in troubleshooting, making the transition smooth and instilling confidence in the adoption of their system. Review collected by and hosted on G2.com.

I find the lack of support for lightweight transactions (LWT) across nodes or partitions challenging. This limitation makes it difficult to handle transactions in a manner similar to PostgreSQL, where transactions can touch multiple tables in the database. Additionally, I would appreciate built-in support for retries, which is currently lacking and impacts the ease of handling transactions effectively. Review collected by and hosted on G2.com.

I use ScyllaDB as the high-performance backend database for my real-time voice chat application, and I am thoroughly impressed by its ability to efficiently manage stateful information with remarkable speed and reliability. ScyllaDB effectively handles a huge volume of reads and writes effortlessly, removing the constant worry of the database becoming a bottleneck. Its reliability allows me to focus on building my application's features instead of constantly tweaking performance. Additionally, its setup was surprisingly straightforward due to Docker, as I simply pulled the official ScyllaDB image and added it to my docker-compose.yml file, which significantly reduced the complexity of installation. Its seamless integration with my Go application via the gocql driver, along with its operation inside Docker as part of my main development stack, underscores its adaptability and ease of use in diverse development environments. Review collected by and hosted on G2.com.

I find the learning curve for ScyllaDB quite steep, especially for newcomers to distributed NoSQL databases. Transitioning from a SQL background requires a significant shift in data modeling approach, and understanding concepts like consistency levels, replication, and query patterns isn't straightforward. This complexity can be daunting and requires a dedicated effort to grasp fully. Review collected by and hosted on G2.com.

I appreciate ScyllaDB for its high performance and low latency in handling NoSQL workloads, making it exceptionally useful for real-time applications such as IoT, finance, and technology, particularly when analyzing time-series data. ScyllaDB stands out as a clear replacement for Apache Cassandra, offering support for the DynamoDB API of AWS, which enhances its applicability and flexibility. I love its thread-per-core architecture, which ensures maximum CPU efficiency by eliminating context switching overhead, resulting in ultra-low latency even under heavy loads. This feature, combined with its predictable performance and automated workload isolation, delivers consistent throughput without stability issues, making ScyllaDB ideal for real-time and high concurrency systems. The database excels in key performance metrics such as scalability and cost efficiency, providing 20x better throughput and more predictable performance under heavy load compared to other NoSQL databases like Cassandra, MongoDB, and DynamoDB. I also value the cloud model ScyllaDB offers, along with its strong support for time series data. Review collected by and hosted on G2.com.

I find the documentation depth lacking, which makes it difficult to fully grasp all aspects of the software. The setup process could be more user-friendly, possibly through an agent-based setup. There is inadequate support for complex use cases such as migration and cloud integrations like AWS. Additionally, the tools for data export or import need improvement. I faced hardware challenges initially, which were a significant hurdle during the setup of ScyllaDB. Review collected by and hosted on G2.com.

ScyllaDB is fast, highly scalable, and efficiently uses all CPU cores. It handles millions of operations per second with consistent performance, scales easily by adding nodes, and has excellent, practical documentation. Review collected by and hosted on G2.com.

It’s self-managed and requires a fairly deep level of technical knowledge to operate effectively. Review collected by and hosted on G2.com.