May 2024 Summaries

External caches are often used to enhance database performance, but they come with significant drawbacks such as added latency, increased costs, decreased availability, and heightened application complexity. Despite the initial appeal of external caching for improving latency, it often results in additional complications, including the need for complex infrastructure and security measures. Three teams—SecurityScorecard, IMVU, and Comcast—transitioned to ScyllaDB, a high-performance database with an internal caching mechanism, achieving lower latency, reduced costs, and simplified infrastructure. This approach allowed them to meet their service-level agreements more effectively than with traditional external caching methods, highlighting the benefits of adopting a modern database that integrates caching internally rather than relying on external solutions.

Organizations looking to transition from AWS DynamoDB to ScyllaDB are driven by the need to reduce costs, avoid latency issues, and seek more flexibility away from cloud vendor lock-in. ScyllaDB offers two main protocols, CQL and ScyllaDB Alternator, with the latter designed to be almost a drop-in replacement for DynamoDB, albeit with some differences and the necessity for minor adjustments during migration. While ScyllaDB is not a native AWS service and may lack some newly launched DynamoDB features, it provides significant advantages like predictable low-latency performance and less infrastructure requirement. Migration typically involves data transfer from DynamoDB and application updates to connect with ScyllaDB, with tools like the ScyllaDB Migrator facilitating the process. Furthermore, ScyllaDB's caching strategy is integrated and efficient, matching the speed of DynamoDB Accelerator (DAX) but with lower latencies and simplified management. Despite some missing features, ScyllaDB's open-source nature and production readiness since 2020 make it a viable alternative for organizations seeking to break free from AWS's ecosystem.

Piotr Dulikowski discusses the unexpected performance issues encountered while implementing per-partition query rate limiting in ScyllaDB, specifically focusing on C++ exception handling. Despite initial attempts to manage overloads by rejecting excess queries to maintain stable goodput, the handling of exceptions, particularly in large volumes, became a bottleneck due to the CPU-intensive nature of rejected operations. This issue was compounded by the use of Seastar, a C++ framework, which, while adept at managing asynchronous tasks, struggled with exception handling during high-load scenarios. To address these issues, the team explored two approaches: using boost::result to return errors instead of throwing exceptions, and implementing a custom utility for inspecting std::exception_ptr without rethrowing, thereby reducing the overhead of exception handling. Benchmarks comparing ScyllaDB versions 5.0.3 and 5.1.0-rc1 demonstrated that these optimizations significantly improved throughput and goodput, allowing the updated version to handle higher write rates with fewer timeouts, thus enhancing overall performance under stress.

ScyllaDB University is an online learning platform providing self-paced courses and interactive live events to help users master ScyllaDB, a high-performance NoSQL database. The platform has updated its offerings, including a popular High Availability lab that simulates node failures to teach data management strategies using varying consistency levels. ScyllaDB’s design includes a peer-to-peer architecture and uses the Gossip protocol for resilience and scalability, with features like multi-datacenter replication ensuring high availability. Upcoming live events will be held in different time zones, offering opportunities to interact with ScyllaDB experts and peers. The platform's courses adapt as the product evolves, ensuring learners access the latest information and practices.

ScyllaDB, a remote-first company, held its annual Internal Engineering Summit in Limassol, Cyprus, bringing together over 120 employees from diverse roles to foster a sense of unity and shared purpose. The event provided an opportunity for team members, including software engineers, product managers, and technical support staff, to engage in technical sessions, share knowledge, and discuss the company's roadmap, security features, and customer experience improvements. Beyond the professional exchanges, the summit included team-building activities, social events, and an off-road jeep safari, allowing participants to enjoy Cyprus' cultural and natural attractions. The gathering emphasized the company's commitment to collaboration and innovation, with the CEO praising the high level of teamwork. ScyllaDB continues to seek new talent for various remote positions across engineering and other departments.