July 2021 Summaries

P99 CONF is a virtual event organized for engineers by engineers, focusing on high-performance architectures and innovative solutions to latency challenges in modern applications. Scheduled for October 6-7, the event features over 30 speakers from various leading technology companies such as Netflix, Twitter, and ScyllaDB, all of whom will share insights on technical topics ranging from kernel optimizations to emerging programming languages like Rust and Go. The conference eschews product pitches, instead offering practical, technical sessions on operating systems, distributed systems, performance testing, and DevOps strategies. Notable speakers include Glauber Costa from Datadog, who will discuss designing for optimal performance in Rust, and others like Waldek Kozaczuk and Tejas Chopra, who will cover topics such as running stateless apps in the cloud and object compaction, respectively. P99 CONF promises a comprehensive lineup of discussions aimed at developers looking to delve into advanced techniques and real-time application solutions, with a commitment to fostering a productive and inclusive environment for all participants.

ScyllaDB Cloud's Bring Your Own Account (BYOA) feature offers a unique managed NoSQL Database-as-a-Service (DBaaS) that operates directly within a user's AWS account, enabling organizations to benefit from pre-negotiated AWS discounts and a consolidated billing process without additional markups. This setup provides a high-performance, cost-effective, and scalable solution that is particularly appealing to companies for its low latency and ease of use, as praised by users like Disney+ Hotstar and GumGum. The BYOA approach also enhances security and regulatory compliance by allowing users to maintain data governance and control over geographic data storage. ScyllaDB Cloud simplifies deployment with a user-friendly setup wizard that integrates ScyllaDB instances seamlessly into existing AWS infrastructure, reducing operational overhead while ensuring data privacy and compliance with regulations such as GDPR and the California Consumer Privacy Act.

ScyllaDB University LIVE Summer School is a free, half-day, instructor-led event designed to provide training on ScyllaDB, featuring sessions from leading engineers and architects. Scheduled for July 28th and 29th in different time zones to accommodate a global audience, the event will cover both foundational and advanced topics, including ScyllaDB architecture, data modeling, integration with Apache Kafka, and collaboration with Apache Spark, as well as building well-architected applications on ScyllaDB Cloud. Attendees will also participate in hands-on labs, roundtable discussions with experts, and networking opportunities, while having access to additional self-paced learning materials and the chance to earn a certification and receive swag. The event aims to enhance participants' understanding of ScyllaDB and improve their application development skills.

The ScyllaDB Rust Driver, initially developed during an internal hackathon, has been officially released as version 0.2.0 on crates.io and has received contributions from the open-source community. It now includes essential features such as authentication, TLS support, configurable load balancing, retry policies, speculative execution, tracing support, and internal logging, making it a robust tool for managing database connections. Benchmarks comparing the ScyllaDB Rust Driver with other drivers, including cpp-driver and cassandra-cpp, show it to be highly competitive in performance and scalability, often outperforming the multithreaded C++ driver. Tests were conducted on a 3-node ScyllaDB cluster, with various configurations to evaluate the driver's ability to handle different workloads and data sizes. The development of the driver was also part of an academic project with the University of Warsaw, involving students in adding features and conducting benchmarks, with their findings published in a thesis.

The blog post explores the complexities and considerations involved in configuring Amazon DynamoDB tables for cost efficiency and operational simplicity, particularly focusing on the choice between on-demand and provisioned capacity modes. It discusses the various factors influencing this decision, such as load patterns, historical data analysis, and the benefits and limitations of autoscaling. The piece argues that autoscaling, while useful in certain scenarios with high amplitude and narrow load peaks, is not a substitute for proper capacity planning and can lead to higher costs if not managed carefully. It highlights that while autoscaling offers flexibility in handling unexpected loads, it may not always deliver significant cost savings compared to upfront reservations and static provisioning, particularly in the context of large, predictable workloads. The post concludes by encouraging readers to engage in thoughtful capacity planning and utilize tools like the ScyllaDB Cloud Calculator for better resource management and cost comparison across different cloud database services.

In June 2021, ScyllaDB made significant progress across its projects, highlighted by the upcoming release of ScyllaDB Open Source 4.5, which introduces features like the ability to load and stream SSTables for easier data migration and restoration. The update also enhances the DynamoDB-compatible API within Project Alternator and implements foundational Raft changes to support future capabilities. Additionally, ScyllaDB optimized its Change Data Capture (CDC) feature, especially for large clusters, and integrated it into ScyllaDB Cloud. Other updates include ScyllaDB Enterprise 2021.1.2, which offers LDAP support and a new GCP image, and improvements in the ScyllaDB Monitoring Stack, Manager, and Operator. The company emphasizes its strong release cadence, which allows for rapid feature addition and bug fixing, and compares its development pace with other open-source projects like Apache Spark and Kafka. ScyllaDB also encourages participation in its University LIVE Summer Session, offering insights into ScyllaDB operations and its integration with big data architectures such as Apache Spark and Kafka.