January 2021 Summaries

YugabyteDB is an open-source distributed relational database designed to handle large-scale data in various environments. It offers strong consistency, high availability, and scalability, making it suitable for mission-critical applications. Unlike traditional monolithic databases, YugabyteDB's distributed architecture allows for horizontal scaling, which can lead to single-digit ms latency. The system provides automatic load balancing for YCQL, a NoSQL API, but recommends external load balancers for YSQL, a fully relational API. Sizing the cluster requires considering read/write transactions per second and running workload generators to determine optimal cluster size. YugabyteDB supports multiple APIs, including YSQL and YCQL, which cater to both relational and NoSQL workloads. The system also offers row-level geo-partitioning for managing data residency across regions.

Yugabyte University offers a variety of instructor-led and on-demand courses that enable learners to acquire distributed SQL fundamentals, development, and administration skills in various formats. The program features virtual 90-minute instructor-led courses delivered by Yugabyte community members, free on-demand video courses, and technical workshops. Corporate training programs are also available for teams, with in-depth multi-hour instructor-led courses delivered over Zoom. A roadmap of upcoming courses and certifications is planned for 2021, including a mix of lectures and hands-on labs to suit different learning styles. The #training channel on YugabyteDB Community Slack provides updates on new trainings and certifications as they become available.

GraphQL provides a query language for APIs, giving UX developers autonomy over querying APIs and the database. The advantages of using GraphQL include retrieving only needed data, strong type system, increased velocity, and minimized feedback cycles. For cloud native API developers, factors like performance, handling infrastructure outages, and scalability are crucial. In enterprise development, scaling and resilience needs must be addressed before sending applications to production. To explore scalable and resilient GraphQL apps with distributed SQL databases, a benchmark was conducted using Hasura GraphQL Engine connecting to YugabyteDB. The results showed that the subscription performance tool tested linear scalability of GraphQL subscriptions without downtime. The setup simulated a real-world e-commerce user order management system, scaling resources from 25K subscribers to 1 million subscribers linearly. High availability was also measured during a database pod failure scenario, observing minimal impact due to the stateless nature of Hasura GraphQL Engine and YugabyteDB's stateful service. A sample application was created to benchmark GraphQL subscriptions, demonstrating the importance of addressing scale and resiliency needs in building scalable and resilient GraphQL apps.

At the Distributed SQL Summit 2020, DaShaun Carter presented a talk on "Cloud Native Spring for Relational Databases", exploring a path to distributed SQL and considering a use case involving a Java app, RESTful services, and single node PostgreSQL. The presentation discussed the cost per transaction and three stakeholder groups: DBAs, app devs, and leadership, highlighting that scaling writes is the tough part of scaling reads. DaShaun recommended using Spring Data R2DBC to solve at-scale problems, considering cost per transaction when making decisions, working together across teams to drive ROI, measuring cost per transaction, and starting with sample code using Spring Data R2DBC for a quick move to a more scalable solution. The presentation concluded by stating that Spring Data and YugabyteDB distributed SQL can quickly address scaling issues without changing much code, with data-driven decisions including cost per transaction.

The Distributed SQL Summit 2020 panel discussion focused on the future of data infrastructure in the telco industry, with experts from Rakuten Mobile, TracFone Wireless, Comcast, and others sharing their views on trends such as IoT, 5G, Edge, cloud, and their impact on data infrastructure. The panelists emphasized the importance of innovation, agility, and resilience in the face of changing technology trends. They discussed various topics including cloud native architecture, Kubernetes, stateful databases, and distributed SQL, highlighting its potential to solve scaling challenges in modern applications. The experts also shared their experiences with deploying cloud-native architectures, automating DevSecOps processes, and addressing data management challenges in a rapidly evolving landscape. Overall, the discussion aimed to provide insights into the future of data infrastructure and how telcos can stay ahead of the curve by adopting innovative technologies and strategies.

The Distributed SQL Summit Asia is a virtual event taking place on January 20-22, featuring an impressive lineup of talks by industry-leading speakers from innovative companies. The event aims to provide professional development opportunities for attendees looking to expand their knowledge of building modern, cloud-native, distributed systems. The schedule includes a Workshop Day on January 20 and Conference Days on January 21-22, covering topics such as achieving internet-scale, operational ease, high performance, and complementary technologies. Attendees can customize their agenda to participate in sessions they find most interesting and actionable.

A growing number of global enterprises are adopting a multi-cloud strategy to support their data infrastructure, driven by the need for reduced operating costs, improved application resilience, enhanced customer experience, and compliance with data regulations. Companies without a multi-cloud strategy risk falling behind competitively, while CTOs face challenges such as optimizing cloud spend, enforcing security across different providers, and migrating workloads between clouds. To design multi-cloud solutions that outweigh the costs, enterprises should choose the right deployment option, select a cross-region data distribution strategy, redefine their disaster recovery (DR) strategy to cater to cloud outages, prioritize security and compliance policies, make full use of cloud elasticity, enable cloud migration and data portability, centralize database tooling and governance, and invest in DevOps and CI/CD tooling. A cloud-native database solution like YugabyteDB can help organizations overcome these challenges and achieve a successful multi-cloud data infrastructure strategy.