September 2022 Summaries

SingleStoreDB's new Code Engine — Powered by WebAssembly (Wasm) enables users to utilize user-defined functions and table-valued functions written in C, C++, Rust, and more languages, with plans for additional supported languages. This feature allows developers to reuse native code in a sandboxed environment at blazing speeds. To get started, users need to set up their local environment using v.s. code, the WASI SDK, and the Rust toolchain. They then create a Wasm file and upload it to an AWS S3 bucket, where they load and call the Rust user-defined function on a Singlestore Helios workspace. The process involves installing the necessary tools, creating a new folder, initializing the Rust source tree, building the .wasm file, uploading the files to object storage, setting up the Singlestore Helios Workspace, and loading and running the Wasm function in Workspaces. Additionally, users can create table-valued functions using the same process. Overall, this feature empowers developers to execute functions directly on SingleStore's distributed system at runtime with near-native performance in a secure environment.

The SingleStore Hackathon has opened, offering $30K in cash prizes and other incentives for developers to contribute their software applications using the SingleStoreDB database. To get started, developers need to register on the hackathon website, create a free Devpost account, and sign up for a free SingleStore account with $600 of credits. They can then choose one of five categories - multi-model, HTAP, Wasm/WebAssembly, Machine Learning/AI, or Database Migration - and build an application that fits their chosen category. The submission requirements include linking to the application code on GitHub, creating a video demonstrating the application's features, and providing contact information. Developers can access resources and documentation for each category to help them get started with SingleStoreDB.

SingleStore is announcing results from its Data Intensity Calculator, an online tool used by 125 companies. The calculator measures data intensity based on five considerations: concurrency, data size, query complexity, data ingest speed, and query latency. Applications assessed showed significant complexity and concurrency, growing data requirements, with nearly 65% of companies expecting their data to grow between 10% and 100% within the next year. Data size was another category in which applications registered high demands, with over a third of tested applications employing more than 50 TB of data. The calculator also found that businesses must support their applications' data ingest speeds and latency control needs, with most applications requiring fast ingest rates and minimal latency. SingleStoreDB is positioned as the database solution that can handle growing data intensity and user expectations, delivering better end-user experiences and enterprise intelligence.

The SingleStoreDB Certified Developer Exam is designed to test a candidate's expertise in developing applications on the world's fastest distributed SQL database, known for its millisecond response times and three-tiered storage architecture. The exam covers four key domains: architecture, design, ingest, and develop, with questions developed by SingleStoreDB's own team of engineers. By passing the exam, candidates can gain an advantage in the job market, recognition, and opportunities for advancement in a crowded database and developer market.

A developer's life is not as comedic as it seems when dealing with single-node open-source databases that struggle with performance issues such as data ingestion, scaling, and storing various types of data. Developers like Jack Ellis from Fathom, Josh Blackburn from IEX Cloud, Gerry Morgan from dailyVest, and Mohammed Radwan from Foodics have experienced challenges with popular open-source databases like MySQL, PostgreSQL, and MariaDB. These challenges include application performance hitting a wall, database sprawl due to the need for specialty databases, and reliability problems that can lead to outages and limited growth. To overcome these issues, developers look for a replacement database that offers streaming data ingestion, low-latency query performance, limitless scalability, robust analytical capabilities, and hybrid workload handling. SingleStoreDB, a real-time, distributed SQL database, is a popular choice among developers due to its MySQL wire-compatibility, ease of migration, and ability to provide significant improvements in speed, performance, and scalability. By migrating to SingleStoreDB, these developers have been able to overcome their challenges and achieve major improvements in their applications and business growth.

SingleStoreDB is a unified, distributed, relational multi-model cloud database designed for real-time applications and analytics, providing low-latency responses, scalability, and simplicity. It powers fast, real-time analytics and applications by addressing performance, scale, and concurrency challenges, making it suitable for various industries such as cybersecurity, IoT, fintech, eCommerce, and more. With its capabilities to access real-time data, deliver fast interactive customer experiences, scale effortlessly, and run anywhere, anytime, SingleStoreDB solves the complexities of modern, real-time applications by providing a solution to the challenges faced by businesses in constantly adding new technologies to accommodate required functionalities. By delivering low-latency responses, scaling 10-100x performance improvements at ⅓ the cost of legacy databases, and boasting 60% better price-performance and TCO over popular database combinations, SingleStoreDB simplifies application architecture with familiar SQL tooling and multi-model capabilities.

Modern tech stacks require developers to manage multiple components of their application, including frontend UI, orchestration, APIs, and databases. As data intensity increases, these components become more complex, but solving the database problem is often a long-term solution for scaling. SingleStoreDB Universal Storage requires careful consideration of shard keys and sort keys to optimize query performance. Choosing the wrong shard key can lead to unbalanced partitions, while missing a sort key can leave tables unorganized and hard to scan. Additionally, mismatched data types across tables can negatively impact query results and performance. By understanding these concepts and avoiding common mistakes, developers can ensure their SingleStoreDB applications perform optimally.

The ACM SIGMOD conference has seen a renewed interest in Hybrid Transactional Analytical Processing (HTAP) this year, with four HTAP papers presented at the conference and an HTAP workshop. SingleStoreDB, one of the widely deployed distributed HTAP SQL databases, was featured in a paper titled "Cloud-Native Transactions and Analytics in SingleStore" which describes its key design decisions enabling it to run a breadth of workloads. HTAP databases like SingleStoreDB are reversing the trend of specialized database systems designed for single use cases, offering benefits such as reduced training requirements for developers, lower data movement and transformation costs, and improved storage efficiency. The paper also highlights SingleStoreDB's unified table storage and separation of storage and compute design, which enable fast scan performance and high availability. A CH-benCHmark benchmark test showed that SingleStoreDB can run both transactional and analytical workloads with strong performance, sharing resources almost equally without impacting each other. This demonstrates the potential of HTAP databases like SingleStoreDB to handle a wide range of applications and use cases.

Reducing cost for historical/unused/archived data is a challenging task due to massive storage requirements, and SingleStoreDB Self-Managed offers an effective solution by providing unlimited storage. By converting storage into a low-cost Glacier storage class, organizations can save up to ~85% on expenses. To achieve this, create a database archiving policy to move historical data from local databases to unlimited storage databases, and then set up a lifecycle policy to transition objects between storage classes. Once the data is moved, detach the database and create a milestone, allowing for easy restoration of the database when needed. The process involves creating an unlimited storage database on AWS S3, setting up archiving policies, configuring lifecycle rules, and attaching the database after restoration in standard storage.

We implemented a dark theme for our Singlestore Helios Portal UI, allowing users to select between light, dark, or system (operating system) themes. The implementation involved providing logic and UI for getting and setting the applied theme in-app and in Storybook, extending design tokens and migrating to CSS Custom Properties, conditionally switching between light and dark graphics, testing and quality assurance, revising our design tokens and component library, discovering edge-cases and collaborating with designers to decide on a solution. We also implemented typed styling props using Class Variance Authority, accessing token values in TypeScript files, generating code derived from design tokens using Style Dictionary, and providing utility classes for applying common styles. The final result was shipped to customers after extensive testing, demonstrating the benefits of implementing a dark theme with CSS Custom Properties.

This article explores the benefits and drawbacks of using multi-model databases. A multi-model database is a type of database that allows you to store and access different types of data, such as relational, time series, geospatial, key-value, and document data, in a single database. This approach can simplify application development, reduce the need for ETL (Extract, Transform, Load), and improve scalability and performance. However, it also requires careful consideration of consistency models, ACID compliance, cost efficiency, ease of handling many data formats, and reduced need for ETL. The article highlights several use cases where multi-model databases can provide significant benefits, including interactive SaaS applications, IoT telemetry and analytics, streaming media analytics and ad optimization, gaming leaderboards, and more. Overall, using a multi-model database like SingleStoreDB can help reduce the time to market for products, improve scalability and performance, and simplify application development.