August 2021 Summaries

Follower reads are a key feature for developers using multi-region databases. They allow read-only queries to be served by any replica, ensuring low-latency access to all data in a database regardless of data homing. CockroachDB initially introduced exact staleness reads, which select how far in the past to read and then send the request to the nearest copy of the requested data. However, this approach can cause issues when the guess is wrong, leading to additional latency or redirection to the leaseholder. To address these issues, CockroachDB is exploring bounded staleness reads, which dynamically determine the optimal timestamp to read at, minimizing staleness and ensuring local query serving without blocking. This new approach provides a more predictable solution in the presence of replication slowdown but comes with a small performance cost and can only be used in single-statement transactions.

The text provides a comprehensive hands-on tutorial for beginners to learn SQL by simulating the development of a database for an animal shelter named "Animals 4 Homes." It covers fundamental SQL concepts and commands, such as creating databases and tables, inserting data, updating records, and retrieving information through SELECT queries. The tutorial introduces data manipulation techniques using SQL statements like INSERT, UPDATE, and DELETE, and guides users through exercises to apply these commands practically. It emphasizes the importance of defining a database schema, managing relationships between tables, and optimizing queries with indexes and JOIN operations. Through the fictional scenario of managing a shelter's animals and adoption records, readers gain practical experience in setting up and interacting with relational databases, understanding SQL's declarative nature, and its powerful application in data management. The guide concludes by encouraging further practice using free online SQL databases and offers additional resources for continued learning.

Cockroach Labs has signed a lease for its first dedicated Toronto office at 33 Yonge Street. The company, which focuses on building the best distributed SQL database, already has a significant presence in Toronto and plans to grow the team significantly over the next year. The new office space is centrally located near Union Station and features large windows, natural light, and ample opportunities for collaboration. Cockroach Labs is currently hiring in Toronto and elsewhere, with open roles available on their website.

The Kubernetes Adoption Trends Survey revealed that companies either have many active workloads running on Kubernetes or just a few, with limited middle ground. The number one challenge running Kubernetes is the deployment of data-intensive transactional workloads, as reported by 46% of surveyed organizations. Serverless adoption is also emerging, with two-thirds of companies already using serverless computing or in the process of doing so. Additionally, multi-cloud environments are becoming more common, with 65% of survey participants working in a multi-cloud or hybrid environment or actively moving workloads to such an environment by the end of 2021.

The history of high availability databases can be divided into three categories: Active-passive, Active-Active, and Multi-Active. Active-passive databases have an active node that processes requests with a hot spare ready to go in case of disaster. Active-Active databases have at least two active nodes that shard data and perform writes to the database. Multi-Active databases require at least three active nodes, each capable of performing reads and writes for any data in the cluster without generating conflicts. The development of these systems has been driven by the need for increased availability, scalability, and consistency in modern distributed computing environments.

This blog post delves into the intricacies of time data types in PostgreSQL, including TIME, TIMETZ, TIMESTAMP, and TIMESTAMPTZ. It covers topics such as session time zones, POSIX time offsets, parsing, casting, precision, functions and operators, AT TIME ZONE, daylight savings, interval math with daylight savings, the Y2K38 problem, and more. The author recommends using TIMESTAMPTZ with a session time zone set to UTC for handling time data types in PostgreSQL.

This article discusses how to update a database schema with no downtime. It highlights the challenges faced when updating traditional relational databases (RDBMS) and NoSQL document-store alternatives, such as MySQL and MongoDB respectively. The author introduces CockroachDB, a next-gen distributed SQL database that offers live schema updates without any downtime or impact on users. It also provides an example of altering the database schema for a Node.js app using CockroachDB's SQL syntax.

A database schema is a blueprint that defines how data in a database is organized. It outlines the structure of the data and establishes relationships between different elements within the database. The schema is crucial for maintaining consistency, facilitating efficient querying, and ensuring data integrity. In SQL databases, schemas can also refer to named groups of database objects or visual representations of a database's structure. Designing an effective schema requires careful consideration of factors such as overall structure, naming conventions, constraints, data types, indexes, security, geo-partitioning, and documentation.