May 2022 Summaries

The Fly.io team has released an API called Fly Machines that enables users to start and stop fast-booting Firecracker VMs directly, bypassing orchestration. This feature is designed for developers who want to quickly spin up applications in a matter of minutes. Additionally, the web UI has undergone significant updates, including new pages, improved information displays, and enhanced organization features. The changelog includes various improvements such as better error handling, more intuitive messages, and additional context for certain errors. Furthermore, Fly.io has released new API Telemetry to track meaningful operations and diagnose performance issues.

Fly.io has introduced Fly Machines, which are VMs with a fast REST API that can boot instances in about 300ms. The platform allows developers to manage individual VMs and create cost-effective applications that scale to zero when not in use. Fly Machines are designed for quick starts and stops, making them suitable for building functions-as-a-service (FaaS). The API is built with efficiency in mind, reducing infrastructure latency and ensuring fast boot times. While there may be some limitations due to hardware pinning, the overall design aims to provide simplicity and reliability.

Fly.io is currently not accepting new applications for the role of Rails developer enthusiastic about Hotwire. The company aims to make it easy to run full-stack apps near users and specializes in geographic distribution frameworks like Elixir's LiveView, Laravel's Livewire, and Ruby on Rail's Hotwire. Fly.io is looking for developers who can contribute substantially both to the platform and Rails/Hotwire itself. The salary ranges from $120k to $200k USD with competitive equity grants. Key responsibilities include maintaining and improving the Getting Started Guide, engaging in the Rails community, creating sample apps, providing support for other developers, and writing helpful articles. Candidates should enjoy using Hotwire, have good instincts for balancing competing customer demands, and be comfortable with software development.

In the past week, Fly.io has made several updates and improvements to its platform. They have provisioned new servers in Sydney, Dulles, and Dallas-Fort Worth to increase capacity for customers deploying apps in these regions. Additionally, they have added account deactivation features, allowing users more control over their non-personal organizations. Fly.io has also deployed an API with Ruby 3.1, which should improve performance and security. Other updates include a first draft of tips and tricks for working with Elixir UI code, improved error messaging from the API, and personalized support email addresses for paid plan users. Furthermore, they have fixed bugs related to community forum registration, IPv6 connection errors, billing issues, and volume metrics.

Ben Johnson, creator of BoltDB and Litestream, discusses the potential for SQLite to serve as a core data layer in full-stack applications. He provides a brief history of application databases and highlights how hardware improvements have made "big data" concepts less relevant for many applications. Johnson argues that SQLite, an embedded database known for its reliability and performance, can be a viable option for production applications with large numbers of users and high availability requirements. He introduces Litestream, an open-source project that enables replication in SQLite databases, making it more resilient to storage failures and concurrency at scale. Johnson emphasizes the simplicity and operational advantages of using SQLite as a data layer and encourages developers to consider this option for their applications.

Fly.io has released a series of updates and features to improve their platform, including faster first deployments with remote builders, enhanced security logging for their API, self-service payment retry feature on the dashboard, new persistent volumes default size, and more. They have also increased capacity in Hong Kong and added another wireguard gateway in IAD to support CI tools. Additionally, they have deployed initial changes to manage GraphQL API performance as queries become more complex.