December 2014 Summaries

Octopus Deploy 2.6, officially released after a few weeks of pre-release and feedback-driven improvements, introduces features such as life cycles and parallel deployments. Founder and CEO Paul Stovell announced the release, encouraging users to explore an extensive blog post detailing the new version's offerings. This marks a festive milestone for Octopus HQ as the team winds down for the holidays, having fine-tuned the software to enhance user experience. Paul, who started Octopus with his wife Sonia in 2012, enjoys carpentry and DIY projects in his free time.

Paul Stovell discusses the transition of Octopus Deploy from RavenDB to SQL Server for version 3.0, detailing the technical and operational motivations behind the switch. The move aims to streamline high availability and backup processes by consolidating data storage into SQL Server, which is familiar to enterprises and DBA's. The shift involves storing deployment logs, in-progress deployment states, and metadata in SQL Server, while NuGet packages remain on disk. Octopus Deploy leverages SQL Server's document store capabilities by storing complex object graphs as JSON blobs, simplifying data access without relying on an ORM. The transition promises improved performance and reliability, although it removes the capability to restart the server during deployments, a feature that was unreliable. The migration from version 2.6 to 3.0 is designed to be straightforward, requiring users to import data from a previous backup. Testing is rigorous, with performance benchmarks ensuring the new version can handle significantly larger installations without increased hardware demands.