August 2026 Summaries

Kestra has become an official DigitalOcean technology partner and released a DigitalOcean plugin with blueprints for automating cloud operations. The plugin provides declarative tasks for managing Droplets, managed database clusters, DNS domains and records, Kubernetes clusters, firewalls, volumes, and load balancers, while a droplet trigger can detect newly created instances across an account. It supports configurable list outputs for iteration, reporting, and analytics, and encourages secure use of DigitalOcean personal access tokens through Kestra secrets rather than embedding credentials in workflow code. The announcement emphasizes that Kestra is designed for multi-step, failure-aware orchestration, such as provisioning temporary compute with guaranteed cleanup, monitoring and governing unapproved infrastructure, automating backups and DNS cutovers, and generating resource inventories. Available starter blueprints include ephemeral droplet runners, database resizing, tagged snapshot backups, new-droplet governance, blue-green DNS cutovers, and weekly account inventory reports.

Kestra 2.0 is a forthcoming open-source Apache 2.0-licensed release that rebuilds the orchestration platform’s execution engine, redesigns its interface, and adds features aimed at distributed deployments, performance, developer productivity, and AI integration. Stateless workers communicate with the control plane through gRPC without direct database access, while on-demand loading of large task outputs is intended to improve UI and database performance at scale. The release introduces cleaner flow semantics, runnable drafts, natural-language AI flow generation, and Model Context Protocol support that can expose flows as tools for AI clients. Kestra is testing release candidates across deployment configurations and in its own production environment, while inviting users to report issues, contribute, or join an Early Adopter Program that provides pre-release builds, office hours, engineering support, and migration assistance. Existing flows are expected to remain compatible, though advanced features such as ForEach and trigger conditions may need guided migration using a public CLI and detailed documentation.

Kestra is presented as an orchestration platform for Electronic Design Automation workflows, addressing hardware teams’ needs for large-scale parallel simulation, costly synthesis runs, artifact handling, and automated Power, Performance, and Area checks that conventional CI/CD tools may not manage effectively. In a nightly RISC-V ALU regression example, Kestra runs separate Verilator containers concurrently to test addition and subtraction operations, then invokes Yosys to synthesize the Verilog design and extract its gate count as a managed artifact. A downstream quality gate compares that count with a 500-gate area budget, causing an initial 32-bit, 903-gate design to fail before it reaches more expensive stages such as place and route. After the ALU datapath is reduced to 8 bits and its testbench updated, the same pipeline passes verification and synthesis with a 230-gate result. The approach uses version-controlled YAML flows and source files, Docker-based tools, scalable worker groups, managed artifacts, and retry mechanisms that can accommodate shared commercial EDA licenses, while remaining compatible with both open-source and command-line commercial tools.