This release marks a major milestone in Transform Hub development: the ability to run workloads on a Kubernetes cluster! Additionally, Python runner can now be used inside Docker containers.

Kubernetes adapter!

We're excited to announce that STH can now run on any Kubernetes cluster. This means there are now three ways for running your sequences, each with a dedicated adapter:

• kubernetes - for deployment in cloud environments and leverage multiple servers
• docker - the default for running sequences in a standard server space
• process - the adapter running non-separated processes for simple deployments

We're still working on the documentation and deployment guides - expect them soon!

Docker support for Python runner

Sequences written in Python can now be run inside Docker containers, using a prebuilt scramjetorg/runner-py image available from Docker hub. This image is based on python:3.8.12-bullseye.

No changes in your sequences are required! Simply start STH with docker runtime adapter, and Python runner will automatically run inside a container.

Groundwork for platform security

For our beta platform users (here's a link if you'd like to become one!) there's one final touch of securing the connection with the platform with SSL.

Minor changes

There were already a couple patch releases after the initial 0.18.0 version, as we addressed several issues that were uncovered.

• Fix for starting the Transform-Hub with mutually exclusive runtime adapters options
• Add responding to OPTIONS requests and setting CORS in the Transform-Hub API server
• Significant improvements in BDD tests (speed up, remove overlapping scenarios)
• Fixes in Python dependency installation and image building
• Docker images are based on newer Debian Bullseye (instead of Buster)
• Restored proper behavior of STOP/KILL control messages for sequences
• Dependencies security upgrade