Hello,
I would like to test Knime on my server by installing the following docker: knime/knime-full - Docker Image . I have read the documentation available on your site ( Business Hub Admin Guide | KNIME Documentation ). Is there an illustrative docker-compose file that would show the configuration/environment variables needed to be filled in and volume mappings?
Thank you in advance for your response.
The knime-full and knime-lts-full container images are not intended to be run stand alone in Docker. Instead they are intended to be run via kubernetes as part of your KNIME Business Hub environment.
It is technically possible to pull the container image and run it via Docker. However no compose files are available, and doing so wouldn’t result in a working executor as none of the control services would be present for it to talk to. As such you wouldn’t be able to send any jobs to the executor for processing.
If you’re looking to add Executor Images from the docker list, you can follow the documentation here:
which uses this data app that can assist you:
Thank you for your response. Thanks to you, I have clearly understood the specificity of this image of the knime-full container now.
I just found some time to devote to my research.
Thanks to your documentation and these ( Business Hub Admin Guide | KNIME Documentation and Workflow Automation in KNIME | KNIME ), I installed the executor on a self-hosted server (with docker-compose and Dockerfile). The tasks are sent to it by the scheduler xyOps (indicating the “.knwf” file to execute at the desired time). Everything works perfectly.
This stack is of course not as optimal as KNIME Business Hub since it multiplies the tools (KNIME Analytics Platform to create the “.knwf” + KNIME executor + scheduler xyOps) and does not allow to directly enter inside the planned workflow to identify and correct potential errors. The administration is therefore a bit more laborious than for KNIME Business Hub. But this allows to evaluate the potential before releasing a possible budget…