Further the adaption of GPU Imaging/Gain Calibration Pipeline as an SDP Workflow

Containerised workflow that can be run with a well defined configuration that does not require recompiling the code. Demo and documented example (added to the API/code documentation) of this code provided so that the steps required to integrate this into the MVP as a workflow callable can be easily assessed and planned for a future PI. Stretch: Integration of the workflow into the SDP MVP.

The GPU Imaging/Gain Calibration Pipeline has been successfully modified to operate as a containerised workflow, and can be executed using dynamic algorithm configurations (via YAML file) without need for source code recompilation. The repository for this pipeline can be found at the following GitLab address: https://gitlab.com/ska-telescope/sep_pipeline_imaging The Dockerfile for containerising the pipeline can be found at the following address: https://gitlab.com/ska-telescope/sep_pipeline_imaging/-/blob/master/Dockerfile The kubernetes helm chart for the pipeline can be found at the following address: https://gitlab.com/ska-telescope/sep_pipeline_imaging/-/tree/master/charts/sep-imaging-pipeline A Makefile is used to simplify the building and pushing of the pipeline Docker image, as well as the deployment and logging of the pipeline (via helm chart) to a running instance of kubernetes. This can be found at the following address (pinned to the deployment of the pipeline): https://gitlab.com/ska-telescope/sep_pipeline_imaging/-/blob/master/Makefile#L148 Please note that the execution of the pipeline via dynamic configuration is only a proof of concept at this point. The demonstration of the pipeline will show that only a handful of algorithm configurations will be set using this technique. Full implementation of dynamic configuration is to be completed in Sprint 6.4 (NZAPP-182). The proof of concept source code for obtaining dynamic configuration via YAML file can be located in the following function call: https://gitlab.com/ska-telescope/sep_pipeline_imaging/-/blob/master/imaging/src/controller.cpp#L232 - It is assumed that a YAML file has been supplied in order for the function to be called. The current proof of concept configuration YAML file can be found at the following address: https://gitlab.com/ska-telescope/sep_pipeline_imaging/-/blob/master/charts/sep-imaging-pipeline/imageconfigs/pipeline.yaml A demonstration for this work will be presented on on Wednesday 14th April, during System Demo 6.2. Documentation for this work will be included into the Pipelines Sphinx documentation during Sprint 6.4 for the NZAPP team (NZAPP-201). The stretch criteria was not met due to time constraints.