The relevant files to look at are:

• across_review/docker/run-scripts/run_ert_venv.sh: This is the entrypoint for the dockerfile and where I start the dask-scheduler, dask-worker and the bokeh server

• across_review/docker/run-scripts/flow_venv.sh this is where I run flow from within ERT. It essentially just calls mpirun -np 2 flow --enable-damaris-output=true "$@"

• across_review/docker/damaris-scripts/damaris.xml this is the xml I provide damaris through flow (will get the regexp treatment)

• across_review/docker/damaris-scripts/publish_data.py this is where I publish the data from damaris

• across_review/docker/damaris-scripts/main.py the bokeh server app.

You should be able to run this yourself by running from the root of the repository

docker run -p 5006:5006 --rm -it -u $(id -u):$(id -g) -v $(realpath ert_example):/workflow -w /workflow -e 'NUMBER_OF_CPUS=2' kjetilly/damaris run_ert_venv.sh ensemble_experiment spe1.ert

where NUMBER_OF_CPUS can be changed (it is passed on to HQ).

You can also alter the XML file that is read from OPM Flow by specifiying the environment variable FLOW_DAMARIS_CONFIG_XML_FILE