DPPS SimPipe: Integration and Release

The CTAO DPPS Simulation Production Pipeline (SimPipe) provides the software, workflows, and data models for generating accurate Monte Carlo simulations of the CTAO observatory.

Installation

The following installation procedures are implemented in the gitlab CI/CD pipeline:

• simtools is installed using pip
• CORSIKA is installed using a tar-file (currently downloaded from a cloud storage)
• sim_telarray is installed using a tar-file (currently downloaded from a cloud storage); planned to be installed from gitlab
• simulation model databases - no installed required; configuration of secrets for access

Download of corsika / sim_telarray is facilitated by a private upload to the DESY Sync&Share. Ask the maintainers to provide the token to you and define it in a .env file in this repository:

SOFTWARE_DOWNLOAD_SECRET=<the token received from the maintainers>

Then run make build-dev-docker to build the simpipe container locally.

Preparing a new release

Open a new branch from main for a release candidate. Use update_simpipe.py to update the simtools image and submodule, the simulation-models Git revision, and the test resources used in CI. Supply the versions you want to change:

• simtools version (see simtools releases)
• simulation-models version (see simulation-models releases)
• simtools-tests Git revision (--tests-revision)
• test resource version (--tests-version); when omitted with --simtools vX.Y.Z, this defaults to vX.Y.0
• dpps-aiv-toolkit version (see dpps-aiv-toolkit releases)

Example:

python ./update_simpipe.py --simtools v0.38.1 --simulation-models v0.18.0 --tests-revision v4.0.0 --aiv-submodule v3.0.0

After this open a merge request, get it reviewed and merge it into main. Create a new release through the Gitlab interface.

Developer Notes

• a local simulation models database is used for testing. Uploading details are defined in chart/templates/bootstrapSimulationModel.yaml