Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Troubleshooting & FAQ

Running simulations

Only one simulation runs at a time

OEDI-SI starts a single HELICS broker on a fixed port, so the backend permits one active run at a time. If a new run is rejected, cancel the current run (the Cancel button on the run detail page, or DELETE /api/runs/{run_id}) or wait for it to finish.

“Address already in use” / broker port conflict

A previous simulation may have left a broker running. Find and stop stray brokers:

pgrep -fl helics_broker      # list broker processes
pkill -f helics_broker       # stop them

Then start your run again. You can also change the broker port with oedisi build --broker-port <port>.

A run fails immediately

Read the per-federate logs. Locally they are written next to the build output (build/<component>.log); in the UI they appear in the Logs section of the run detail page, and via GET /api/runs/{run_id}/logs/{component}. A component that exits at startup is usually missing a Python dependency or a required static_input.

ModuleNotFoundError from a component (local runs)

When running locally (not in Docker), each component runs in your Python environment, so its dependencies must be installed. Check the component’s pyproject.toml and install the missing packages, or use the multi-container workflow for isolation.

Setup

OEDISI_COMPONENTS is not set

The frontend backend resolves components through OEDISI_COMPONENTS, which must point at the Components/ directory of oedisi-components:

export OEDISI_COMPONENTS="/path/to/oedisi-components/Components"

oedisi: command not found

The CLI ships with the oedisi package. Install it (pip install oedisi) and confirm the environment’s bin directory is on your PATH.

helics run is missing

Install the HELICS launcher and broker: pip install "helics[cli]" helics-apps. Verify with helics run --help and which helics_broker.

Docker / multi-container

Builds fail or hang on Apple Silicon (M1/M2/M3)

Force the image platform so multi-arch base images resolve correctly:

export DOCKER_DEFAULT_PLATFORM=linux/amd64

Stale containers or networks

Prune before a fresh multi-container run so old networks don’t collide:

docker system prune -f
docker network prune -f

Still stuck?