Skip to main content

Camera Discovery and Client Worker Checks

Synopsis​

Intended end result: Restore camera ingestion when workers are idle or cameras are missing.

How this page gets you there: Inspect worker logs and environment inputs, then restart only the ingestion path and validate new event flow.

Try VS Code tunnel first for reviewing logs and editing config files. It is usually faster and less error-prone than running repeated shell commands.

When tunneled into the VM, start with:

  • /mnt/datadisk/condor/client/logs/calfire_inference.out.log
  • /mnt/datadisk/condor/client/logs/calfire_inference.err.log
  • /mnt/datadisk/condor/client/logs/calfire_inference_panorama.out.log
  • /mnt/datadisk/condor/client/.env

Fallback: VM command-line checks​

If you cannot use a VS Code tunnel, run these VM commands instead.

Inspect supervisord-managed client logs directly:

sudo docker exec calfire_client sh -lc 'tail -n 120 /app/logs/calfire_inference.out.log'
sudo docker exec calfire_client sh -lc 'tail -n 120 /app/logs/calfire_inference.err.log'
sudo docker exec calfire_client sh -lc 'tail -n 120 /app/logs/calfire_inference_panorama.out.log'

Confirm the client container received expected camera and backend environment variables:

sudo docker exec calfire_client sh -lc 'env | egrep "^(ALERTCA|FIRECAMS|CAMERAS|POLL_SECONDS|TRITON_URL|DB_)=" | sort'

If the client reports No cameras to process. Sleeping., update /mnt/datadisk/condor/client/.env with explicit camera values, then restart only the ingestion path:

cd /mnt/datadisk/condor
sudo docker compose --project-name condor -f docker-compose.yaml restart client server_collect
sudo docker exec -it calfire_db sh -lc 'psql -U "$POSTGRES_USER" -d "$POSTGRES_DB" -c "select max(ts_utc) from inference_events;"'