From 5b4f3967f7012384cc2c8b7ae4227f492b0a3941 Mon Sep 17 00:00:00 2001
From: Christian Chwala <christian.chwala@kit.edu>
Date: Tue, 31 Mar 2026 20:51:32 +0200
Subject: [PATCH] fix(mno-simulator): set loop_duration to full 3-month dataset

- Set loop_duration_seconds to 7948860 (~92 days) so the simulator
  only restarts after exhausting the entire source dataset instead
  of looping every hour
- Update README: document auto-download behaviour, add actual
  NETCDF_FILE_URL for standalone use, fix stale loop_duration example
---
 mno_data_source_simulator/README.md  | 25 +++++++++++++++++++++++--
 mno_data_source_simulator/config.yml |  6 ++++--
 2 files changed, 27 insertions(+), 4 deletions(-)

diff --git a/mno_data_source_simulator/README.md b/mno_data_source_simulator/README.md
index facf1fb..1944630 100644
--- a/mno_data_source_simulator/README.md
+++ b/mno_data_source_simulator/README.md
@@ -44,7 +44,9 @@ Edit `config.yml` for local/standalone use, or set environment variables for Doc
 ```yaml
 data_source:
   netcdf_file: "/app/example_data/openMRG_cmls_20150827_3months.nc"
-  loop_duration_seconds: 3600  # Lookback window for real-time replay
+  # Set to the full dataset duration so the simulator only loops after all
+  # source data is exhausted (openMRG_cmls_20150827_3months.nc ≈ 92 days).
+  loop_duration_seconds: 7948860
 
 generator:
   generation_frequency_seconds: 10  # How often to generate files
@@ -61,7 +63,26 @@ sftp:
 | Variable | Description |
 |---|---|
 | `NETCDF_FILE` | Path to the NetCDF source file |
-| `NETCDF_FILE_URL` | URL to download the file if not already present |
+| `NETCDF_FILE_URL` | URL to download the file automatically if not already present |
+
+### NetCDF dataset auto-download
+
+The simulator downloads the 3-month source file automatically on startup if the
+path set in `NETCDF_FILE` does not exist **and** `NETCDF_FILE_URL` is set.
+
+**Docker (`docker-compose up mno_simulator`):** `NETCDF_FILE_URL` is already
+configured in `docker-compose.yml`, so the file is downloaded automatically —
+no extra steps needed.
+
+**Standalone:** export the URL before starting the simulator:
+
+```bash
+export NETCDF_FILE_URL=https://bwsyncandshare.kit.edu/s/jSAFftGXcJjQbSJ/download
+python main.py
+```
+
+The file (~193 MB) is saved to the path given by `NETCDF_FILE` (or
+`data_source.netcdf_file` in `config.yml`) and is only downloaded once.
 
 ### Authentication
 
diff --git a/mno_data_source_simulator/config.yml b/mno_data_source_simulator/config.yml
index a17150b..d8227a4 100644
--- a/mno_data_source_simulator/config.yml
+++ b/mno_data_source_simulator/config.yml
@@ -7,8 +7,10 @@ data_source:
   # downloaded automatically at startup if NETCDF_FILE_URL is set.
   # Override via the NETCDF_FILE environment variable.
   netcdf_file: "/app/example_data/openMRG_cmls_20150827_3months.nc"
-  # Loop duration in seconds (1 hour = 3600 seconds)
-  loop_duration_seconds: 3600
+  # Loop duration in seconds — set to the full 3-month dataset duration so the
+  # simulator only restarts after exhausting all source data.
+  # (openMRG_cmls_20150827_3months.nc: 7 948 860 s ≈ 92 days)
+  loop_duration_seconds: 7948860
 
 # SFTP upload configuration
 sftp: