Enable Swell to use GMAO AWS R2D2 server and S3 data store

docs/configuring_aws_server.md

@@ -0,0 +1,175 @@
+# Quick guide to test R2D2 server with Swell on Discover
+
+[Test outside of Swell](#2-to-test-outside-of-swell)
+
+## 1. To test within Swell:
+
+Make sure `~/.swell/r2d2_credentials.yaml` exists with your user/api_key/host/compiler and AWS credentials for S3 access.
+
+#### a. Set ~/.swell/r2d2_credentials.yaml
+
+```bash
+# R2D2 API credentials
+user: <your_username>
+api_key: <your_key>
+r2d2_host: discover
+r2d2_compiler: intel
+r2d2_server_host: "<enter_ip_address>"
+r2d2_server_port: "8080"
+
+# For S3 access
+aws_access_key_id : <access_key_id>
+aws_secret_access_key : <secret_access_key>
+aws_session_token : "<session_token>"
+
+```
+
+#### b. Quick test
+
+Run `IngestObs` directly without launching a full workflow:
+
+```bash
+# Create the experiment
+swell create ingest_obs_marine
+
+# Edit the generated experiment.yaml:
+#   - dry_run: false
+#   - obs_to_ingest: ['adt_cryosat2n']
+
+# Run the task
+swell task IngestObs <experiment_root>/swell-ingest_obs/swell-ingest_obs-suite/experiment.yaml \
+  -d 2021-07-02T06:00:00Z \
+  -m geos_marine
+```
+
+### Verify it is stored
+
+```python
+python3 -c "
+import r2d2
+results = r2d2.search(
+    item='observation',
+    observation_type='adt_cryosat2n',
+    window_start='20210702T060000Z',
+    window_length='PT6H'
+)
+print(f'Found {len(results)} records')
+for r in results:
+    print(r)
+"
+```
+
+### Test fetch
+
+```python
+python3 -c "
+import r2d2
+r2d2.fetch(
+    item='observation',
+    provider='odas',
+    observation_type='adt_cryosat2n',
+    file_extension='nc',
+    window_start='20210702T090000Z',
+    window_length='PT6H',
+    target_file='./test_fetch.nc'
+)
+print('Fetch OK')
+"
+```
+
+### Full workflow test
+
+```bash
+swell create ingest_obs_marine
+# Edit experiment.yaml: dry_run: false
+# Run the suite
+swell launch /path/to/suite/swell-ingest_obs/swell-ingest_obs-suite
+```
+
+This runs `IngestObs` for every cycle time across the date range.
+
+## 2. To test outside of Swell:
+
+#### a. Set up your environment
+
+Copy or create the required scripts into your working directory:
+- [load_r2d2.sh](../src/swell/utilities/scripts/load_r2d2.sh)
+- [prod_setup_env.sh](../src/swell/utilities/scripts/prod_setup_env.sh)
+
+Then source them to load R2D2 and set environment variables:
+
+```bash
+source load_r2d2.sh
+source prod_setup_env.sh
+```
+
+#### b. Configure AWS credentials:
+
+Set up AWS credentials for S3 access.
+
+```bash
+mkdir -p ~/.aws
+
+cat >> ~/.aws/credentials << 'EOF'
+[aws-us-east-1]
+aws_access_key_id = ACCESS_KEY
+aws_secret_access_key = SECRET_KEY
+EOF
+
+cat >> ~/.aws/config << 'EOF'
+[profile aws-us-east-1]
+region = us-east-1
+EOF
+```
+
+#### c. Test R2D2 store/fetch from Discover
+
+```bash
+python3 << 'EOF'
+import r2d2
+
+# Test metadata (API only)
+print("Data hubs:")
+for h in r2d2.search(item='data_hub'):
+    print(f"  {h.get('name')} ({h.get('platform')})")
+
+print("Data stores:")
+for s in r2d2.search(item='data_store'):
+    print(f"  {s.get('name')}")
+
+print("Compute hosts:")
+for c in r2d2.search(item='compute_host'):
+    print(f"  {c.get('name')}")
+
+# Test store (API + S3)
+import tempfile, os
+test_file = os.path.join(tempfile.gettempdir(), 'r2d2_test.txt')
+with open(test_file, 'w') as f:
+    f.write("test from Discover\n")
+
+r2d2.store(
+    item='observation',
+    provider='test',
+    observation_type='test_obs',
+    file_extension='txt',
+    window_start='20240101T120000Z',
+    window_length='PT6H',
+    source_file=test_file
+)
+print("Store OK")
+
+# Test fetch (API + S3)
+fetch_file = os.path.join(tempfile.gettempdir(), 'r2d2_fetched.txt')
+r2d2.fetch(
+    item='observation',
+    provider='test',
+    observation_type='test_obs',
+    file_extension='txt',
+    window_start='20240101T120000Z',
+    window_length='PT6H',
+    target_file=fetch_file
+)
+print(f"Fetch OK: {open(fetch_file).read().strip()}")
+EOF
+```
+

docs/examples/r2d2/ingest_obs.md

@@ -14,8 +14,8 @@ Add your observation name to the list and set the date range:
 
 ```python
 qd.obs_to_ingest(['adt_cryosat2n', 'my_obs'])   # Add yours here
-qd.start_cycle_point("2021-07-02T06:00:00Z")    # When to start
-qd.final_cycle_point("2021-07-03T06:00:00Z")    # When to stop
+qd.start_cycle_point("2023-07-02T06:00:00Z")    # When to start
+qd.final_cycle_point("2023-07-03T06:00:00Z")    # When to stop
 ```
 
 ### Step 2: Make sure your observation is registered

src/swell/tasks/ingest_obs.py

@@ -153,7 +153,8 @@ def process_obs_config(
         provider = get_provider_for_observation(
             obs_name, self.ioda_names_list, self.logger)
 
-        retrieval_method = config.get('retrieval_method')  # cp or s3
+        # where the files are located
+        retrieval_method = config.get('retrieval_method')
 
         # Determine source pattern based on method
         source_pattern = config.get(

src/swell/utilities/r2d2.py

@@ -106,6 +106,7 @@ def _get_platform_r2d2_config(logger: Logger, platform: str = None) -> tuple:
         return None, None
 
     # Platform-specific R2D2 configurations
+    # Note: ~/.swell/r2d2_credentials.yaml overrides these values if specified
     platform_configs = {
         'nccs_discover_sles15': {
             'host': 'discover-gmao',
@@ -115,10 +116,6 @@ def _get_platform_r2d2_config(logger: Logger, platform: str = None) -> tuple:
             'host': 'discover-gmao',
             'compiler': 'intel'
         },
-        'aws': {
-            'host': 'aws-gmao',
-            'compiler': 'intel'  # or 'gnu' depending on AWS setup
-        },
         'generic': {
             'host': None,
             'compiler': None
@@ -180,21 +177,37 @@ def load_r2d2_credentials(
     if 'api_key' in credentials and 'R2D2_API_KEY' not in os.environ:
         os.environ['R2D2_API_KEY'] = credentials['api_key']
 
+    if 'r2d2_server_host' in credentials and 'R2D2_SERVER_HOST' not in os.environ:
+        os.environ['R2D2_SERVER_HOST'] = credentials['r2d2_server_host']
+
+    if 'r2d2_server_port' in credentials and 'R2D2_SERVER_PORT' not in os.environ:
+        os.environ['R2D2_SERVER_PORT'] = str(credentials['r2d2_server_port'])
+
+    if 'aws_access_key_id' in credentials and 'AWS_ACCESS_KEY_ID' not in os.environ:
+        os.environ['AWS_ACCESS_KEY_ID'] = credentials['aws_access_key_id']
+
+    if 'aws_secret_access_key' in credentials and 'AWS_SECRET_ACCESS_KEY' not in os.environ:
+        os.environ['AWS_SECRET_ACCESS_KEY'] = credentials['aws_secret_access_key']
+
+    if 'aws_session_token' in credentials and 'AWS_SESSION_TOKEN' not in os.environ:
+        os.environ['AWS_SESSION_TOKEN'] = credentials['aws_session_token']
+
     # Set host and compiler (YAML config takes precedence over platform detection)
-    if 'host' in credentials and 'R2D2_HOST' not in os.environ:
-        os.environ['R2D2_HOST'] = credentials['host']
-        logger.info(f"Using platform host '{r2d2_host}' (overriding YAML '{credentials['host']}')")
+    if 'r2d2_host' in credentials and 'R2D2_HOST' not in os.environ:
+        os.environ['R2D2_HOST'] = credentials['r2d2_host']
+        logger.info(f"Using platform host '{r2d2_host}' \
+                    (overriding YAML '{credentials['r2d2_host']}')")
         logger.warning("Using host from YAML file")
 
     elif r2d2_host and 'R2D2_HOST' not in os.environ:
         os.environ['R2D2_HOST'] = r2d2_host
         logger.info(f"Set R2D2_HOST={r2d2_host} from platform configuration")
 
     # Set compiler
-    if 'compiler' in credentials and 'R2D2_COMPILER' not in os.environ:
-        os.environ['R2D2_COMPILER'] = credentials['compiler']
+    if 'r2d2_compiler' in credentials and 'R2D2_COMPILER' not in os.environ:
+        os.environ['R2D2_COMPILER'] = credentials['r2d2_compiler']
         logger.info(f"Using platform compiler '{r2d2_compiler}' \
-                    (overriding YAML '{credentials['compiler']}')")
+                    (overriding YAML '{credentials['r2d2_compiler']}')")
         logger.warning("Using compiler from YAML file")
 
     elif r2d2_compiler and 'R2D2_COMPILER' not in os.environ:

src/swell/utilities/scripts/prod_setup_env.sh

@@ -3,15 +3,14 @@ unset R2D2_SERVER_PORT
 
 export R2D2_USER=username
 export R2D2_API_KEY=api_key
-export R2D2_HOST=discover-gmao
-export R2D2_COMPILER=intel
-
-source venv_client/bin/activate
+export R2D2_HOST="discover"
+export R2D2_COMPILER="intel"
+export R2D2_SERVER_HOST="http://13.217.72.149"
+export R2D2_SERVER_PORT="8080"
 
 echo “  R2D2 Production environment:”
 echo “  R2D2_API_KEY:  [set]”
-echo “  R2D2_SERVER_HOST: $R2D2_SERVER_HOST '(should be empty)'”
-echo “  R2D2_SERVER_PORT:  $R2D2_SERVER_PORT '(should be empty)'”
-echo “  - Client should default to https://r2d2-api.jcsda.org”
+echo “  R2D2_SERVER_HOST: $R2D2_SERVER_HOST ”
+echo “  R2D2_SERVER_PORT:  $R2D2_SERVER_PORT ”
 echo “  R2D2_HOST: $R2D2_HOST”
 echo “  R2D2_COMPILER: $R2D2_COMPILER”