Insitu query logging & request timeouts

CHANGELOG.md

@@ -30,6 +30,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Bumped ingress timeout in Helm chart to reflect AWS gateway timeout
 - SDAP-399: Updated quickstart guide for standalone docker deployment of SDAP.
 - SDAP-399: Updated quickstart Jupyter notebook
+- Added logging message for start of insitu query + added status code & elapsed time to post query log message.
+- Added explicit timeouts for all insitu related queries to prevent hanging issue.
 ### Deprecated
 ### Removed
 - removed dropdown from matchup doms endpoint secondary param

analysis/webservice/algorithms/doms/insitu.py

@@ -5,6 +5,8 @@
 import requests
 from datetime import datetime
 from webservice.algorithms.doms import config as insitu_endpoints
+from urllib.parse import urlencode
+from webservice.webmodel import NexusProcessingException
 
 
 def query_insitu_schema():
@@ -15,7 +17,11 @@ def query_insitu_schema():
     """
     schema_endpoint = insitu_endpoints.getSchemaEndpoint()
     logging.info("Querying schema")
-    response = requests.get(schema_endpoint)
+    try:
+        response = requests.get(schema_endpoint, timeout=(15.05, 331))
+    except (requests.exceptions.ReadTimeout, requests.exceptions.ConnectTimeout):
+        raise NexusProcessingException(code=504, reason=f'Insitu schema request timed out')
+
     response.raise_for_status()
     return response.json()
 
@@ -60,12 +66,22 @@ def query_insitu(dataset, variable, start_time, end_time, bbox, platform, depth_
     # Page through all insitu results
     next_page_url = insitu_endpoints.getEndpoint(provider, dataset)
     while next_page_url is not None and next_page_url != 'NA':
-        if session is not None:
-            response = session.get(next_page_url, params=params)
+        thetime = datetime.now()
+        if params == {}:
+            logging.info(f"Starting insitu request: {next_page_url}")
         else:
-            response = requests.get(next_page_url, params=params)
+            logging.info(f"Starting insitu request: {next_page_url}?{urlencode(params)}")
+
+
+        try:
+            if session is not None:
+                response = session.get(next_page_url, params=params, timeout=(15.05, 331))
+            else:
+                response = requests.get(next_page_url, params=params, timeout=(15.05, 331))
+        except (requests.exceptions.ReadTimeout, requests.exceptions.ConnectTimeout):
+            raise NexusProcessingException(code=504, reason=f'Insitu request timed out after {str(datetime.now() - thetime)} seconds')
 
-        logging.info(f'Insitu request {response.url}')
+        logging.info(f'Insitu request {response.url} finished. Code: {response.status_code} Time: {str(datetime.now() - thetime)}')
 
         response.raise_for_status()
         insitu_page_response = response.json()