From ee49c54525aae584a100e043713e6fd7e87070af Mon Sep 17 00:00:00 2001
From: Sigurd Pettersen <sigurd.pettersen@ceetronsolutions.com>
Date: Thu, 21 Nov 2024 21:59:00 +0100
Subject: [PATCH] Add performance metrics to surface query service and adjust
 Radix configuration for scaling

---
 backend_py/primary/primary/main.py             |  1 +
 .../surface_query_service.py                   | 18 +++++++++++++++++-
 radixconfig.yml                                | 10 +++++-----
 3 files changed, 23 insertions(+), 6 deletions(-)

diff --git a/backend_py/primary/primary/main.py b/backend_py/primary/primary/main.py
index d2e9fa22e..6f12bf57a 100644
--- a/backend_py/primary/primary/main.py
+++ b/backend_py/primary/primary/main.py
@@ -46,6 +46,7 @@
 logging.getLogger("primary.services.sumo_access").setLevel(logging.DEBUG)
 logging.getLogger("primary.services.user_session_manager").setLevel(logging.DEBUG)
 logging.getLogger("primary.services.user_grid3d_service").setLevel(logging.DEBUG)
+logging.getLogger("primary.services.surface_query_service").setLevel(logging.DEBUG)
 logging.getLogger("primary.routers.grid3d").setLevel(logging.DEBUG)
 logging.getLogger("primary.routers.dev").setLevel(logging.DEBUG)
 
diff --git a/backend_py/primary/primary/services/surface_query_service/surface_query_service.py b/backend_py/primary/primary/services/surface_query_service/surface_query_service.py
index ca8e19f37..4841ac36c 100644
--- a/backend_py/primary/primary/services/surface_query_service/surface_query_service.py
+++ b/backend_py/primary/primary/services/surface_query_service/surface_query_service.py
@@ -11,6 +11,9 @@
 from primary import config
 from primary.services.sumo_access.sumo_blob_access import get_sas_token_and_blob_store_base_uri_for_case
 
+from webviz_pkg.core_utils.perf_metrics import PerfMetrics
+
+
 LOGGER = logging.getLogger(__name__)
 
 
@@ -51,6 +54,9 @@ async def batch_sample_surface_in_points_async(
     x_coords: list[float],
     y_coords: list[float],
 ) -> List[RealizationSampleResult]:
+    
+    perf_metrics = PerfMetrics()
+
     realization_object_ids = await _get_object_uuids_for_surface_realizations(
         sumo_access_token=sumo_access_token,
         case_uuid=case_uuid,
@@ -59,9 +65,15 @@ async def batch_sample_surface_in_points_async(
         surface_attribute=surface_attribute,
         realizations=realizations,
     )
-
+    
+    perf_metrics.record_lap("get-obj-ids")
+    
     sas_token, blob_store_base_uri = get_sas_token_and_blob_store_base_uri_for_case(sumo_access_token, case_uuid)
 
+    #LOGGER.debug(f"Token: {sas_token}")
+
+    perf_metrics.record_lap("sas-token")
+
     request_body = _PointSamplingRequestBody(
         sasToken=sas_token,
         blobStoreBaseUri=blob_store_base_uri,
@@ -74,6 +86,8 @@ async def batch_sample_surface_in_points_async(
         LOGGER.info(f"Running async go point sampling for surface: {surface_name}")
         response: httpx.Response = await client.post(url=SERVICE_ENDPOINT, json=request_body.model_dump())
 
+    perf_metrics.record_lap("main-call")
+
     json_data: bytes = response.content
     response_body = _PointSamplingResponseBody.model_validate_json(json_data)
 
@@ -82,6 +96,8 @@ async def batch_sample_surface_in_points_async(
         values_np = np.asarray(res.sampledValues)
         res.sampledValues = np.where((values_np < response_body.undefLimit), values_np, np.nan).tolist()
 
+    LOGGER.debug(f"------------------ batch_sample_surface_in_points_async() took: {perf_metrics.to_string_s()}")
+
     return response_body.sampleResultArr
 
 
diff --git a/radixconfig.yml b/radixconfig.yml
index ccd4a8d1e..4a2257e3e 100644
--- a/radixconfig.yml
+++ b/radixconfig.yml
@@ -134,7 +134,7 @@ spec:
           cpu: 2000m
       horizontalScaling:
         minReplicas: 1
-        maxReplicas: 16
+        maxReplicas: 8
         resources:
           memory:
             averageUtilization: 60
@@ -145,13 +145,13 @@ spec:
           resources:
             requests:
               memory: 32Gi
-              cpu: 8000m
+              cpu: 32000m
             limits:
               memory: 32Gi
-              cpu: 8000m
+              cpu: 32000m
           horizontalScaling:
-            minReplicas: 16
-            maxReplicas: 16
+            minReplicas: 8
+            maxReplicas: 8
             resources:
               memory:
                 averageUtilization: 60