Merge branch 'feature/PI-506-upgrade_lambdas' into release/2024-09-13

infrastructure/terraform/etc/prod.tfvars

@@ -1,3 +1,4 @@
-account_name = "prod"
-environment  = "prod"
-domain       = "api.cpm.national.nhs.uk"
+account_name       = "prod"
+environment        = "prod"
+domain             = "api.cpm.national.nhs.uk"
+lambda_memory_size = 1536

infrastructure/terraform/etc/qa.tfvars

@@ -1,3 +1,4 @@
-account_name = "qa"
-environment  = "qa"
-domain       = "api.cpm.qa.national.nhs.uk"
+account_name       = "qa"
+environment        = "qa"
+domain             = "api.cpm.qa.national.nhs.uk"
+lambda_memory_size = 1536

infrastructure/terraform/per_workspace/main.tf

@@ -112,6 +112,7 @@ module "lambdas" {
       resources = local.permission_resource_map[replace(file, ".json", "")]
     }
   }
+  memory_size = var.lambda_memory_size
 }
 
 module "authoriser" {

infrastructure/terraform/per_workspace/modules/api_worker/api_lambda/lambda.tf

@@ -7,6 +7,7 @@ module "lambda_function" {
   handler       = "api.${var.name}.index.handler"
   runtime       = var.python_version
   timeout       = 10
+  memory_size   = var.memory_size
 
   timeouts = {
     create = "5m"

infrastructure/terraform/per_workspace/modules/api_worker/api_lambda/vars.tf

@@ -40,3 +40,7 @@ variable "attach_policy_statements" {
 variable "policy_statements" {
   default = {}
 }
+
+variable "memory_size" {
+  default = 128
+}

infrastructure/terraform/per_workspace/vars.tf

@@ -65,3 +65,7 @@ variable "python_version" {
 variable "domain" {
   type = string
 }
+
+variable "lambda_memory_size" {
+  default = 128
+}

scripts/test/test.mk

@@ -6,6 +6,7 @@ SDS_DEV_APIKEY =
 USE_CPM_PROD ?= FALSE
 TEST_COUNT =
 COMPARISON_ENV ?= local
+RUN_SPEEDTEST = ?= FALSE
 
 _pytest:
 	AWS_DEFAULT_REGION=$(AWS_DEFAULT_REGION) AWS_ACCESS_KEY_ID=$(AWS_ACCESS_KEY_ID) AWS_SECRET_ACCESS_KEY=$(AWS_SECRET_ACCESS_KEY) AWS_SESSION_TOKEN=$(AWS_SESSION_TOKEN) poetry run python -m pytest $(PYTEST_FLAGS) $(_INTERNAL_FLAGS) $(_CACHE_CLEAR)
@@ -40,4 +41,4 @@ test--feature--%--auto-retry:  ## Autoretry of failed feature (gherkin) tests
 	$(MAKE) test--feature--$* _INTERNAL_FLAGS="--define='auto_retry=true'"
 
 test--sds--matrix: ## Run end-to-end smoke tests that check data matches betweeen cpm and ldap
-	SDS_PROD_APIKEY=$(SDS_PROD_APIKEY) SDS_DEV_APIKEY=$(SDS_DEV_APIKEY) USE_CPM_PROD=$(USE_CPM_PROD) TEST_COUNT=$(TEST_COUNT) COMPARISON_ENV=$(COMPARISON_ENV) poetry run python -m pytest $(PYTEST_FLAGS) -m 'matrix' --ignore=src/layers --ignore=src/etl $(_CACHE_CLEAR)
+	SDS_PROD_APIKEY=$(SDS_PROD_APIKEY) SDS_DEV_APIKEY=$(SDS_DEV_APIKEY) USE_CPM_PROD=$(USE_CPM_PROD) TEST_COUNT=$(TEST_COUNT) COMPARISON_ENV=$(COMPARISON_ENV) RUN_SPEEDTEST=$(RUN_SPEEDTEST) poetry run python -m pytest $(PYTEST_FLAGS) -m 'matrix' --ignore=src/layers --ignore=src/etl $(_CACHE_CLEAR)

src/api/tests/sds_data_tests/calculation.py

@@ -0,0 +1,133 @@
+import ast
+import glob
+import json
+import math
+import os
+import statistics
+
+from event.json import json_loads
+
+
+def preprocess_json_file(file_path):
+    with open(file_path, "r") as f:
+        content = f.read().strip()
+        # Add brackets around the content to form a valid JSON array
+        content = f"[{content}]"
+
+        # Replace ',{' with '},{' to ensure proper JSON array formatting
+        content = content.replace("},{", "},{")
+
+        # Remove any trailing commas (`,]` is not valid JSON)
+        content = content.replace(",]", "]")
+
+        try:
+            data = json_loads(content)  # noqa: T201
+        except json.JSONDecodeError as e:
+            print(f"Error parsing JSON in file {file_path}: {e}")  # noqa: T201
+            raise
+    return data
+
+
+def transform_params(params_str):
+    # Convert the string representation of the dictionary to an actual dictionary
+    try:
+        params_dict = ast.literal_eval(params_str)
+    except (ValueError, SyntaxError) as e:
+        print(f"Error parsing string to dictionary: {e}")  # noqa: T201
+        return {}
+
+    # Transform the dictionary, excluding "use_cpm"
+    transformed_dict = {
+        f"request.params.{key}": value
+        for key, value in params_dict.items()
+        if key != "use_cpm"  # Ignore the "use_cpm" key
+    }
+
+    return transformed_dict
+
+
+def extract_response_times(json_files):
+    ldap_times = []
+    cpm_times = []
+    params = []
+
+    for file in json_files:
+        data = preprocess_json_file(file)
+        for entry in data:
+            ldap_times.append(entry["ldap_response_time"])
+            cpm_times.append(entry["cpm_response_time"])
+            params_dict = transform_params(entry["params"])
+            params_dict["path"] = entry["path"]
+            params.append(params_dict)
+
+    return ldap_times, cpm_times, params
+
+
+def format_value(value):
+    return f"{value:.2f} ms"
+
+
+def calculate_statistics(times_list):
+    if not times_list:
+        return {
+            "mean": "0.00 ms",
+            "mean_under_1s": "0.00 ms",
+            "mode": "N/A",
+            "lowest": "0.00 ms",
+            "highest": "0.00 ms",
+            "median": "0.00 ms",
+        }
+
+    try:
+        mode_value = statistics.mode(times_list)
+    except statistics.StatisticsError:
+        mode_value = "N/A"  # No unique mode found
+
+    geometric_mean = math.exp(sum(math.log(x) for x in times_list) / len(times_list))
+
+    # Filter times under 1000ms
+    times_under_1s = [time for time in times_list if time < 1000]
+
+    mean_under_1s = sum(times_under_1s) / len(times_under_1s) if times_under_1s else 0
+
+    return {
+        "mean": format_value(sum(times_list) / len(times_list)),
+        "mean_under_1s": format_value(mean_under_1s),
+        "mode": format_value(mode_value) if mode_value != "N/A" else mode_value,
+        "lowest": format_value(min(times_list)),
+        "highest": format_value(max(times_list)),
+        "median": format_value(statistics.median(times_list)),
+    }
+
+
+def write_to_json_file(output_file_path, data_list):
+    # Check if the file already exists
+    if os.path.exists(output_file_path):
+        print(  # noqa: T201
+            f"The file '{output_file_path}' already exists. No action will be taken."  # noqa: T201
+        )  # noqa: T201
+        return
+
+    # Write the list to the JSON file if it does not exist
+    with open(output_file_path, "w") as file:
+        json.dump(data_list, file, indent=4)
+
+
+# Get all JSON files in the directory
+# json_files = ["test_success_0.json", "test_success_1.json", "test_success_2.json"]
+json_files = glob.glob("src/api/tests/sds_data_tests/test_success_*.json")
+# json_files = glob.glob("*.json")
+
+# Extract response times
+ldap_times, cpm_times, params = extract_response_times(json_files)
+output_file_path = (
+    "src/api/tests/sds_data_tests/data/sds_fhir_api.speed_test_queries.device.json"
+)
+
+# Calculate statistics
+ldap_stats = calculate_statistics(ldap_times)
+cpm_stats = calculate_statistics(cpm_times)
+
+print(f"LDAP Response Time Stats: {ldap_stats}")  # noqa: T201
+print(f"CPM Response Time Stats: {cpm_stats}")  # noqa: T201
+write_to_json_file(output_file_path, params)

src/api/tests/sds_data_tests/conftest.py

@@ -1,11 +1,22 @@
 import json
+import subprocess
 
 import pytest
 
 current_file_index = 0
 entry_count = 0
 max_entries_per_file = 2000
-file_name_template = "test_success_{}.json"
+file_name_template = "src/api/tests/sds_data_tests/test_success_{}.json"
+
+
+@pytest.fixture(scope="session", autouse=True)
+def run_after_tests():
+    # This code will run after all tests have finished
+    yield
+    # Code to run after tests
+    subprocess.run(
+        ["python", "src/api/tests/sds_data_tests/calculation.py"], check=True
+    )
 
 
 def get_current_file():
@@ -41,7 +52,7 @@ def pytest_runtest_logreport(report):
             str(report.longreprtext)
         )
         output = {"failed_request": failed_request, "error": assertion_error}
-        with open("test_failure.json", "a") as f:
+        with open("src/api/tests/sds_data_tests/test_failure.json", "a") as f:
             f.write(json.dumps(output))
             f.write(",")
     if report.when == "call" and report.passed and hasattr(pytest, "success_message"):