Fix ECH and Serverless test workflows (#3086)

* update serverless wf

* fix str

* fix lint issues

* update test suite

* add retry mechanism for aws provider

* update provider config

* revert bump version change

Author: gurevichdmitry

.github/workflows/weekly-serverless.yml

@@ -11,12 +11,22 @@ jobs:
     runs-on: ubuntu-latest
     outputs:
       deployment_name: ${{ steps.set_deployment_name.outputs.date-name }}
+      stack_version: ${{ steps.set_deployment_name.outputs.stack-version }}
     steps:
+      - name: Check out the repo
+        uses: actions/checkout@v4
+
+      - name: Hermit Environment
+        uses: ./.github/actions/hermit
+
       - name: Set deployment name
         id: set_deployment_name
+        env:
+          STACK_VERSION: ${{ env.ELK_VERSION }}
         run: |
           date_name=$(echo "prd-env-$(date +'%d%b%H%M')" | tr '[:upper:]' '[:lower:]')
           echo "date-name=$date_name" >> $GITHUB_OUTPUT
+          echo "stack-version=$STACK_VERSION" >> $GITHUB_OUTPUT
 
   deploy:
     needs: naming
@@ -30,7 +40,7 @@ jobs:
       deployment_name: ${{ needs.naming.outputs.deployment_name }}
       # For now, the region is not used because it's overridden in the tf, but it's here for future compatibility.
       ess-region: "aws-us-east-1"
-      elk-stack-version: 9.0.0-SNAPSHOT
+      elk-stack-version: ${{ needs.naming.outputs.stack_version }}
       serverless_mode: true
       run-sanity-tests: true
       expiration_days: 0

deploy/test-environments/cis/main.tf

@@ -1,5 +1,6 @@
 provider "aws" {
-  region = var.region
+  region      = var.region
+  max_retries = 30
 }
 
 locals {

tests/integration/tests/test_sanity_checks.py

@@ -30,6 +30,8 @@
 COMPONENTS_BACKOFF = 10
 
 AGENT_VERSION = elasticsearch.agent_version
+if AGENT_VERSION.endswith("SNAPSHOT"):
+    AGENT_VERSION = AGENT_VERSION.split("-")[0]
 
 # Check if AGENT_VERSION is provided
 if not AGENT_VERSION:
@@ -181,7 +183,6 @@ def test_cspm_aws_findings(
         query_list = build_query_list(
             benchmark_id="cis_aws",
             match_type=match_type,
-            version=AGENT_VERSION,
             agent=agent,
         )
         query, sort = cspm_client.build_es_must_match_query(must_query_list=query_list, time_range="now-24h")
@@ -244,7 +245,6 @@ def test_cspm_gcp_findings(
         query_list = build_query_list(
             benchmark_id="cis_gcp",
             match_type=match_type,
-            version=AGENT_VERSION,
             agent=agent,
         )
         query, sort = cspm_client.build_es_must_match_query(must_query_list=query_list, time_range="now-24h")
@@ -279,7 +279,6 @@ def test_cspm_azure_findings(
     for agent in azure_agents:
         query_list = build_query_list(
             benchmark_id="cis_azure",
-            version=AGENT_VERSION,
             match_type=match_type,
             agent=agent,
         )

tests/integrations_setup/configuration_fleet.py

@@ -29,6 +29,7 @@
     "Key=project,Value=test-environments"
 )
 
+DOCKER_ARTIFACTORY_AGENT = "docker.elastic.co/elastic-agent/elastic-agent"
 
 elk_config = Munch()
 elk_config.user = os.getenv("ES_USER", "NA")
@@ -40,6 +41,10 @@
 
 kspm_config = Munch()
 kspm_config.docker_image_override = os.getenv("DOCKER_IMAGE_OVERRIDE", "")
+if not kspm_config.docker_image_override and elk_config.stack_version.endswith("SNAPSHOT"):
+    kspm_config.docker_image_override = f"{DOCKER_ARTIFACTORY_AGENT}:{elk_config.stack_version}"
+    if elk_config.agent_version:
+        kspm_config.docker_image_override = f"{DOCKER_ARTIFACTORY_AGENT}:{elk_config.agent_version}"
 
 aws_config = Munch()
 aws_config.access_key_id = os.getenv("AWS_ACCESS_KEY_ID", "NA")