[FEATURE] More relevant approach to solving the non-responsive server problem

.ci/opensearch/Dockerfile.opensearch

@@ -3,5 +3,11 @@ FROM opensearchproject/opensearch:${OPENSEARCH_VERSION}
 
 ARG opensearch_path=/usr/share/opensearch
 ARG SECURE_INTEGRATION
+ENV SECURE_INTEGRATION=$SECURE_INTEGRATION
 
 RUN if [ "$SECURE_INTEGRATION" != "true" ] ; then $opensearch_path/bin/opensearch-plugin remove opensearch-security; fi
+
+HEALTHCHECK --start-period=20s --interval=5s --retries=5 --timeout=1s \
+  CMD if [ "$SECURE_INTEGRATION" != "true" ]; \
+  then curl -f http://localhost:9200/_cat/health || exit 1; \
+  else curl -fk https://localhost:9200/_cat/health -u admin:admin || exit 1; fi

.github/workflows/test-compatibility.yml

@@ -40,9 +40,12 @@ jobs:
           export OPENSEARCH_VERSION=${{ matrix.entry.opensearch_version }}
           export SECURE_INTEGRATION=${{ matrix.secured }}
           make cluster.clean cluster.build cluster.start
-      - name: Check Opensearch cluster health
-        run: |
-          make cluster.check SECURE_INTEGRATION=${{ matrix.secured }} NUM_CHECKS=30 CHECK_INTERVAL=1
+
+      - uses: zethuman/[email protected]
+        with:
+          name: opensearch_opensearch_1
+          timeout: 300
+          interval: 5
 
       - name: Integration test without security
         if: ${{ matrix.secured == 'false'}}

CHANGELOG.md

@@ -33,6 +33,7 @@ Inspired from [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
 - Adds `advanced_index_actions` guide ([#288](https://github.com/opensearch-project/opensearch-go/pull/288))
 - Adds testcases to check UpdateByQuery functionality ([#304](https://github.com/opensearch-project/opensearch-go/pull/304))
 - Adds additional timeout after cluster start ([##303](https://github.com/opensearch-project/opensearch-go/pull/303))
+- Adds more relevant healthchecker ([#315](https://github.com/opensearch-project/opensearch-go/pull/315))
 
 ### Changed
 

Makefile

@@ -205,37 +205,6 @@ cluster.clean: ## Remove unused Docker volumes and networks
 	docker network prune --force
 	docker system prune --volumes --force
 
-cluster.check:
-	@printf "\033[2m→ Checking up cluster health...\033[0m\n"
-	SECURE_INTEGRATION=$(SECURE_INTEGRATION) \
-    NUM_CHECKS=$(NUM_CHECKS) \
-    CHECK_INTERVAL=$(CHECK_INTERVAL) \
-    echo "SECURE_INTEGRATION: $$SECURE_INTEGRATION"; \
-    for ((i=1; i<=NUM_CHECKS; i++)); do \
-      	if [[ $$SECURE_INTEGRATION == "true" ]]; \
-      	then response=$$(curl --insecure -sS "https://localhost:9200/_cluster/health" -u admin:admin); \
-  		else response=$$(curl -sS "http://localhost:9200/_cluster/health"); \
-  		fi; \
-		if [[  $$response == *"green"* || $$response == *"yellow"* ]]; then \
-			echo "Cluster is healthy."; \
-			exit 0; \
-		fi; \
-		echo "Check #$$i: Cluster is not healthy. Status: $$response"; \
-		if [[ $$i == $$((NUM_CHECKS / 5)) ]]; then \
-			echo "Restarting OpenSearch Container..."; \
-			docker-compose --project-directory .ci/opensearch restart; \
-		fi; \
-		sleep $$CHECK_INTERVAL; \
-	done; \
-	echo "Cluster is still not healthy after $$NUM_CHECKS checks."; \
-	exit 1;
-
-cluster.kek:
-	if [[ "$$SECURE_INTEGRATION" == "false" ]] \
-	then HEALTH_CHECK_URL="https://localhost:9200/_cluster/health" \
-	else HEALTH_CHECK_URL="http://localhost:9200/_cluster/health" \
-	fi; \
-
 linters:
 	./bin/golangci-lint run ./... --timeout=5m