CI: retry docker containers startup

.circleci/config.yml

@@ -85,7 +85,7 @@ jobs:
       - load_cache
       - run:
           name: Create Kafka Environment
-          command: ./docker/startup.sh
+          command: ./docker/startup_retry.sh
       - run:
           name: Integration Tests
           command: mvn integration-test
@@ -122,7 +122,7 @@ jobs:
           command: mvn package -Ddistributed
       - run:
           name: Create Kafka Environment
-          command: ./docker/startup.sh
+          command: ./docker/startup_retry.sh
       - run:
           name: Integration Tests
           command: mvn integration-test -Ddistributed -Darango.topology=$STARTER_MODE
@@ -149,7 +149,7 @@ jobs:
           command: mvn package -Ddistributed
       - run:
           name: Create Kafka Environment
-          command: ./docker/startup.sh
+          command: ./docker/startup_retry.sh
       - run:
           name: Integration Tests
           command: mvn integration-test -Ddistributed -DSslTest=true -Dit.test=com.arangodb.kafka.SslIT
@@ -173,7 +173,7 @@ jobs:
       - load_cache
       - run:
           name: Create Kafka Environment
-          command: ./docker/startup.sh
+          command: ./docker/startup_retry.sh
       - run:
           name: Start Toxiproxy
           command: ./bin/startProxy.sh

docker/start_db.sh

@@ -55,14 +55,14 @@ if [ "$EXTENDED_NAMES" == "true" ]; then
 fi
 
 # data volume
-docker create -v /data --name arangodb-data alpine:3 /bin/true
-docker cp "$LOCATION"/jwtSecret arangodb-data:/data
-docker cp "$LOCATION"/server.pem arangodb-data:/data
+docker create -v /data --name adb-data alpine:3 /bin/true
+docker cp "$LOCATION"/jwtSecret adb-data:/data
+docker cp "$LOCATION"/server.pem adb-data:/data
 
 docker run -d \
     --name=adb \
     -p 8528:8528 \
-    --volumes-from arangodb-data \
+    --volumes-from adb-data \
     -v /var/run/docker.sock:/var/run/docker.sock \
     -e ARANGO_LICENSE_KEY="$ARANGO_LICENSE_KEY" \
     $STARTER_DOCKER_IMAGE \

docker/start_kafka_zk.sh

@@ -11,7 +11,7 @@ ZK_DOCKER_IMAGE=docker.io/bitnami/zookeeper:3.8
 docker pull $ZK_DOCKER_IMAGE
 
 docker run -d \
-  --name zk-1 -h zk-1 \
+  --name kafka-zk-1 -h kafka-zk-1 \
   --network arangodb \
   -e BITNAMI_DEBUG=true \
   -e ALLOW_ANONYMOUS_LOGIN="yes" \
@@ -22,7 +22,7 @@ docker run -d \
   --network arangodb \
   -p 19092:19092 \
   -e BITNAMI_DEBUG=true \
-  -e KAFKA_CFG_ZOOKEEPER_CONNECT=zk-1 \
+  -e KAFKA_CFG_ZOOKEEPER_CONNECT=kafka-zk-1 \
   -e ALLOW_PLAINTEXT_LISTENER="yes" \
   -e KAFKA_CFG_LISTENERS=PLAINTEXT://:9092,PLAINTEXT_HOST://0.0.0.0:19092 \
   -e KAFKA_CFG_ADVERTISED_LISTENERS="PLAINTEXT://kafka-1:9092,PLAINTEXT_HOST://172.28.0.1:19092" \
@@ -35,7 +35,7 @@ docker run -d \
   --network arangodb \
   -p 29092:29092 \
   -e BITNAMI_DEBUG=true \
-  -e KAFKA_CFG_ZOOKEEPER_CONNECT=zk-1 \
+  -e KAFKA_CFG_ZOOKEEPER_CONNECT=kafka-zk-1 \
   -e ALLOW_PLAINTEXT_LISTENER="yes" \
   -e KAFKA_CFG_LISTENERS=PLAINTEXT://:9092,PLAINTEXT_HOST://0.0.0.0:29092 \
   -e KAFKA_CFG_ADVERTISED_LISTENERS="PLAINTEXT://kafka-2:9092,PLAINTEXT_HOST://172.28.0.1:29092" \
@@ -48,7 +48,7 @@ docker run -d \
   --network arangodb \
   -p 39092:39092 \
   -e BITNAMI_DEBUG=true \
-  -e KAFKA_CFG_ZOOKEEPER_CONNECT=zk-1 \
+  -e KAFKA_CFG_ZOOKEEPER_CONNECT=kafka-zk-1 \
   -e ALLOW_PLAINTEXT_LISTENER="yes" \
   -e KAFKA_CFG_LISTENERS=PLAINTEXT://:9092,PLAINTEXT_HOST://0.0.0.0:39092 \
   -e KAFKA_CFG_ADVERTISED_LISTENERS="PLAINTEXT://kafka-3:9092,PLAINTEXT_HOST://172.28.0.1:39092" \

docker/start_schema_registry.sh

@@ -6,7 +6,7 @@ docker pull $DOCKER_IMAGE
 KAFKA_BOOTSTRAP_SERVERS=PLAINTEXT://kafka-1:9092,PLAINTEXT://kafka-2:9092,PLAINTEXT://kafka-3:9092
 
 docker run -d \
-  --name schema-registry -h schema-registry \
+  --name kafka-schema-registry -h kafka-schema-registry \
   --network arangodb \
   -p 8081:8081 \
   -e SCHEMA_REGISTRY_KAFKASTORE_BOOTSTRAP_SERVERS="$KAFKA_BOOTSTRAP_SERVERS" \

docker/startup_retry.sh

@@ -0,0 +1,18 @@
+#!/bin/bash
+
+# exit when any command fails
+set -e
+
+until ./docker/startup.sh
+do
+    # rm all containers, in case of retries
+    # retries might be needed in case of port-mapping errors in CircleCI, e.g. for errors like:
+    # docker: Error response from daemon: driver failed programming external connectivity on endpoint kafka-3:
+    # Error starting userland proxy: listen tcp4 0.0.0.0:39092: bind: address already in use.
+
+    docker ps -a -f name=kafka-.* -q | while read x ; do docker rm -f $x ; done
+    docker rm -f adb
+    docker ps -a -f name=adb-.* -q | while read x ; do docker rm -f $x ; done
+    echo "Startup failed, retrying in 10 seconds..."
+    sleep 10
+done

src/connect-distributed/java/spi/ClusterKafkaConnectDeployment.java

@@ -57,7 +57,7 @@ public KafkaConnectOperations client() {
 
     @Override
     public String getSchemaRegistryUrlConnect() {
-        return "http://schema-registry:8081";
+        return "http://kafka-schema-registry:8081";
     }
 
     @Override