Add the option to start Hadoop docker container when running integration tests (#9513)

* hadoop docker it

* hadoop docker container it

* fix hadoop container
This commit is contained in:
Maytas Monsereenusorn 2020-03-16 12:04:05 -07:00 committed by GitHub
parent e7b3dd9cd1
commit 09600db8f2
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 56 additions and 4 deletions

View File

@ -23,7 +23,7 @@ LC_ALL=C.UTF-8
# JAVA OPTS
COMMON_DRUID_JAVA_OPTS=-Duser.timezone=UTC -Dfile.encoding=UTF-8 -Dlog4j.configurationFile=/shared/docker/lib/log4j2.xml
DRUID_DEP_LIB_DIR=/shared/docker/lib/*:/usr/local/druid/lib/mysql-connector-java.jar
DRUID_DEP_LIB_DIR=/shared/hadoop_xml/*:/shared/docker/lib/*:/usr/local/druid/lib/mysql-connector-java.jar
# Druid configs
druid_auth_authenticator_basic_authorizerName=basic
@ -62,3 +62,4 @@ druid_zk_service_host=druid-zookeeper-kafka
druid_auth_basic_common_maxSyncRetries=20
druid_indexer_logs_directory=/shared/tasklogs
druid_sql_enable=true
druid_extensions_hadoopDependenciesDir=/shared/hadoop-dependencies

View File

@ -266,6 +266,7 @@
<profile>
<id>integration-tests</id>
<properties>
<start.hadoop.docker>false</start.hadoop.docker>
<override.config.path></override.config.path>
</properties>
<build>
@ -282,6 +283,7 @@
<phase>pre-integration-test</phase>
<configuration>
<environmentVariables>
<DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER>${start.hadoop.docker}</DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER>
<DRUID_INTEGRATION_TEST_JVM_RUNTIME>${jvm.runtime}</DRUID_INTEGRATION_TEST_JVM_RUNTIME>
<DRUID_INTEGRATION_TEST_GROUP>${groups}</DRUID_INTEGRATION_TEST_GROUP>
<DRUID_INTEGRATION_TEST_OVERRIDE_CONFIG_PATH>${override.config.path}</DRUID_INTEGRATION_TEST_OVERRIDE_CONFIG_PATH>

View File

@ -16,7 +16,7 @@
# Cleanup old images/containers
{
for node in druid-historical druid-coordinator druid-overlord druid-router druid-router-permissive-tls druid-router-no-client-auth-tls druid-router-custom-check-tls druid-broker druid-middlemanager druid-zookeeper-kafka druid-metadata-storage;
for node in druid-historical druid-coordinator druid-overlord druid-router druid-router-permissive-tls druid-router-no-client-auth-tls druid-router-custom-check-tls druid-broker druid-middlemanager druid-zookeeper-kafka druid-metadata-storage druid-it-hadoop;
do
docker stop $node
docker rm $node
@ -29,6 +29,7 @@
{
# environment variables
DIR=$(cd $(dirname $0) && pwd)
HADOOP_DOCKER_DIR=$DIR/../examples/quickstart/tutorial/hadoop/docker
DOCKERDIR=$DIR/docker
SERVICE_SUPERVISORDS_DIR=$DOCKERDIR/service-supervisords
ENVIRONMENT_CONFIGS_DIR=$DOCKERDIR/environment-configs
@ -45,6 +46,8 @@
cp -r client_tls docker/client_tls
# Make directories if they dont exist
mkdir -p $SHARED_DIR/hadoop_xml
mkdir -p $SHARED_DIR/hadoop-dependencies
mkdir -p $SHARED_DIR/logs
mkdir -p $SHARED_DIR/tasklogs
@ -53,6 +56,12 @@
cp -R docker $SHARED_DIR/docker
mvn -B dependency:copy-dependencies -DoutputDirectory=$SHARED_DIR/docker/lib
# Pull Hadoop dependency if needed
if [ -n "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" ] && [ "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" == true ]
then
java -cp "$SHARED_DIR/docker/lib/*" -Ddruid.extensions.hadoopDependenciesDir="$SHARED_DIR/hadoop-dependencies" org.apache.druid.cli.Main tools pull-deps -h org.apache.hadoop:hadoop-client:2.8.5 -h org.apache.hadoop:hadoop-aws:2.8.5
fi
# install logging config
cp src/main/resources/log4j2.xml $SHARED_DIR/docker/lib/log4j2.xml
@ -114,6 +123,12 @@ else
esac
fi
# Build Hadoop docker if needed
if [ -n "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" ] && [ "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" == true ]
then
docker build -t druid-it/hadoop:2.8.5 $HADOOP_DOCKER_DIR
fi
# Start docker containers for all Druid processes and dependencies
{
@ -149,5 +164,39 @@ fi
# Start Router with custom TLS cert checkers
docker run -d --privileged --net druid-it-net --ip 172.172.172.12 ${COMMON_ENV} ${ROUTER_CUSTOM_CHECK_TLS_ENV} ${OVERRIDE_ENV} --hostname druid-router-custom-check-tls --name druid-router-custom-check-tls -p 8891:8891 -p 9091:9091 -v $SHARED_DIR:/shared -v $SERVICE_SUPERVISORDS_DIR/druid.conf:$SUPERVISORDIR/druid.conf --link druid-zookeeper-kafka:druid-zookeeper-kafka --link druid-coordinator:druid-coordinator --link druid-broker:druid-broker druid/cluster
}
# Start Hadoop docker if needed
if [ -n "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" ] && [ "$DRUID_INTEGRATION_TEST_START_HADOOP_DOCKER" == true ]
then
# Start Hadoop docker container
docker run -d --privileged --net druid-it-net --ip 172.172.172.13 -h druid-it-hadoop --name druid-it-hadoop -p 2049:2049 -p 2122:2122 -p 8020:8020 -p 8021:8021 -p 8030:8030 -p 8031:8031 -p 8032:8032 -p 8033:8033 -p 8040:8040 -p 8042:8042 -p 8088:8088 -p 8443:8443 -p 9000:9000 -p 10020:10020 -p 19888:19888 -p 34455:34455 -p 49707:49707 -p 50010:50010 -p 50020:50020 -p 50030:50030 -p 50060:50060 -p 50070:50070 -p 50075:50075 -p 50090:50090 -p 51111:51111 -v $SHARED_DIR:/shared druid-it/hadoop:2.8.5 sh -c "/etc/bootstrap.sh && tail -f /dev/null"
# wait for hadoop namenode to be up
echo "Waiting for hadoop namenode to be up"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -mkdir -p /druid"
while [ $? -ne 0 ]
do
sleep 2
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -mkdir -p /druid"
done
echo "Finished waiting for Hadoop namenode"
# Setup hadoop druid dirs
echo "Setting up druid hadoop dirs"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -mkdir -p /druid"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -mkdir -p /druid/segments"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -mkdir -p /quickstart"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -chmod 777 /druid"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -chmod 777 /druid/segments"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -chmod 777 /quickstart"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -chmod -R 777 /tmp"
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -chmod -R 777 /user"
# Copy data files to Hadoop container
docker exec -t druid-it-hadoop sh -c "./usr/local/hadoop/bin/hdfs dfs -put /shared/wikiticker-it/wikiticker-2015-09-12-sampled.json.gz /quickstart/wikiticker-2015-09-12-sampled.json.gz"
echo "Finished setting up druid hadoop dirs"
echo "Copying Hadoop XML files to shared"
docker exec -t druid-it-hadoop sh -c "cp /usr/local/hadoop/etc/hadoop/*.xml /shared/hadoop_xml"
echo "Copied Hadoop XML files to shared"
fi
}

View File

@ -14,7 +14,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.
for node in druid-historical druid-coordinator druid-overlord druid-router druid-router-permissive-tls druid-router-no-client-auth-tls druid-router-custom-check-tls druid-broker druid-middlemanager druid-zookeeper-kafka druid-metadata-storage;
for node in druid-historical druid-coordinator druid-overlord druid-router druid-router-permissive-tls druid-router-no-client-auth-tls druid-router-custom-check-tls druid-broker druid-middlemanager druid-zookeeper-kafka druid-metadata-storage druid-it-hadoop;
do
docker stop $node