Skip to main content
Version: 0.6.1-incubating

User Docker images

Apache Gravitino Docker image​

You can deploy the service with the Gravitino Docker image.

Container startup commands

docker run --rm -d -p 8090:8090 -p 9001:9001 apache/gravitino:0.6.1-incubating

Changelog

  • apache/gravitino:0.6.1-incubating

    • Based on Gravitino 0.6.1-incubating, you can know more information from 0.6.1-incubating release notes.

  • apache/gravitino:0.6.0-incubating (Switch to Apache official DockerHub repository)

    • Use the latest Gravitino version 0.6.0-incubating source code to build the image.

  • datastrato/gravitino:0.5.1

    • Based on Gravitino 0.5.1, you can know more information from 0.5.1 release notes.

  • datastrato/gravitino:0.5.0

    • Based on Gravitino 0.5.0, you can know more information from 0.5.0 release notes.

  • datastrato/gravitino:0.4.0

    • Based on Gravitino 0.4.0, you can know more information from 0.4.0 release notes.

  • datastrato/gravitino:0.3.1

    • Fix some issues

  • datastrato/gravitino:0.3.0

    • Docker image datastrato/gravitino:0.3.0
    • Gravitino Server
    • Expose ports:
      • 8090 Gravitino Web UI
      • 9001 Iceberg REST service

Apache Gravitino Iceberg REST server Docker image​

You can deploy the standalone Gravitino Iceberg REST server with the Docker image.

Container startup commands

docker run --rm -d -p 9001:9001 apache/gravitino-iceberg-rest:0.6.1-incubating

Changelog

  • apache/gravitino-iceberg-rest:0.6.1-incubating

    • Based on Gravitino 0.6.1-incubating, you can know more information from 0.6.1-incubating release notes.

  • apache/gravitino-iceberg-rest:0.6.0-incubating.

    • Gravitino Iceberg REST Server with memory catalog backend.
    • Expose ports:
      • 9001 Iceberg REST service

Playground Docker image​

You can use the playground to experience the whole Gravitino system with other components.

The playground consists of multiple Docker images.

The Docker images of the playground have suitable configurations for users to experience.

Apache Hive image​

Changelog

  • apache/gravitino-playground:hive-2.7.3 (Switch to Apache official DockerHub repository)

    • Use datastrato/hive:2.7.3-no-yarn Dockerfile to rebuild the image.

  • datastrato/hive:2.7.3-no-yarn

    • Docker image datastrato/hive:2.7.3-no-yarn
    • hadoop-2.7.3
    • hive-2.3.9
    • Don't start YARN when container startup

Trino image​

Changelog

  • apache/gravitino-playground:trino-435-gravitino-0.6.1-incubating

    • Use Gravitino release 0.6.1-incubating Dockerfile to build the image.

  • apache/gravitino-playground:trino-435-gravitino-0.6.0-incubating (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/trino:435-gravitino-0.5.1

    • Based on Gravitino 0.5.1, you can know more information from 0.5.1 release notes.

  • datastrato/trino:426-gravitino-0.5.0

    • Based on Gravitino 0.5.0, you can know more information from 0.5.0 release notes.

  • datastrato/trino:426-gravitino-0.4.0

    • Based on Gravitino 0.4.0, you can know more information from 0.4.0 release notes.

  • datastrato/trino:426-gravitino-0.3.1

    • Fix some issues

  • datastrato/trino:426-gravitino-0.3.0

    • Docker image datastrato/trino:426-gravitino-0.3.0
    • Base on trino:426
    • Added Gravitino trino-connector-0.3.0 libraries into the /usr/lib/trino/plugin/gravitino

Developer Docker images

You can use these kinds of Docker images to facilitate integration testing of all catalog and connector modules within Gravitino.

Apache Gravitino CI Apache Hive image with kerberos enabled​

You can use this kind of image to test the catalog of Apache Hive with kerberos enable

Changelog

  • apache/gravitino-ci:kerberos-hive-0.1.5 (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/gravitino-ci-kerberos-hive:0.1.5

    • Start another HMS for the Hive cluster in the container with port 19083. This is to test whether Kerberos authentication works for a Kerberos-enabled Hive cluster with multiple HMS.
    • Refresh ssh keys in the startup script.
    • Add test logic to log in localhost via ssh without password.

  • datastrato/gravitino-ci-kerberos-hive:0.1.4

    • Increase the total check time for the status of DataNode to 150s.
    • Output the log of the DataNode fails to start

  • datastrato/gravitino-ci-kerberos-hive:0.1.3

    • Add more proxy users in the core-site.xml file.
    • fix bugs in the start.sh script.

  • datastrato/gravitino-ci-kerberos-hive:0.1.2

    • Add ${HOSTNAME} >> /root/.ssh/known_hosts to the startup script.
    • Add check for the status of DataNode, if the DataNode is not running or ready within 100s, the container will exit.

  • datastrato/gravitino-ci-kerberos-hive:0.1.1

    • Add a principal for Gravitino web server named 'HTTP/localhost@HADOOPKRB'.
    • Fix bugs about the configuration of proxy users.

  • datastrato/gravitino-ci-kerberos-hive:0.1.0

    • Set up a Hive cluster with kerberos enabled.
    • Install a KDC server and create a principal for Hive. For more please see kerberos-hive

Apache Gravitino CI Apache Hive image​

You can use this kind of image to test the catalog of Apache Hive.

Changelog

  • apache/gravitino-ci:hive-0.1.13 (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/gravitino-ci-hive:0.1.13

    • Support Hive 2.3.9 and HDFS 2.7.3
      • Docker environment variables:
        • HIVE_RUNTIME_VERSION: hive2 (default)
    • Support Hive 3.1.3, HDFS 3.1.0 and Ranger plugin version 2.4.0
      • Docker environment variables:
        • HIVE_RUNTIME_VERSION: hive3
        • RANGER_SERVER_URL: Ranger admin URL
        • RANGER_HIVE_REPOSITORY_NAME: Hive repository name in Ranger admin
        • RANGER_HDFS_REPOSITORY_NAME: HDFS repository name in Ranger admin
      • If you want to enable Hive Ranger plugin, you need both set the RANGER_SERVER_URL and RANGER_HIVE_REPOSITORY_NAME environment variables. Hive Ranger audit logs are stored in the /tmp/root/ranger-hive-audit.log.
      • If you want to enable HDFS Ranger plugin, you need both set the RANGER_SERVER_URL and RANGER_HDFS_REPOSITORY_NAME environment variables. HDFS Ranger audit logs are stored in the /usr/local/hadoop/logs/ranger-hdfs-audit.log
      • Example: docker run -e HIVE_RUNTIME_VERSION='hive3' -e RANGER_SERVER_URL='http://ranger-server:6080' -e RANGER_HIVE_REPOSITORY_NAME='hiveDev' -e RANGER_HDFS_REPOSITORY_NAME='hdfsDev' ... datastrato/gravitino-ci-hive:0.1.13

  • datastrato/gravitino-ci-hive:0.1.12

    • Shrink hive Docker image size by 420MB

  • datastrato/gravitino-ci-hive:0.1.11

    • Remove yarn from the startup script; Remove yarn-site.xml and yarn-env.sh files;
    • Change the value of mapreduce.framework.name from yarn to local in the mapred-site.xml file.

  • datastrato/gravitino-ci-hive:0.1.10

    • Remove SSH service from the startup script.
    • Use hadoop-daemon.sh to start HDFS services.

  • datastrato/gravitino-ci-hive:0.1.9

    • Remove cache after installing packages.

  • datastrato/gravitino-ci-hive:0.1.8

    • Change the value of hive.server2.enable.doAs to true

  • datastrato/gravitino-ci-hive:0.1.7

    • Download MySQL JDBC driver before building the Docker image
    • Set hdfs as HDFS superuser group

  • datastrato/gravitino-ci-hive:0.1.6

    • No starting YARN when container startup
    • Removed expose ports:
      • 22 SSH
      • 8088 YARN Service

  • datastrato/gravitino-ci-hive:0.1.5

    • Rollback Map container hostname to 127.0.0.1 before starting Hadoop of datastrato/gravitino-ci-hive:0.1.4

  • datastrato/gravitino-ci-hive:0.1.4

    • Configure HDFS DataNode data transfer address to be 0.0.0.0:50010
    • Map the container hostname to 127.0.0.1 before starting Hadoop
    • Expose 50010 port for the HDFS DataNode

  • datastrato/gravitino-ci-hive:0.1.3

    • Change MySQL bind-address from 127.0.0.1 to 0.0.0.0
    • Add iceberg to MySQL users with password iceberg
    • Export 3306 port for MySQL

  • datastrato/gravitino-ci-hive:0.1.2

    • Based on datastrato/gravitino-ci-hive:0.1.1
    • Modify fs.defaultFS from local to 0.0.0.0 in the core-site.xml file.
    • Expose 9000 port in the Dockerfile file.

  • datastrato/gravitino-ci-hive:0.1.1

    • Based on datastrato/gravitino-ci-hive:0.1.0
    • Modify HDFS/YARN/HIVE MaxPermSize from 8GB to 128MB
    • Modify HADOOP_HEAPSIZE from 8192 to 128

  • datastrato/gravitino-ci-hive:0.1.0

    • Docker image datastrato/gravitino-ci-hive:0.1.0
    • hadoop-2.7.3
    • hive-2.3.9
    • Expose ports:
      • 22 SSH
      • 9000 HDFS defaultFS
      • 50070 HDFS NameNode
      • 50075 HDFS DataNode HTTP server
      • 50010 HDFS DataNode data transfer
      • 8088 YARN Service
      • 9083 Hive metastore
      • 10000 HiveServer2
      • 10002 HiveServer2 HTTP

Apache Gravitino CI Trino image​

You can use this image to test Trino.

Changelog

  • apache/gravitino-ci:trino-0.1.6 (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/gravitino-ci-trino:0.1.6

    • Upgrade trino:426 to trino:435

  • datastrato/gravitino-ci-trino:0.1.5

    • Add check for the version of gravitino-trino-connector

  • datastrato/gravitino-ci-trino:0.1.4

    • Change -Xmx1G to -Xmx2G in the config file /etc/trino/jvm.config

  • datastrato/gravitino-ci-trino:0.1.3

    • Remove copy content in folder gravitino-trino-connector to plugin folder /usr/lib/trino/plugin/gravitino

  • datastrato/gravitino-ci-trino:0.1.2

    • Copy JDBC driver 'mysql-connector-java' and 'postgres' to /usr/lib/trino/iceberg/ folder

  • datastrato/gravitino-ci-trino:0.1.0

    • Docker image datastrato/gravitino-ci-trino:0.1.0
    • Based on trinodb/trino:426 and removed some unused plugins from it.
    • Expose ports:
      • 8080 Trino JDBC port

Apache Gravitino CI Doris image​

You can use this image to test Apache Doris.

Changelog

  • apache/gravitino-ci:doris-0.1.5 (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/gravitino-ci-doris:0.1.5

    • Remove the chmod command in the Dockerfile to decrease the size of the Docker image.

  • datastrato/gravitino-ci-doris:0.1.4

    • remove chmod in start.sh to accelerate the startup speed

  • datastrato/gravitino-ci-doris:0.1.3

    • To adapt to the CI framework, don't exit container when start failed, logs are no longer printed to stdout.
    • Add report_disk_state_interval_seconds config to decrease report interval.

  • datastrato/gravitino-ci-doris:0.1.2

    • Add a check for the status of Doris BE, add retry for adding BE nodes.

  • datastrato/gravitino-ci-doris:0.1.1

    • Optimize start.sh, add disk space check before starting Doris, exit when FE or BE start failed, add log to stdout

  • datastrato/gravitino-ci-doris:0.1.0

    • Docker image datastrato/gravitino-ci-doris:0.1.0
    • Start Doris BE & FE in one container
    • Please set table properties "replication_num" = "1" when creating a table in Doris, because the default replication number is 3, but the Doris container only has one BE.
    • Username: root, Password: N/A (password is empty)
    • Expose ports:
      • 8030 Doris FE HTTP port
      • 9030 Doris FE MySQL server port

Apache Gravitino CI Apache Ranger image​

You can use this image to control Trino's permissions.

Changelog

  • apache/gravitino-ci:ranger-0.1.1 (Switch to Apache official DockerHub repository)

    • Use Gravitino release 0.6.0 Dockerfile to build the image.

  • datastrato/gravitino-ci-ranger:0.1.1

    • Docker image datastrato/gravitino-ci-ranger:0.1.1
    • Use ranger-admin release from datastrato/apache-ranger:2.4.0 to build docker image.
    • Remove unnecessary hack in start-ranger-service.sh.
    • Reduce docker image build time from ~1h to ~5min.
    • How to debug Ranger admin service:
      • Use docker exec -it <container_id> bash to enter the docker container.
      • Add these context export JAVA_OPTS=-agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=*:5001 into /opt/ranger-admin/ews/webapp/WEB-INF/classes/conf/ranger-admin-env-debug.sh in the docker container.
      • Execute ./opt/ranger-admin/stop-ranger-admin.sh and ./opt/ranger-admin/start-ranger-admin.sh to restart Ranger admin.
      • Clone the Apache Ranger project from GiHub and checkout the 2.4.0 release.
      • Create a remote debug configuration (Use model classpath = EmbeddedServer) in your IDE and connect to the Ranger admin container.

  • datastrato/gravitino-ci-ranger:0.1.0

    • Docker image datastrato/gravitino-ci-ranger:0.1.0
    • Support Apache Ranger 2.4.0
    • Use environment variable RANGER_PASSWORD to set up Apache Ranger admin password, Please notice Apache Ranger Password should be minimum 8 characters with min one alphabet and one numeric.
    • Expose ports:
      • 6080 Apache Ranger admin port