Hadoop, Big Data, NoSQL & DevOps Docker Images

Contains 45 DockerHub repos with 340+ tags, many different versions of standard official open source software, see Full Inventory futher down.

These docker images are tested by hundreds of tools and also used in the full functional test suites of various other GitHub repos.

Overview - this repo contains:

• Hadoop & Big Data ecosystem technologies (Spark, Kafka, Presto, Drill, Nifi, ZooKeeper)
• NoSQL datastores (HBase, Cassandra, Riak, SolrCloud)
• OS & development images (Alpine, CentOS, Debian, Ubuntu)
• DevOps tech, anything cool and open source (RabbitMQ Cluster, Mesos, Consul)
• My GitHub repos containing hundreds of tools related to these technologies with all dependencies pre-built

These images are all available pre-built on My DockerHub - https://hub.docker.com/u/harisekhon/.

Quality and Testing - this repo has entire test suites run against it from various GitHub repositories to validate the docker images' functionality, branches vs tagged versions align, latest contains correct version from master branch, syntax checks against all build files found in the repo covering all common file formats (JSON/INI/Make/XML/YAML configurations) etc.

These are reusable tests that can anybody can implement and can be found in my PyTools and Bash-Tools repos as well as the Advanced Nagios Plugins Collection which contains hundreds of technology specific API-level test programs to ensure the docker images are functioning as intended.

Continuous Integration in run on this and adjacent repos that form a bi-directional validation between these docker images and several other repositories full of hundreds of programs. All of this is intended to keep the quality of this repo as high as possible.

Hari Sekhon

Big Data Contractor, United Kingdom

(ex-Cloudera, Hortonworks Consultant)

https://www.linkedin.com/in/harisekhon

(you're welcome to connect with me on LinkedIn)

Ready to run Docker images

docker search harisekhon
docker run harisekhon/nagios-plugins

To see more than the 25 DockerHub repos limited by docker search (docker issue 23055) I wrote dockerhub_search.py using the DockerHub API, available in my PyTools github repo and as a pre-built docker image:

docker run harisekhon/pytools dockerhub_search.py harisekhon

There are lots of tagged versions of official software in my repos to allow development testing across multiple versions, usually more versions than available from the official repos (and new version updates available on request, just raise a GitHub issue).

DockerHub tags are not shown by docker search (docker issue 17238) so I wrote dockerhub_show_tags.py available in my PyTools github repo and as a pre-built docker image - eg. to see an organized list of all tags for the official CentOS & Ubuntu repos dynamically using the DockerHub API:

docker run harisekhon/pytools dockerhub_show_tags.py centos ubuntu

(you might need to add --timeout 60 as ubuntu has a lot of tags now I've noticed to stop it timing out while fetching all the pages)

For service technologies like Hadoop, HBase, ZooKeeper etc for which you'll also want port mappings, each directory in the GitHub project contains both a standard docker-compose configuration as well as a make run shortcut (which doesn't require docker-compose to be installed) - either way you don't have to remember all the command line switches and port number specifics:

cd zookeeper
docker-compose up

or for technologies with interactive shells like Spark, ZooKeeper, HBase, Drill, Cassandra where you want to be dropped in to an interactive shell, use the make run shortcut instead:

cd zookeeper
make run

which is much easier to type and remember than the equivalent bigger commands like:

docker run -ti -p 2181:2181 harisekhon/zookeeper

which gets much worse for more complex services like Hadoop / HBase:

docker run -ti -p 2181:2181 -p 8080:8080 -p 8085:8085 -p 9090:9090 -p 9095:9095 -p 16000:16000 -p 16010:16010 -p 16201:16201 -p 16301:16301 harisekhon/hbase

docker run -ti -p 8020:8020 -p 8032:8032 -p 8088:8088 -p 9000:9000 -p 10020:10020 -p 19888:19888 -p 50010:50010 -p 50020:50020 -p 50070:50070 -p 50075:50075 -p 50090:50090 harisekhon/hadoop

Full Inventory:

Official Standard Open Source Technologies:

More specific information can be found in the readme page under each respective directory in the Dockerfiles git repo.

• Alluxio - distributed in-memory filesystem for cluster computing frameworks by UC Berkely's AMPLab - readme
• Apache Drill - distributed SQL engine by MapR (opens Drill SQL shell) - readme
• Cassandra - distributed NoSQL datastore by Facebook and DataStax (opens CQL shell, bundled with nagios-plugins)
• Consul - distributed service discovery by HashiCorp
• H2O - distributed machine learning framework by 0xdata
• Hadoop (HDFS + Yarn) - distributed storage and compute cluster by Yahoo, Cloudera and Hortonworks
• HBase - distributed NoSQL datastore by Facebook (opens HBase shell)
• Jython - Python on Java JVM (useful for Hadoop python utilities using Hadoop's Java API. eg. PyTools)
• Kafka - pub-sub data broker by LinkedIn and Confluent
• Mesos - datacenter resource manager by Mesosphere (mostly obsoleted by more free Hortonworks / Hadoop Yarn resource manager)
• Nifi - IOT data flow engine by NSA and Hortonworks
• Presto - distributed SQL engine by Facebook (opens Presto SQL shell) - readme
• Presto (Teradata distribution) - Teradata's Presto distribution including ODBC and JDBC drivers (opens Presto SQL shell) - readme
• RabbitMQ Cluster - pub-sub message queue broker by Pivotal (extension of RabbitMQ official image with added plugins)
• Riak KV - distributed NoSQL datastore by Basho
• Riak KV (bundled with nagios-plugins)
• Serf - decentralized cluster coordination engine by HashiCorp
• Solr - mature indexing engine built on Lucene search library
• SolrCloud - clustered distributed indexing engine version of Solr
• Spark - fast distributed cluster compute engine usually used on Hadoop, by UC Berkely's AMPLab and Databricks (opens Spark shell)
• Superset - data visualization by Airbnb
• Tachyon (Alluxio < 1.0) - distributed in-memory filesystem for cluster computing frameworks by UC Berkely's AMPLab
• ZooKeeper (opens ZK shell) - distributed coordination and sychronization service by Yahoo

Repos suffixed with -dev are the official technologies + development & debugging tools + my github repos with all dependencies pre-built.

My GitHub Repos (with all libs + deps pre-built):

• Advanced Nagios Plugins Collection - 350+ nagios plugins for every Hadoop distribution and every major NoSQL technology - Hadoop, Redis, Elasticsearch, Solr, HBase, Cassandra & DataStax OpsCenter, MongoDB, MySQL, Kafka, Riak, Memcached, Couchbase, CouchDB, Mesos, Spark, Neo4j, Datameer, H2O, WanDisco, Yarn, HDFS, Impala, Apache Drill, Presto, ZooKeeper, Cloudera, Hortonworks, MapR, IBM BigInsights, Infrastructure - Linux, DNS, Whois, SSL Certs etc

  • Tags:
    • nagios-plugins:latest (centos)
    • nagios-plugins:alpine
    • nagios-plugins:centos
    • nagios-plugins:debian
    • nagios-plugins:ubuntu

• Python Tools - 50+ tools for Hadoop, Spark, Pig, Ambari Blueprints, AWS CloudFormation, Linux, Data Converters & Validators (Avro/Parquet/JSON/CSV/XML/YAML), Elasticsearch, Solr, IPython - CLI tools

• Perl Tools - 25+ tools for Hadoop, Hive, Solr, Linux, SQL, Ambari, Datameer, Web and various Linux CLI Tools

• Spotify Tools - Backup & Play Automation: Spotify Lookup - converts Spotify URIs to 'Artist - Track' form by querying the Spotify Metadata API. Spotify Cmd - command line control of Spotify on Mac via AppleScript for automation, auto timed track flick through etc.

• CentOS + all Github repos pre-built

• Debian + all Github repos pre-built

• Ubuntu + all Github repos pre-built

• Alpine + all Github repos pre-built

Base Images:

Dev images:

• CentOS latest with Java JDK, Perl, Python, Jython, Ruby, Scala, Groovy, GCC, Maven, SBT, Gradle, Make, Expect, EPEL etc.
• Debian latest with Java JDK, Perl, Python, Jython, Ruby, Scala, Groovy, GCC, Maven, SBT, Gradle, Make, Expect etc.
• Ubuntu latest with Java JDK, Perl, Python, Jython, Ruby, Scala, Groovy, GCC, Maven, SBT, Gradle, Make, Expect etc.
• Alpine latest with Java JDK, Perl, Python, Jython, Ruby, Scala, Groovy, GCC, Maven, SBT, Gradle, Make, Expect etc.

Base Images of Java / Scala:

All builds use OpenJDK with jre and jdk numbered tags. See this article below for why it might be illegal to bundle Oracle Java (and why no Linux distributions do this either):

https://www.javacodegeeks.com/2016/03/running-java-docker-youre-breaking-law.html

• CentOS latest combinations of Java 7/8 and Scala 2.10/2.11
• Debian latest with Java 7
• Ubuntu 14.04 with Java 7
• Ubuntu latest with Java 8, 9

Build from Source

All images come pre-built on DockerHub but if you want to compile from source for any reason such as developing improvements, I've made this easy to do:

git clone https://github/harisekhon/Dockerfiles

cd Dockerfiles

To build all Docker images, just run the make command at the top level:

make

To build a specific Docker image, enter its directory and run make:

cd nagios-plugins

make

You can also build a specific version by checking out the git branch for the version and running the build:

cd consul
git checkout consul-0.9
make

or build all versions of a given software project like so:

cd hadoop
make build-versions

See the top level Makefile as well as the Makefile.in which is sourced per project with any project specific overrides in the <project_directory>/Makefile.

Support

Please raise tickets for issues and improvements at https://github.com/harisekhon/dockerfiles/issues