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.
Big Data Contractor, United Kingdom
(ex-Cloudera, Hortonworks Consultant)
(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
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 memory base storage by UC Berkely's AMPLab
- Apache Drill - distributed SQL engine by MapR (opens Drill SQL shell)
- 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)
- Presto (Teradata distribution) - Teradata's Presto distribution including ODBC and JDBC drivers (opens Presto SQL shell)
- 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 memory based storage 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, Mesos, Spark, Neo4j, Datameer, H2O, WanDisco, Yarn, HDFS, Impala, Apache Drill, ZooKeeper, Cloudera, Hortonworks, MapR, IBM BigInsights, Infrastructure - Linux, DNS, Whois, SSL Certs etc
- nagios-plugins:latest (centos)
- 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
- 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
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):
- 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:
To build a specific Docker image, enter its directory and run
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
Please raise tickets for issues and improvements at https://github.com/harisekhon/dockerfiles/issues