Accessing Accumulo 2.0.0 using a Docker Container

Question

I have an Accumulo 2.0.0 instance up and running using the UNO build process. I have run the helloworld jar from the accumulo-examples to write 50K records to the Hadoop tables.

I am looking for an example of a docker container that can be used to access my Accumulo instance to read the information from that instance. Does anyone know of an example of an existing container with the API calls to Accumulo 2.0?

I have found discussions about running Accumulo within a container. I am trying to keep Accumulo running outside of a container and access the instance from a docker container.

Thanks

Answer 1

Uno is a tool for running things on a single node for testing and development purposes. Running the client in another node, whether Docker or otherwise, is simply not the intended use case Uno was written to satisfy.

Additionally, please be aware that Uno is unreleased software at the Apache Software Foundation, and using it is not recommended by the ASF or the Fluo PMC for users outside the development team at this time. See this policy on releases at the ASF for more information.

All that said, you may be able to run an Accumulo client inside a Docker container (assuming you have all the class path items to run a client there). However, you will probably need to set a specific hostname in your uno.conf file (UNO_HOST) instead of localhost. If you use a DNS name and IP address that is reachable from within the Docker container, it should work with an appropriate Docker networking configuration.

I am not an expert on Docker networking, but a quick internet search resulted in this documentation.

Further details regarding Uno should be available on the Fluo mailing lists from Uno's software developers.