lapax/riak repository overview

⁠docker-riak

This is based on dockerfile and scripts from https://github.com/hectcastro/docker-riak⁠

The changes for my needs were significant enough to not make it a true fork..

Changes from hectcastro/docker-riak

• enable search

• use leveldb instead of bitcask backend

• enable cluster admin control panel with user and password

• removed dependency on pipeworks

• uses docker native networking with icc enabled (inter container comm)

• riak node names are set up based on ip address -- hackish but works

• the cluster startup scripts uses data mounts from the host into the containers

• start-cluster and test-cluster takes an optional numeric argument for number of nodes - defaults to 5

• uses dynamic dns to update an external dyndns server from docker events - see docker_dns file

⁠major features retained

• use of supervisord

• use of ssh

• scripts and makefile

⁠possible todos

• use link names

⁠Installation

⁠standalone (single container)

docker pull lapax/riak
docker run -d -t -i \
	-h "riak1" \
	-v /path/to/local/directory:/var/lib/riak \
	-name "riak1" \
	"lapax/riak"

or just

docker run -d -t -i lapax/riak

⁠Cluster installation

⁠Install Docker

If you're running Ubuntu, use the instructions for installing Docker on Linux⁠.

If you're not on a Ubuntu host, use Vagrant⁠ to spin up a Ubuntu virtual machine with Docker installed⁠.

Then, login to the virtual machine:

$ vagrant ssh

⁠Install dependencies

Once you're on a Ubuntu machine, install the following dependencies:

$ sudo apt-get install -y git curl make sshpass

⁠Clone repository

$ git clone https://github.com/lapax/dockerfile-riak.git
$ cd docker-riak
$ make
$ make riak-container

⁠Launch cluster

$ make start-cluster 3

⁠Test cluster

$ make test-cluster 3

⁠Admin cluster

log in via browser to https://<nodeip>:8069/admin with default user name password (admin / adminpass)

⁠Tear down cluster

$ make stop-cluster

⁠Troubleshooting

Spinning up Docker containers consumes memory. If the memory allocated to your Ubuntu [virtual] machine is not adaquate, make start-cluster will fail with something like:

runtime: panic before malloc heap initialized
fatal error: runtime: cannot allocate heap metadata