Clean and simple containerized Apache Cassandra cluster for local testing. A modern alternative to ccm.
This approach is based on official image (maintained by Docker Hub ), but still lets you control all the config files without a need to build a custom image. Which is described in more details in this blog post
./setup-config.sh
docker-compose up -d
# WARNING:
# ./setup-config.sh doesn't work on MacOS atm, because Mac version of docker command line doesn't allow passing variables
# in order to fix it, replace all instances of ${CASSANDRA_VERSION} in the script with an explicit version number
# works fine on Linux
# Check the cluster status
docker exec cass1 nodetool status
- That will bring up a 3 node Cassandra cluster. You can configure the number of Cassandra containers in docker-compose.yml
- Cassandra data is stored under
./data/on the host and kept even if the cluster is destroyed - Cassandra configuration files are under
./etc/. - You can access the cluster from your app on
localhost:9042 - Or can add a container with your app to docker-compose.yml
If you want to start fresh again (no data, vanilla config):
sudo rm -r ./data/* ./etc/*
docker-compose down
docker exec cass1 nodetool ...
docker exec cass1 cqlsh -e "DESCRIBE keyspaces"
# Or in interactive shell mode
docker exec -it cass1 cqlsh
docker logs -f cass1
Edit docker-compose.yml for your needs, keeping in mind the following:
- An official docker image for cassandra must be used
- You should specify an explicit version tag
- Each cassandra container needs to have two volumes configured, one for data and one for config files
- Some basic Cassandra configuration can be done through
CASSANDRA_environment variables - Check the syntax with
docker-compose configbefore running./setup-config.sh
If you need more advanced configuration, you can edit config files (eg. cassandra.yaml) for each node under etc/<node>/ and simply run
docker-compose restart
A good example can be found in this blog post