This repo provides a composable set of Docker Compose scripts for running KillrVideo in various configurations
You need to have Docker and docker-compose installed in order to launch the application. Experience with Docker is recommended but not required.
Here's how to run the default configuration, which uses the Java backend
git clone https://github.com/KillrVideo/killrvideo-docker-common.git
cd killrvideo-docker-common
docker-compose up -d
It'll take some time to pull the images and launch the application. After that, these web interfaces should be available:
- http://localhost:3000 - KillrVideo Web Interface
- http://localhost:9091 - DataStax Studio
If you use Docker on mac/windows, docker-machine or similar setup, you may need to replace 'localhost' with the IP address of your docker machine.
It is frequently useful for development and deployment purposes to make changes to this common configuration. This repo also provides a set of override Docker Compose files that operate according to the behaviors described at: https://docs.docker.com/compose/extends/.
The basic way to use this configuration is to do the following:
docker-compose up -d -f docker-compose.yaml <override files>
Alternatively, you can define a .env
file that lists the yaml
files you wish to use, along with any environment
variables needed. For example, the following .env
would override the Java backend with the Python backend:
COMPOSE_PROJECT_NAME=killrvideo
COMPOSE_FILE=./docker-compose.yaml:./docker-compose-backend-python.yaml
The override files override various aspects of the default configuration, for example, specifying different images, or different values for environment variables, and so on. These can be used in various combinations
Quick descriptions of the override files:
docker-compose-backend-external.yaml
- adding this file disables the backend services in the compose so you can run the implementation of your choice locally in your IDE.docker-compose-backend-nodejs.yaml
- adding this file overrides the backend services to the Node.js implementationdocker-compose-backend-python.yaml
- adding this file overrides the backend services to the Python implementationdocker-compose-dse-external.yaml
- adding this file disables the DSE image in the compose (regardless of any additionaldocker-compose-dse-*.yaml
files that may have been specified). If using this make sure to define theKILLRVIDEO_EXTERNAL_IP
environment variabledocker-compose-dse-metrics.yaml
- adding this file enables metrics collection on DSE and runs Prometheus (port X) and Grafana (port X)docker-compose-dse-secure.yaml
- adding this file swaps in DSE configuration files to enable the DSE authentication, authorization and role management modules. If using this compose file, make sure to define theKILLRVIDEO_DSE_USERNAME
andKILLRVIDEO_DSE_PASSWORD
environment variables and setKILLRVIDEO_CREATE_DSE_USER=true
, so that thekillrvideo_dse_config
container will create the user. You may also choose to override the default administrator accountcassandra/cassandra
by settingKILLRVIDEO_CREATE_ADMIN_USER=true
and providing values forKILLRVIDEO_ADMIN_USERNAME
andKILLRVIDEO_ADMIN_PASSWORD
. PLEASE NOTE: if you have previously used thedocker-compose-dse-volumes.yaml
described below, you will want to make sure to delete any existing DSE container and the contents of thedse_data
directory, as the security configuration logic inkillrvideo-dse-config
is designed to be run on a fresh node/cluster.docker-compose-dse-volumes.yaml
- adding this causes DSE to store its data files in the directory dse-data so that you may keep the data even when recreating the container (note warning above underdocker-compose-dse-secure.yaml
)docker-compose-ops-center.yaml
- adding this runs an instance of OpsCenter
To get the effective configuration that results from application of all of your selected config files, use
the command docker-compose config
.
You can deploy your effective Docker configuration to Kubernetes in two steps:
- First, run the command
./kompose-convert.sh
. This will generate resource definitions in thek8s
subdirectory - Then, deploy the configuration by running
kubectl apply -f ./k8s
Optionally, you may wish to delete the contents of the k8s
directory between runs of kompose-convert.sh