The recommended way of deploying Featuris is via a Docker image on Kubernetes (k8s) since that’s how we’re deploying it. The following deployment methods are covered:
The Docker image for Featuris resides at https://hub.docker.com/r/zephinzer/featuris/.
The Docker image for Featuris resides at https://hub.docker.com/r/zephinzer/featuris/.
The k8s-featuris.yaml assumes you’ve already got a k8s cluster up and running with your credentials saved into kubectl.
Replace the variable names denoted by __double_underscores__ before and after, with your own values and run kubectl apply -n %NAMESPACE% -f ./path/to/k8s-featuris.yaml.
Create a new directory and place your feature manifests in it. In the same directory, create a docker-compose.yml and add the following contents:
version: "3"
services:
# ...
# other services
# ...
Featuris:
image: zephinzer/featuris:latest
volumes:
- ./:/app/data/features
expose:
- "3000"
ports:
- 3000:3000
environment:
- NODE_ENV=production
- LOGS_DISABLED="false"
- LOGS_FORMAT="json"
# ...
Then run:
docker-compose -f ./docker-compose.yml up
Create a new directory and place your feature manifests in it.
Run the following from the directory, which pulls in the Docker image from Docker Hub and uses it to instantiate a container that is running in "production" with logs enabled (the -e flags).
docker pull zephinzer/featuris:latest
docker run \
-e "NODE_ENV=production"
-e "LOGS_DISABLED=false"
-v "$(pwd)":/app/data/features
-p 3000:3000
zephinzer/featuris:latest;
You should be able to check Featuris is up by going to http://localhost:3000.
The -v flag indicates that the folder you are at, should be used as the directory for /app/data/features, which means the current directory you are at with feature manifests will be mapped to the container’s /app/data/features.
Lastly the -p flag indicates to expose port 3000 on the host (first ‘3000’) from the container’s port 3000 (second ‘3000’)
To make it run in the background, add a -d flag and to make sure it is running, use:
docker ps
Sometimes it is nice to be able to bundle your configuration into an image which you can deploy.
Like usage via pulling from Docker Hub, create a directory and place your feature manifests into it. Next, create a new Dockerfile with the following contents:
FROM zephinzer/featuris:latest
ENV NODE_ENV=__preffered_environment__ \
LOGS_FORMAT=__format_of_logs__
COPY . /app/data/features
EXPOSE 3000
VOLUME ["/app/data/features"]
ENTRYPOINT ["npm", "start"]
Replace all variables (surrounded by double underscores __like_this__) with your own configurations. You can build and push your own application to a private registry using:
REPO_PATH="your_org/featuris";
VERSION=__version__;
docker build -f ./Dockerfile -t ${REPO_PATH}:${VERSION} .
docker login nexus-docker.yourdomain.com -uusername -ppassword;
docker tag ${REPO_PATH}:${VERSION} ${REPO_PATH}:latest;
docker push ${REPO_PATH}:${VERSION};
docker push ${REPO_PATH}:latest;;
docker logout
Clone this repository into a server of your choice.
Run yarn install to install the dependencies.
Two template configurations for running Featuris are available at provisioning/ecosystem.yaml. If you need your own, make a copy of the template configurations and add your environment variables in them.
Start Featuris by running:
pm2 start ./provisioning/ecosystem.yaml --only __APP_NAME_IN_ECOSYSTEM_FILE__.
Read more on pm2 at their official documentation.
Need further configurations? Check out the section on how to configure Featuris.