cubicweb-fresh 1.5.2

expense tracking application built on the CubicWeb framework

Expense tracking application built on the CubicWeb framework.

Developping with docker

Assuming you have a postgresql running on your machine where you can connect with peer authentication, run “make dev” it will spawn an interactive shell inside a docker container with the code mounted in develop mode. It mean you can edit the code locally and run it in the container.

Some useful commands:

• cubicweb-ctl db-create -a fresh will create and initialize the database

• cubicweb-ctl pyramid -D -l info fresh will start the instance on http://localhost:8080

Deploying on kubernetes

The file deployment.yaml contains several containers split in several deployments:

• nginx to serve static files directly

• fresh to run the application (and compile the latest translation)

• an initContainers upgrade before fresh that upgrade the database schema if there is a new fresh version.

• fresh-scheduler to have the scheduler running.

To create the initial database from an existing empty database:

kubectl run -it fresh-dbcreate \
   --env CW_DB_HOST=db \
   --env CW_DB_USER=user \
   --env CW_DB_PASSWORD=pass \
   --env CW_DB_NAME=fresh \
   --image=hub.extranet.logilab.fr/logilab/fresh --command -- \
   cubicweb-ctl db-create --automatic --create-db=n fresh
kubectl delete deployment fresh-dbcreate

Then generate a secret named “fresh” from where environment variables are set:

kubectl create secret generic fresh-env \
   --from-literal CW_DB_HOST=db
   --from-literal CW_DB_USER=user \
   --from-literal CW_DB_PASSWORD=pass \
   --from-literal CW_DB_NAME=fresh \
   --from-literal CW_BASE_URL=https://fresh.example.com

You need to mount a cwclientlib configuration file to make CWClientLibDataFeedParser. If you don’t use this feature, just create an empty file.

Create a file named cwclientlibrc and run:

kubectl create secret generic fresh-cwclientlibrc --from-file=./cwclientlibrc

Create a persistent volume for bfss data:

kubectl apply -f deploy/pvc.yaml

Then deploy fresh with:

kubectl apply -f deploy/deployment.yaml