Environment Setup

IMPORTANT: The vagrant documentation is not up to date. We have moved everything to a docker-compose setup

Docker-compose setup

1. Make sure you have docker on your machine
2. Clone this project
3. cd into the cloned project
4. Write docker-compose up -d to run the entire system. this starts up a rabbitmq instance, two http-api applications, one to accept student_tester.py requests (helge-api) and one for the frontend to use (backend), a frontend project and a consumer for the rabbitmq

Vagrant

1. Clone project

git clone https://github.com/ProjectHackernewsGroup04/Ops

2. Go to /vm directory

cd vm

3. Start up vagrant

vagrant up

4. SSH into this machine

vagrant ssh

5. Have FUN!

Run docker

1. Inside Vagrant vm

vagrant@vagrant:$ cd /app/

vagrant@vagrant:/app$ docker-compose up --build

The docker-compose.yaml file collects all dockerfile from other reposetories and runs them; sets ports and containers: 1 for backend (backend-app), 1 for frontend (frontend-app), 1 for mongodb (db).

Importing testdata

vagrant@vagrant:/app$ sh database/testdata/import.sh

Test if all containers runs

2. Open new terminal and vagrant ssh
3. Test backend curl http://0.0.0.0:5000
4. Test backend and mongodb ex: curl http://0.0.0.0:5000/api/item/all
5. Test frontend curl http://0.0.0.0:3000

Go to http://localhost:9000/ to watch frontend in action

Enjoy the HackerNews!

CI/CD chain

For CI/CD we're using CircleCI.

CircleCI allows us to version control the deployment process, as it uses simple yaml files to configure the workflow. There's a deployment process for the database, backend and frontend. The backend and frontend builds, tests and deploys the software, while the database project configures a MongoDB instance with Docker and runs it.

All of the services are started using a command like: docker service create --name http-api --publish=5001:5001 --network=hnclone-network -e 'CLOUDAMQP_URL=<CLOUDAMQP_URL>' -e 'BACKEND_URL=http://backend:5000' -d hnclonecphb/http-api:master

After that, every time an update to the master is pushed, it will run a docker service update --image <IMAGE_NAME>:$CIRCLE_WORKFLOW_ID

The CI/CD workflow works like the following:

Deployment Sequence diagram

System Sequence diagram

Operators guideline

Server access

There is a guest user on the server that has access only to docker commands.

The server can be accesses through SSH at the IP address: 46.101.43.21. User and private key can be handed over upon request.

Docker operations

This section is docker specific operation guidelines.

Important operations

• docker service scale <SERVICE>=<NUMBER>
• docker service update <SERVICE>

How to operate

On the server there are multiple docker containers running:

• backend
• frontend
• http-api
• rabbitmq
• rabbitmq-consumer
• database

The names are pretty selfexplanatory if you've read the above documentation.

It's important all containers are up and running at all times. If one of them is down or something is wrong, you can check the logs by running docker service logs -f <container-name>. This will output and follow the logs from that container.

IMPORTANT! THE FOLLOWING SECTION IS DEPRECATED

Important operations

• docker start <container-name>
• docker logs -f <container-name>

How to operate

On the server there are multiple docker containers running:

• backend
• frontend
• http-api
• rabbitmq
• rabbitmq-consumer
• database

The names are pretty selfexplanatory if you've read the above documentation.

It's important all containers are up and running at all times. If one of them is down or something is wrong, you can check the logs by running docker logs -f <container-name>. This will output and follow the logs from that container.

If a docker container is not running, you can start it with docker start <container-name>