Collective Service Server

Backend Server for Collective Service Project

Sync Filter Options

Whenever there is an update in AWS RDS Production Database (e.g. addition of Indicators), the Filter Options need to be synced. This can be achieved by login into the Django Admin Panel and click on the button SYNC FILTER OPTIONS on the top right.

Deployment

• Staging
  • Login to the VM.

      # Update client and server
      cd ~/services/client
      git pull
      
      cd ~/services/server
      git pull
      
      # Update docker containers
      cd ~/services/
      docker-compose up --build -d

• Production
  • The Github Actions deployment pipeline is triggered whenever changes are pushed to the branch release.
  • Requires approval from the administrator.
  • These actions update the resources in AWS.

Access AWS ECS Container

• Environment
  • Linux or MacOS
• Installations
  • AWS CLI: Install latest version of AWS command line tool
  • Session Manager: Install latest version of Session Manager plugin for AWS CLI
• Setup AWS Credentials
  • Ask AWS Admin to create a user for you and give you the credentials to login to AWS
  • Run the command:

  • $ aws configure

  • Provide the credentials
  • Check in the directory if the setup is done correctly.

  • $ cd ~/.aws (Two files must be created viz. config and credentials)

  • Note that if you don't define the profile, a default profile will be created.
• Login to AWS
  • Login to the AWS Console
  • Ask the Admin to assign ECS access and other ECS policies in the IAM.
  • Navigate to the Elastic Container Service(ECS)
  • In the Clusters tab, note the few names:
    • Cluster Name: e.g. cluster-backend-prod
    • Task Definition: e.g. backend-task-prod:3
    • Service Name: e.g. backend-service-prod
    • Task ID: e.g. 4d038f255f2f41da9eer186a1b976xxx
    • Container Name: e.g. backend-container-prod
  • Once those information is gathered, run the following command in the terminal.

  • aws ecs --profile <profile> update-service --cluster <cluster_name> --task-definition <task_def_name> --enable-execute-command --service <service_name> --desired-count 1

  • Then, try to access the container using the following command in your terminal.

  • aws ecs --profile <profile> execute-command --cluster <cluster_name> --task <task_id> --container <container_name> --interactive --command "/bin/sh"

  • If for some reason you don't have access to the container, you need remove/delete/stop the task within the Cluster. After that, a new task will be created automatically (just wait for a couple of minutes until the new container status reaches RUNNING)
  • Then, repeat above steps.
  • Once you have access to the container, run commands for maintenance or debug.