Kanvas Api

Kanvas Ecosystem API powered by PhalconPHP

[

Installation

• Clone the project
• Copy storage/ci/.env.prod and paste it in the root of the project and rename it .env
• On phalcon-api/.env in MYSQL_ROOT_PASSWORD and DATA_API_MYSQL_PASS assign the root password for MySQL.
• On phalcon-api/.env, update MySQL credentials (DATA_API_MYSQL_NAME,DATA_API_MYSQL_USER,DATA_API_MYSQL_PASS)
• On phalcon-api/.env, change DATA_API_MYSQL_HOST = localhost to DATA_API_MYSQL_HOST = mysql
• Run Docker containers with the docker-compose up --build command
• After the build, access the project main container with docker exec -it id_of_docker_container sh
• Inside the container's console run get inside the apps folder, cd app/
• Copy storage/ci/phinx.php to phinx.php
• Copy storage/ci/phinx-kanvas.php to phinx-kanvas.php
• To finish the setup run ./runCli setup start this will run migration, seed and acl

NOTE : This requires docker to be present in your system. Visit their site for installation instructions.

CREATE NEW APP

• To create a new app run ./runCli setup newapp {{AppName}}

CLI

• Clear model and temp cache ./app/php cli/cli.php clearcache
• Update db migartion ./app/vendor/bin/phinx migrate -e production
• Clear old Sessions on db ./app/php cli/cli.php clearcache sessions

QUEUES

The Kanvas Core uses RabbitMQ to manage our queue process. Internally we handle 3 queue jobs to start

• php cli/cli.php queue jobs
• php cli/cli.php queue events
• php cli/cli.php queue notifications

Jobs : will handle normal Jobs run on any moment during the runtime of the app

Events : will handle events we run send to the queue $this->events->fireToQueue('user:test', Users::findFirst(), ['test'])

Notifications : will handle notifications we send to the queue Users::findFirst(18)->notify(new CanvasSubscription(Companies::findFirst(10)))

ACL

By Default the Canvas will assign all register user the Admin role but if you want to define a specific roles , you will need to add to your app settings

defaultAdminRole : App.RoleName

Requests

Ecosystem

When working with other local apps we have created a docker network called canvas_network , this will allow other local ecosystem apps to connect to it if needed

Add to your local docker-compose file on the app network

  my-proxy-net:
    external:
      name: canvas_network

And on your contianer network info

    networks:
      - local-network
      - my-proxy-net

TEST

• Inside the container's console run ./vendor/bin/codecept run

CI/CD

There's a CI/CD already defined for this api that is composed by:

1. Helm Chart templates
2. GitHub Actions Pipeline

The technologies selected to work fine with this CI/CD are:

1. EKS (AWS's Kubernetes)
2. GitHub/GitHub Actions
3. AWS CLI
4. AWS ECR (container registry )
5. Bash Scripting (so the agent worker must be Unix)

The steps of this pipeline are programed inside the repo at ./.github/workflows/actions.yml

This pipeline needs to setup some spesific vars before run:

At GitHub secrets scope:

Go to the Settings/Secrets section of your GitHub Repo and configure the following secret vars:

AWS_ACCESS_KEY_ID= Access key of the aws account with access rigths to the EKS cluster

AWS_SECRET_ACCESS_KEY= Secret of the Access Key ID

DEVELOPMENT_VARS= All content of your .env file filled with the access and configurations needed by the api (you can find an example at the repo ./.env.example

At ./.github/workflows/actions.yml file scope:

Must replace the var cluster_namein the actions.yaml file with the name of your api's destination cluster that is in the AWS org of your configured aws account.

Replace account_id: ******** at "AWS ECR" step with the account ID of your Access KEY

Replace <account_id> inside repo name (and region if necesary) --set apiImage=<account_id>.dkr.ecr.us-east-1.amazonaws.com/${{ env.instance_name }}:latest you'll find that line at "Deploy Helm chart" step.