Local Contexts Hub Main Repository

Live Prod Link:

• Local Contexts Hub

Hub Sandbox Link:

• Local Contexts Hub Sandbox

Develop Live Link:

• Develop

Data sources:

• Data used by Local Contexts Hub
• ROR API

Getting Started

What you need to run this app:

• virtualenv
• Django 3.1
• PostGreSQL 15+
• Python 3.8+ with the latest version of pip

Local Deployment

First Time: set up database

NOTE: See section below on Heroku vs. Local Deploymet and setting environment variables

psql> create database localcontextsdb # create the database

Setup virtualenvironment

pyenv virtualenv 3.8.3 envi

Activate virtualenvironment:

source envi/bin/activate

Install or Update Dependencies

Once in virtualenv, in the project root run: pip install -r requirements.txt

Initialize Environment Variables

Set localcontexts Environment Variables using EXPORT Statements This will set local environment variables accessible via PYTHON

source env-local.sh sets up environment variables for locally hosted database source env-localprod.sh sets up environment variables for production database to access it locally source env-localdev.sh sets up environment variables for development/testing database to access it locally source env-localstage.sh sets up environment variables for staging database to access it locally

Migration

python manage.py makemigrations
python manage.py migrate # sets up database
python manage.py createsuperuser # so you can start using database
  Username: <Fill in your username>
  Email address: *<INSERT SITE_ADMIN_EMAIL> that is specified in the previous step, 'Initialize Environment Variables'*

Production Maintenance Mode

A superuser can enable / disable maintenance mode by going to /maintenance-mode/on or /maintenance-mode/off which will show the 503.html template to every user on the site letting users know site is under construction. Superuser will still have access to /admin and is the only user that can still make changes to the database at this time. Maintenance mode should be turned on for production deplyment and turned off when deployment is complete.

Production Deployment and Migration

Every push to master will automatically trigger a new version build. Migrations should be done before pushing to master!! In order to make migrations or migrate the production database:

1. source env-localprod.sh
2. CLOUD PROXY CONNECTION STRING
3. python manage.py makemigrations
4. python manage.py migrate
5. Merge develop into master

Running the Server

python manage.py runserver

Locally, the app will start on port 8000.

Configuration can be found in the settings.py file. Database models are located in /<app name>/models.py.

GCP Deployment

Prerequisites

1. Install gcloud utility and login user appropriate account: gcloud auth login --update-adc.

2. Configure the default GCP project: gcloud set project <projectId>.

3. Download Cloud SQL proxy.

4. Start the proxy: cloud_sql_proxy --instances=<cloud-sql-connection-name>=tcp:5432.

   You can find the connection name at the Cloud SQL instance overview page. It has the following structure: <gcp-project>:<region>:<instance-name> e.g. biocode-localcontests-db:us-central1:biocode-db.

Now you're able to start the local env but point to the Cloud DB. This may be useful if migrations should be applied.

Configuring GCP environment

The application is deployed into App Engine Standard serverless platform and uses managed Postgres database via Cloud SQL deployment.

Prior to deploying the app one must go over the following steps in order to set up and configure the GCP project.

1. Configure the GCP project ID to be used by the gcloud: gcloud config set <project-id>.

2. Create App Engine app within the current GCP project and region.

   There may be only one app per GCP project and only inside one region so pick the app region wisely. You will not be able to change it afterwards.

   You may use gcloud app regions list to list all the available App Engine regions.

   export REGION="us-central1"
   gcloud app create --region="${REGION}"

   One may also create a new App via the web console UI here.

3. Create a new Cloud SQL Postgres instance.

   In order to create a new SQL instance it is recommended to use the web UI wizard here.

   The wizard default should cover some more-or-less generic workloads, but it may be a good idea to check out all the configuration options.

   After the instance creation navigate to the instance dashboard and copy the connection name.

4. Create Postgres DB.

   In order to create a new DB one may use the following command, or the Cloud SQL web UI.

   export DB_NAME="localcontextsdb"
   export INSTANCE="biocode-localcontests-db-stage:us-central1:biocode-db" # Change to real instance name
   gcloud sql databases create "${DB_NAME}" --instances="${INSTANCE}"

5. Migrate the DB.

   When the infrastructure is up it should be a good idea to roll out all the Django DB migrations as well as to create a Django superuser.

   To roll out the migrations do the following.

   • Start the Cloud SQL proxy: cloud_sql_proxy --instances=<cloud-sql-connection-name>=tcp:5432.
   • Make the migrations: python manage.py makemigrations.
   • Apply the migrations: python manage.py migrate

   To create a superuser run: python manage.py createsuperuser.

Deploying the app

When the initial setup is done the app can be deployed with a single command: gcloud app deploy.

By default, the deployment promotes the newly deployed version and routes all the traffic to it.

In order to perform a deployment without routing the traffic add --no-promote option to the deployment command: gcloud app deploy --no-promote.

App Engine creates a new version for each deployment. The version name is the deployment date and time. In order to supply a custom version name use --version option: gcloud app deploy --version="2021-02-06-v14".

Setting Up Git Pre-Commit Hook

At the project root, run this command to setup the GIT pre-commit hook. This lets our codebase run formatting and linting before completing a commit. Also, whenever the scripts/pre-commit.sh file is updated, the command below must be rerun for the pre-commit hook to be updated.

cp -p scripts/pre-commit.sh .git/hooks/pre-commit