Career Network

Development

Ensure node is installed
It is suggested to install an eslint plugin for your chosen editor

Clone repo and run (specifying the desired environment, if not dev1):

npm install
cd functions
npm install
cp .runtimeconfig.json-example .runtimeconfig.json
cd ..
npx firebase login
npm run env:dev1
npm run dev

Enter your personal secret Airtable API key into functions/.runtimeconfig.json (from https://airtable.com/account)

Enter your Intercom test identity verification secret into functions/.runtimeconfig.json

See for identity verification secret info: https://www.intercom.com/help/en/articles/183-enable-identity-verification-for-web-and-mobile

Enter LinkedIn App ID and secret into functions/.runtimeconfig.json

Testing

Tests reside in the __tests__ folder and use Jest as the main testing framework, as well as React Testing Library.

The project contains both unit and integration tests, which can be run separately.

To only run unit tests, execute:

npm run test:unit

The integration tests depend on the Firestore emulator. So, before attempting to run them, you must start an emulator instance on your local machine:

npx firebase emulators:start --only firestore

Now, you can launch the integration tests by running:

npm run test:integration

If you just want to run all tests, use:

npm run test

Deployment

Clone repo and install:

npm install

Specify the Firebase project to which you would like to deploy (ppe shown here as an example):

npm run env:ppe

Set production Intercom identity verification secret (or test secret if not deploying to production):

npx firebase functions:config:set intercom.identity_verification_secret="IDENTITY VERIFICATION SECRET"

Set Airtable API key (generally only needed in dev, since npm run airtable:dump should be used to pull data from Airtable and commit to public/static/api/ for production deployments). This is generally only needed if you want to test new Airtable content locally or dump Airtable content to the persisted data files using the aforementioned command.

npx firebase functions:config:set airtable.api_key="AIRTABLE API KEY"

Set

See for identity verification secret info: https://www.intercom.com/help/en/articles/183-enable-identity-verification-for-web-and-mobile

(Optional) Preview the exported static site locally:

The LinkedIn Auth uses the default service accounts for signing custom tokens, if that account don't have the iam.serviceAccounts.signBlob permission, you may use IAM and admin of the Google Cloud Platform Console to grant it with the necessary permissions.

npm run preview

If all looks good, ship it:

npm run deploy

Upload activities data

The recommended activities templates/data are stored in activity-templates/templates/ directory.

In order to view these activities on this site, you need to upload the activity templates to the firestore activityTemplates collection by running node ./scripts/updateActivityTemplates.js. You can update templates in the collection by running the same command.

You can change the firebase environment you are updating by editing the info found in the firebaseConfig for updateActivityTemplates.js. You will most likely have to generate a private key and download a serviceAccount.json.

DNS

DNS is maintained by Terraform in the /terraform directory.

For details, see: https://developers.cloudflare.com/terraform/

In a nutshell (API token created at https://dash.cloudflare.com/profile/api-tokens):

export CLOUDFLARE_API_TOKEN=your-api-token
cd terraform
terraform init
terraform plan
terraform apply

New method: DNS is now managed through GCP.

Get a service account key:

https://console.cloud.google.com/iam-admin/serviceaccounts?project=nj-career-network
Find "terraform" service account
Create a new key for the "terraform" service account
Move the downloaded key to terraform/credentials
Update google.tf to reflect the filename of your key (if many people are managing this someday, maybe we instead just each export our key paths to GOOGLE_CLOUD_KEYFILE_JSON).

Then the same as above:

cd terraform
terraform init
terraform plan
terraform apply

You must manually configure the DC records (DNSSEC) created by GCP in your registrar: https://console.cloud.google.com/net-services/dns/zones/njcareers-org?project=nj-career-network

Updating API snapshots

In production, configuration data from Airtable is read from static JSON files that live in this repo rather than directly from the Airtable API (for stability reasons and better versioning of the content).

To update these local JSON files, run:

npm run dev

Then, in a separate shell:

npm run airtable:dump

To update the live content, commit the changes and deploy normally.

Scheduled Firestore backups

We've added a mechanism that simplifies creating daily backups of the Firestore database. It works by registering a job in Google Cloud Scheduler that is executed every 24 hours. This job has an associated cloud function that performs the actual export and saves the results into a Google Storage bucket.

Here are a few caveats worth having in mind:

All collections are backed up, and only for the default database ((default)).
The target bucket where the exported documents will be stored needs to follow this convention: <PROJECT-ID>-backups. So, for example, in the case of the nj-career-network-dev2 project, the expected bucket name will be nj-career-network-dev2-backups. Don't forget to create it before attempting to deploy the function.
The firestoreBackup cloud function reads the current project from the GCLOUD_PROJECT environment variable that is automatically set by the service. This means it's not tied to any specific Firebase project.
You'll need to check your IAM permissions in the GCP Console to grant access to the default service account for App Engine. More specifically, you need to add the role Cloud Datastore Import Export Admin and Storage Admin; otherwise you'll receive a permission error when the scheduled function tries to perform the export. See more: https://firebase.google.com/docs/firestore/solutions/schedule-export
It's advisable to set up error notifications through GCP Stackdriver, to be informed of any execution issuess: https://console.cloud.google.com/errors/ (visit this page after deploying the function to enable alerts).

In order to deploy the function, you can run npm run deploy from the root project folder (or from within the functions folder, to only deploy the functions and not the rest of the project). It should take care of building and preparing the code for deployment as well as setting up the scheduler and the necessary permissions in your account. You might want to view the function in the Firebase "Functions" UI, follow the link to "View in Cloud Scheduler," and manually run it a couple times to verify proper execution (logs in Firebase UI console).

Big Thanks

Cross-browser Testing Platform and Open Source ❤️ provided by Sauce Labs

newjersey/career-network