service-inspector

🔎 Microservice Inspector 🔎

Purpose

To push information about Microservices to AirTable after CI deployment

Design goals

• Microservices should themselves describe what they are, should not be perscriptive about how they self identify
• Should be able to be installed and ran without a package.json (through npx), should be able to run in ruby projects just with a nodejs dependency

Get up and running

Service inspector uses a plugin based architecture and runs every JavaScript file in a directory specifically called inspector-plugins within the current working directory

Plugins must expose a run method and return an object which matches the following format:

const run = () => {
  return {
    columnName: 'Last Synced',
    columnValue: Date.now(),
  }
};

module.exports = { run };

Once set up you can run to see what the outputs will be:

npx service-inspector print

To update AirTable with those values run:

npx service-inspector update --airtable-api-key <API-KEY> --airtable-base-id appnEws1PUhekz5jh  --airtable-table-name Microservices --airtable-row-id recTrwiBViF1PS6vm

Improvements

• Output data which was pushed to AirTable (Print command)
• Debug statements
• Tests

Unknowns

• How do we know which values are actively being updated?
• How should we manage replacing values when there is something on air table but not something locally?
• How should I manage dependencies? -- currently they're coming from the inspector project though this feels kind of wrong