OlympusDAO Treasury Subgraph

Purpose

It proved to be enormously difficult to aggregate GraphQL results across multiple endpoints and date ranges (due to limitations with Graph Protocol pagination). The code in this monorepo seeks to address that.

Architecture

The monorepo contains two components:

apps/
- server/
  - Code to configure and deploy a Wundergraph API server that aggregates and transforms GraphQL results
  - Offers a number of 'operations' (in Wundergraph terminology):
    - Earliest: the earliest record from each blockchain
      - earliest/metrics
      - earliest/tokenRecords
      - earliest/tokenSupplies
    - Latest: the latest record from each blockchain
      - latest/metrics
      - latest/tokenRecords
      - latest/tokenSupplies
    - Paginated: records from the startDate input variable until the present. Aggregates all blockchains and handles pagination. Specify crossChainDataComplete to restrict results up to the latest date with cross-chain data. Specify includeRecords to include records used to calculate each metric (large response size).
      - paginated/metrics
      - paginated/tokenRecords
      - paginated/tokenSupplies
- client/
  - Generates a React-compatible client that can be used in olympus-frontend
  - Publishes the client to NPM

The API server is currently hosted on Google Cloud Run, with Firebase Hosting to provide a static domain name.

HTTP-layer caching is provided by Google Cloud Run. An attempt was made to utilise caching in Redis (on Upstash), but it performed slower than the OOTB HTTP-layer caching. It can be enabled through the CACHE_ENABLED environment variable.

Developer Tasks

The repo is setup using turbo to make handling tasks easier.

Setup

Run yarn in the root directory.
Copy .env.sample to .env and replace any required values

Building

During local development, you can trigger a build with yarn build:local. This requires the setup tasks to have been completed.

yarn build requires an environment variable to be set, and is used in the Wundergraph Cloud deploy process.

Running

During local development, you can run an API endpoint locally with yarn server:start.

This requires environment variables to be set, so follow the instructions in setup.

Unit Tests

Run yarn test:local.

This requires environment variables to be set, so follow the instructions in setup.

Testing the Olympus Frontend

Running the frontend against a different API endpoint requires jumping through some (small) hoops:

Run the API endpoint locally. See Running.
Pass the API endpoint to the frontend: VITE_WG_PUBLIC_NODE_URL=http://localhost:9991 yarn start

Deployment - Wundergraph Server

Orchestration is performed using Pulumi. To deploy, follow these steps:

Change to the apps/server directory.
Authenticate with Pulumi, using pulumi login
Run pulumi up --stack <dev | prod>

NOTE: the Upstash credentials in the production project and environment should be different to that of all other projects/environments, so that the production cache is not polluted.

Deployment - Client NPM Package

Set the required values in .env.prod
Authenticate with the NPM package registry: npm login
Update the version in apps/client/package.json
Update the changelog
Run the following command: yarn publish-package
- Prefix the command with YARN_OTP=<OTP VALUE>

NOTE: You must be a member of the @olympusdao org in NPM in order to publish.

Wishlist / TODO

When testing/developing new subgraph versions, it would be nice to be able to provide a URL parameter with the subgraph deployment ID and have the Typescript operation use that deployment ID instead of the configured data source.

OlympusDAO/treasury-subgraph