filecoin-station / spark-api

💥 SPARK API
https://github.com/filecoin-station/spark
Other
3 stars 1 forks source link

spark-api

SPARK API

CI

Routes

POST /retrievals

Start a new retrieval.

Body:

{
  sparkVersion: String,
  zinniaVersion: String
}

Response:

```typescript
{
  id: String,
  cid: String,
  providerAddress: String,
  protocol: 'graphsync'|'bitswap'|'http'
}

PATCH /retrievals/:id

Parameters:

Body:

{
  participantAddress: String,
  timeout: Boolean,
  startAt: String,       // ISO 8601
  statusCode: Number,
  firstByteAt: String,   // ISO 8601
  endAt: String,         // ISO 8601
  byteLength: Number,
  attestation: String,
  stationId: String
}

Dates should be formatted as ISO 8601 strings.

Response:

OK

GET /miner/:minerId/deals/eligible/summary

Parameters:

Response:

Number of deals grouped by client IDs.

{
  "minerId": "f0814049",
  "dealCount": 13878,
  "clients": [
    { "clientId": "f02516933", "dealCount": 6880 },
    { "clientId": "f02833886", "dealCount": 3126 }
  ]
}

GET /client/:clientId/deals/eligible/summary

Parameters:

Response:

Number of deals grouped by miner IDs.

{
  "clientId": "f0215074",
  "dealCount": 38977,
  "providers": [
    { "minerId": "f01975316", "dealCount": 6810 },
    { "minerId": "f01975326", "dealCount": 6810 }
  ]
}

GET /allocator/:allocatorId/deals/eligible/summary

Parameters:

Response:

Number of deals grouped by client IDs.

{
  "allocatorId": "f03015751",
  "dealCount": 4088,
  "clients": [
    { "clientId": "f03144229", "dealCount": 2488 },
    { "clientId": "f03150656", "dealCount": 1600 }
  ]
}

Development

Database

Set up PostgreSQL with default settings:

Alternatively, set the environment variable $DATABASE_URL with postgres://${USER}:${PASS}@${HOST}:${POST}/${DATABASE}.

The Postgres user and database need to already exist, and the user needs full management permissions for the database.

You can also the following command to set up the PostgreSQL server via Docker:

docker run -d --name spark-db \
  -e POSTGRES_HOST_AUTH_METHOD=trust \
  -e POSTGRES_USER=$USER \
  -e POSTGRES_DB=$USER \
  -p 5432:5432 \
  postgres

When working on multiple Spark-related services, we recommend to use the following commands to create or reset the Postgres instance:

docker rm -f meridian-db && docker run --name meridian-db -e POSTGRES_HOST_AUTH_METHOD=trust -e POSTGRES_USER=$USER -e POSTGRES_DB=$USER -p 5432:5432 -d postgres && sleep 1; psql postgres://localhost:5432/ -c "CREATE DATABASE spark_evaluate" && psql postgres://localhost:5432/ -c "CREATE DATABASE spark_stats" && psql postgres://localhost:5432/ -c "CREATE DATABASE spark"

api

Start the API service:

npm start --workspace api

Run tests and linters:

npm test --workspace api
npm run lint --workspace api

Deployment

Pushes to main will be deployed automatically.

Perform manual devops using Fly.io:

$ fly deploy api