Orakl Network VRF

Description

The Orakl Network VRF is one of the main Orakl Network solutions. It provides an access to provably random number generator.

The code is located under core directory, and separated to three independent microservices: listener, worker and reporter.

State Setup

The Orakl Network VRF requires an access to state of listeners and VRF keys.

Listener

The Orakl Network API holds information about all listeners. The command below adds a single VRF listener to the Orakl Network state to listen on vrfCoordinatorAddress for RandomWordsRequested event. The chain parameter specifies a chain on which we expect to operate with the Orakl Network VRF Listener.

orakl-cli listener insert \
    --service VRF \
    --chain ${chain} \
    --address ${vrfCoordinatorAddress} \
    --eventName RandomWordsRequested

• example

orakl-cli listener insert --service VRF --chain baobab --address 0xDA8c0A00A372503aa6EC80f9b29Cc97C454bE499 --enventName RandomWordsRequested

Reporter

The Orakl Network API holds information about all reporters. The command below adds a single VRF reporter to the Orakl Network state to report to oracleAddress. The chain parameter specifies a chain on which we expect to operate. Reporter is defined by an address and a privateKey parameters.

orakl-cli reporter insert \
  --service VRF \
  --chain ${chain} \
  --address  ${address} \
  --privateKey ${privateKey} \
  --oracleAddress ${oracleAddress}

• example

orakl-cli reporter insert \
  --service VRF \
  --chain baobab \
  --address  0x12 \
  --privateKey abc \
  --oracleAddress 0xDA

VRF Keys

To be able to run VRF as a node operator, one must have registered VRF keys in VRFCoordinator, and VRF keys has to be in Orakl Network state as well. VRF worker will load them from the Orakl Network API when it is launched.

If you do not have VRF keys, you can generate them with the Orakl Network CLI using the following command.

orakl-cli vrf keygen

The output of generated command will be similar to the one below, but including the keys on the right side of the keys (sk, pk, pkX,pkY, and keyHash). VRF keys are generated randomly, therefore every time you call the keygen command, you receive a different output. sk represents a secret key which is used to generate the VRF beta and pi. This secret key should never be shared with anybody except the required personnel.

sk=
pk=
pkX=
pkY=
keyHash=

To store VRF keys in Orakl Network state use orakl-cli vrf insert command. Parameter --chain corresponds to the network name to which VRF keys will be associated.

orakl-cli vrf insert \
    --chain ${chain} \
    --pk ${pk} \
    --sk ${sk} \
    --pkX ${pkX} \
    --pkY ${pkY} \
    --keyHash ${keyHash}

Configuration

Before we launch the Orakl Network VRF, we must specify several environment variables. The environment variables are automatically loaded from a .env file.

• NODE_ENV=production

• CHAIN

• PROVIDER_URL

• ORAKL_NETWORK_API_URL

• LOG_LEVEL

• REDIS_HOST

• REDIS_PORT

• HEALTH_CHECK_PORT

• SLACK_WEBHOOK_URL

The Orakl Network VRF is implemented in Node.js which uses NODE_ENV environment variable to signal the execution environment (e.g. production, development). Setting the environment to production generally ensures that logging is kept to a minimum, and more caching levels take place to optimize performance.

CHAIN environment variable specifies on which chain the Orakl Network VRF will be running, and which resources will be collected from the Orakl Network API.

PROVIDER_URL defines an URL string representing a JSON-RPC endpoint that listener and reporter communicate through.

ORAKL_NETWORK_API_URL corresponds to url where the Orakl Network API is running. The Orakl Network API interface is used to access Orakl Network state such as listener and VRF key configuration.

Setting a level of logs emitted by a running instance is set through LOG_LEVEL environment variable, and can be one of the following: error, warning, info, debug and trace, ordered from the most restrictive to the least. By selecting any of the available options you subscribe to the specified level and all levels with lower restrictiveness.

REDIS_HOST and REDIS_PORT represent host and port of Redis to which all Orakl Network VRF microservices connect. The default values are localhost and 6379, respectively.

The Orakl Network VRF does not offer a rich REST API, but defines a health check endpoint (/) served under a port denoted as HEALTH_CHECK_PORT.

Errors and warnings emitted by the Orakl Network VRF can be sent to Slack channels through a slack webhook. The webhook URL can be set with the SLACK_WEBOOK_URL environment variable.

Launch

Before launching the VRF solution, the Orakl Network API has to be accessible from the Orakl Network VRF to load VRF keys, and listener settings.

After the Orakl Network API is healthy, launch the VRF service, which consists of listener, worker, and reporter microservices, with the command below. Microservices communicate with each other through the BullMQ - job queue.

yarn start:core:vrf

Run in dev mode through the following command:

yarn dev:core:vrf

It's also possible to run the microservices separately in any arbitrary order:

yarn start:listener:vrf
yarn start:worker:vrf
yarn start:reporter:vrf

Quick launch with Docker

From orakl repository's root, run the following command to build all images:

docker-compose -f docker-compose.local-core.yaml build

Set wallet credentials, ADDRESS and PRIVATE_KEY values, in the .core-cli-contracts.env file. Keep in mind that the default chain is localhost. If changes are required, update CHAIN (other options being baobab and cypress) and PROVIDER_URL values. Note that if the chain is not localhost, Coordinator and Prepayment contracts won't be deployed. Instead, Bisonai's already deployed contract addresses will be used. After setting the appropriate .env values, run the following command to start the VRF service:

SERVICE=vrf docker-compose -f docker-compose.local-core.yaml up --force-recreate

Note that the current docker implementation is designed to run a single service, either rr or vrf, at a time. Therefore, it's highly recommended to add --force-recreate when running docker-compose up command. That will restart all containers thus removing all the modified data in those containers.

Here is what happens after the above command is run:

• api, postgres, redis, and json-rpc services will start as separate docker containers

• postgres will get populated with necessary data:

  • chains

  • services

  • vrf keys

  • listener (after contracts are deployed)

  • reporter (after contracts are deployed)

• migration files in contracts/v0.1/migration/ get updated with provided keys and other values

• if the chain is localhost:

  • contracts/v0.1/hardhat.config.cjs file gets updated with PROVIDER_URL

  • relevant coordinator and prepayment contracts get deployed

Architecture

Orakl Network VRF