Deploying to a cleared local hardhat node should ignore previous deployment

[kanej]

Ignition treats deploying to a local Hardhat network node (chainId: 31337) like other networks. It creates deployment files under the ./ignition/deployments folder.

However, if the Hardhat node is stopped and restarted, then the deployment state will need to be wiped to allow for a redeployment of the module.

It would be better if, when deploying against a network with chainId 31337, Ignition automatically detects that the network node has restarted since the previous run. If the network node has restarted, the deployment should be reset with a complete wipe of the previous deployment state.

A potential approach

Pull the Hardhat node instance id if the network has chainId 31337:

const metadata = await network.provider.send("hardhat_metadata");
const instanceId = metadata.instanceId; // random identifier of the session

Then to record that instance id against the deployment initialize message to the journal

export interface DeploymentInitializeMessage {
  type: JournalMessageType.DEPLOYMENT_INITIALIZE;
  chainId: number;
  instanceId?: number;
}

On a rerun of the deployment, Ignition should compare the instance ids, if they differ, the deployment state should be cleared.

[alcuadrado]

This is a great idea!