How to configure the Data Availability Committee (DAC) in your chain

PUBLIC PREVIEW DOCUMENT

This document is currently in public preview and may change significantly as feedback is captured from readers like you. Click the Request an update button at the top of this document or join the Arbitrum Discord to share your feedback.

AnyTrust chains rely on an external Data Availability Committee (DAC) to store data and provide it on-demand instead of using its parent chain as the Data Availability (DA) layer. The members of the DAC run a Data Availability Server (DAS) to handle these operations. Once the DA servers are running, the chain needs to be configured with their information to effectively store and retrieve data from them.

In this how-to, you'll learn how to configure the DAC in your chain. Refer to the Introduction for the full process of running DA servers and configuring the chain.

This how-to assumes that you're familiar with:

The DAC's role in the AnyTrust protocol. Refer to Inside AnyTrust for a refresher.
Kubernetes. The examples in this guide use Kubernetes to containerize your DAS.
How to deploy a Data Availability Server (DAS). This is needed to understand where the data we'll be handling in this guide comes from.
The Foundry toolkit

Step 0: Prerequisites

Before starting to generate the keyset and configuring the nodes and chain, you'll need to gather the following information from all the DA servers run by the DAC members:

Public BLS Key
URL of the RPC endpoint
URL(s) of the REST endpoint(s)

You should also make sure that at least one DAS is running as an archive DAS, otherwise the information will not be available after the expiry time.

Step 1: Generate the keyset and keyset hash with all the information from the servers

What is a keyset?

The AnyTrust protocol assumes that for the n members of the DAC, a minimum of h members maintain integrity. So h is then the minimum number of trusted committee members on an AnyTrust chain. In scenarios where k = (n + 1) - h members of the DAC pledge to grant access to a specific piece of information, these k members must sign and attest they have stored the data to be considered successful.

To perform this signing operation, each DAC member must generate their own set of BLS public and private keys. They should do this independently and ensure these keys are random and only used by them. You can find more information about how to generate a BLS pair of keys in Generating BLS Keys.

An Anytrust chain needs to know all DAC members' public keys to validate the integrity of the data being batched and posted. A keyset is a list of all DAC members' RPC endpoint and BLS public key. Additionally, it also contains information about how many signatures are needed to approve a Data Availability Certificate (DACert), via a special assumed-honest parameter (i.e., the h parameter we mentioned above). This design lets the chain owner modify the DAC membership over time, and DAC members change their keys if needed. See Inside AnyTrust for more information.

We use this keyset, and its hash to configure the SequencerInbox contract with the valid keyset, and also the batch poster (to request storing information) and full nodes (to request information already stored).

How to generate a keyset and a keyset hash

Nitro comes with a special tool to generate both the keyset and the keyset hash. To use it, you need to first structure the keyset information in a JSON object with the following structure:

{
    "keyset": {
      "assumed-honest": h,
      "backends": [
        {
          "url": "https://rpc-endpoint-of-member-1/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_1"
        },
        {
          "url": "https://rpc-endpoint-of-member-2/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_2"
        },

        ...

        {
          "url": "https://rpc-endpoint-of-member-n/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_N"
        }
      ]
    }
}

The JSON fields represent the following:

assumed_honest is the amount of members that we assume are honest from the n members of the DAC. This is the h variable we mentioned in the previous section.
backends contain information about each member of the DAC:
- url contains the RPC endpoint of the DAS run by that member
- pubkey contains the base64-encoded BLS public key used in the DAS run by that member

Once you have the JSON structure, save it into a file, for example, keyset-info.json.

Finally, we'll use Nitro's datool dumpkeyset utility inside Docker to generate the keyset and keyset hash.

docker run -v $(pwd):/data/keyset --entrypoint datool offchainlabs/nitro-node:v3.1.2-309340a dumpkeyset --conf.file /data/keyset/keyset-info.json

This command will output two results: Keyset and KeysetHash. Save them to use in the next steps.

Example with mocked-up data

Here's an example that uses mocked-up data:

The JSON file is:

{
  "keyset": {
    "assumed-honest": 2,
    "backends": [
      {
        "url": "http://example",
        "pubkey": "YAbcteVnZLty5qRebeswHKhdjEMVwdou+imSfyrI+yVXHOMdLWA3Nf4DGW9tVry/mhmZqJp01TaYIsREXWdsFe1S5QCNqnddyag5yZ/5Y6GZRqx0BXmHTaxPY5kHrhvGnwxmlJVbUk1xjKRFgxxTdTk3c0AfM3JaeWYTed3avV//KGGdwHC+/Z7XPWmeXCNsGhY75YuoEAK2EwcJvAZK9de6lHEwtyBWvxcmOADxo6siacalEO+OdBL9VtHvG5FqEwbjsdnILAmTcb2YYVgqyq2joW6d/uXQ685hCWWYqC8RLQqTXoyrXEjYLjEEsMe6eRV9rRoBmj5/atB3uOYwixFv7A9YI5YiRjw2MfoB4rQnJAkhW4AJQiwWcV2+3lkJBg=="
      },
      {
        "url": "http://example",
        "pubkey": "YAg1+ZXyR48kiS0FDaoon4trnBsYW80oUy+I1hDCZCotxvNQl0AjbTPD4tkTaqsX+BnIxnEpO7ondxd2Lo0cH3usnhfdKNKTmpWbs45QD5wRw4zrvEJuLeqXxAF1plXRdACubHX/SeiEx5RpJJ5wlTJYhUtk+oRFxYWtRdxtxpdVAcavfP9wdCAsaH+Ke/GjrBkmiXVfIyJ1tMhCGxpWaem5BMKaKSzflht4OnwLTOc2kA3k2MY8X4WmXLRK80vvhArO+Eq3X0TEyRN2ELaBB6/zu9zBkRnHqSfBFbe5v7J9hcUA7nfRPsWpejrmv1HTtwpVAuhBbee1646f7uN2QRyjXIp/P1l8dgZXjPlqRxXOWjXPSOOcCh+qLe4i105oGQ=="
      }
    ]
  }
}

And when running the command we obtain:

$ docker run -v $(pwd):/data/keyset --entrypoint datool offchainlabs/nitro-node:v3.1.2-309340a dumpkeyset --conf.file /data/keyset/keyset-info.json
Keyset: 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
KeysetHash: 0xfdca3e4e2de25f0a56d0ced68fd1cc64f91b20cde67c964c55105477c02f49be

Step 2: Update the SequencerInbox contract

Once we have the keyset and its hash, we can configure the SequencerInbox contract so it accepts DACerts signed by the DAC members.

The SequencerInbox can be configured with the new keyset by invoking the setValidKeyset method. Note that only the chain owner can call this method.

Here's an example of how to use Foundry to configure the SequencerInbox with the keyset generated in the previous step:

cast send --rpc-url $PARENT_CHAIN_RPC --private-key $CHAIN_OWNER_PRIVATE_KEY $SEQUENCERINBOX_ADDRESS "setValidKeyset(bytes)" 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

Step 3: Craft the new configuration for the batch poster

To configure the batch poster, we'll use the JSON structure we created in Step 1. This will allow the batch poster to send RPC requests to all the DA servers (to store the information of the transactions being included in the next batch), craft the DACert, and store it in the SequencerInbox.

The configuration to enable the DAC in the batch poster looks like this:

{
  ...

  "data-availability": {
    "enable": true,
    "rpc-aggregator": {
      "enable": true,
      "assumed-honest": h,
      "backends": [
        {
          "url": "https://rpc-endpoint-of-member-1/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_1"
        },
        {
          "url": "https://rpc-endpoint-of-member-2/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_2"
        },

        ...

        {
          "url": "https://rpc-endpoint-of-member-n/",
          "pubkey":"PUBLIC_KEY_OF_MEMBER_N"
        }
      ]
    }
  },

  ...
}

The following parameters are used:

data-availability.enable: tells the batch poster to handle information stored in a DAC
data-availability.rpc-aggregator: includes information of the RPC endpoints of all the DA servers run by DAC members.
- enable: tells the batch poster that the RPC aggregator will be used
- assumed and backends: include information from the DA servers (following the same format as specified in Step 1)

Once the configuration is in place, you can restart your batch poster so it begins communicating with the DA servers to store transaction data, while storing the DACert in the SequencerInbox.

Step 4: Craft the new configuration for your chain's nodes

Finally, we also need to configure all other nodes so they can communicate with the DAC. To do that, we'll also use the JSON structure we created in Step 1.

The configuration to enable the DAC in a full node looks like this:

{
  ...

  "data-availability": {
    "enable": true,
    "rest-aggregator": {
      "enable": true,
      "urls": [
          "https://rest-endpoint-of-member-1/",
          "https://rest-endpoint-of-member-2/",
          ...
          "https://rest-endpoint-of-member-n/",
      ],
      "online-url-list": "https://url-of-list-of-rest-endpoints"
    }
  },

  ...
}

The following parameters are used:

data-availability.enable: tells the node to query information from the DAC
data-availability.rest-aggregator: includes information on the REST endpoints of all the DA servers run by DAC members.
- enable: tells the node that the REST aggregator will be used
- urls or online-url-list: usually only one of these is used, although both parameters can be used and the information will be aggregated together. urls is a list of all REST endpoints of the DA servers, and online-url-list is a URL to a list of URLs of the REST endpoints of the DA servers.

Once the configuration is in place, you can restart your node so it begins communicating with the DA servers to retrieve transaction data.

Step 0: Prerequisites​

Step 1: Generate the keyset and keyset hash with all the information from the servers​

What is a keyset?​

How to generate a keyset and a keyset hash​

Example with mocked-up data​

Step 2: Update the SequencerInbox contract​

Step 3: Craft the new configuration for the batch poster​

Step 4: Craft the new configuration for your chain's nodes​