Skip to main content

Off-chain data

The operator can store information on-chain or off-chain. Off-chain storage has two main use cases:

  • Data that may need to be deleted, such as personal data that a person can ask you to erase.
  • Data too large to be stored on the chain.

The principle: the application stores the data in clear, and only a digest of the data is anchored on-chain. Anyone given the clear data can later check that it matches the anchored digest.

Get the digest of your data​

Suppose you want to anchor the identity of the debtor off-chain:

{ "lastName": "Durand", "firstName": "Pierre" }

You send this data to the operator, which answers with the digest of the data:

{
  "digestData": {
    "ABCD1234": {
      "__salt__": "123456AB",
      "lastName": "Durand",
      "firstName": "Pierre"
    }
  },
  "digest": {
    "digest": "ABCD1234",
    "type": "json",
    "encoding": "canonical",
    "digestAlg": "sha256"
  }
}
  • digestData is the data to keep in your application, indexed by its digest. It contains a random __salt__, so that the digest cannot be guessed from the data alone.
  • digest describes the digest to anchor.

The operator guarantees the exact string that represents the JSON. It handles whitespace and sorts the keys (canonical encoding), so that the same JSON always gives the same digest.

Anchor the digest​

Use the __digest__ key to anchor a receivable that references the debtor without any personal data:

{
  "applicationId": "...",
  "data": {
    "name": "Energy supply invoice",
    "reference": "F-2025-11873",
    "issuer": "EDF",
    "date": "2023-12-05",
    "principal": 358.2,
    "interest": 12.5,
    "debtor": {
      "__digest__": {
        "digest": "ABCD1234",
        "type": "json",
        "encoding": "canonical",
        "digestAlg": "sha256"
      }
    }
  }
}

Resolve​

At resolution time, you provide the clear data in the digestData field of the request sent to the resolver:

{
  "linkedMicroblock": "cmts://resolve/al/A1480947BF02147CC62FE8163C/h/259",
  "policies": {
    "onResolutionFailure": "throw",
    "onCyclicResolution": "throw"
  },
  "digestData": {
    "ABCD1234": {
      "__salt__": "123456AB",
      "lastName": "Durand",
      "firstName": "Pierre"
    }
  }
}

The resolver replaces the __digest__ with the clear data, and checks that it matches the digest. The result is a resolution proof that carries the digestData, so the proof checker can repeat the check:

{
  "type": "resolutionProof",
  "proof": {
    "linkedMicroblock": "cmts://resolve/al/A1480947BF02147CC62FE8163C/h/259",
    "resolvedMicroblock": {
      "name": "Energy supply invoice",
      "reference": "F-2025-11873",
      "issuer": "EDF",
      "date": "2023-12-05",
      "principal": 358.2,
      "interest": 12.5,
      "debtor": { "lastName": "Durand", "firstName": "Pierre" }
    },
    "proofs": {},
    "digestData": {
      "ABCD1234": {
        "__salt__": "123456AB",
        "lastName": "Durand",
        "firstName": "Pierre"
      }
    }
  }
}

The proofs object holds the authenticity proofs; it is shown empty here for brevity.

Re-anchor off-chain data on-chain​

Sometimes off-chain data must later be anchored on-chain for good. For instance, the debtor's digest can be anchored with no expiry, since it holds no personal data (chainStorageInDays is null):

{
  "applicationId": "...",
  "chainStorageInDays": null,
  "data": {
    "__digest__": {
      "digest": "ABCD1234",
      "type": "json",
      "encoding": "canonical",
      "digestAlg": "sha256"
    }
  },
  "channels": [],
  "actors": [],
  "channelAssignations": [],
  "actorAssignations": [],
  "author": "..."
}

Suppose this microblock is referenced by cmts://resolve/al/3008327736206FA139A/h/598.

Now suppose you must anchor a report that lives 25 years, and about which the data subject cannot ask for deletion. The report must then contain the clear data too. You reference the debtor's microblock with __resolution_link__ and carry the clear data in __digest_data__:

{
  "applicationId": "...",
  "chainStorageInDays": 9000,
  "data": {
    "name": "Report",
    "receivable": {
      "reference": "F-2025-11873",
      "issuer": "EDF",
      "date": "2023-12-05",
      "principal": 358.2,
      "interest": 12.5
    },
    "debtor": {
      "__resolution_link__": "cmts://resolve/al/3008327736206FA139A/h/598",
      "__digest_data__": {
        "ABCD1234": {
          "__salt__": "123456AB",
          "lastName": "Durand",
          "firstName": "Pierre"
        }
      }
    }
  },
  "channels": [],
  "actors": [],
  "channelAssignations": [],
  "actorAssignations": [],
  "author": "..."
}

This has two effects:

  • the debtor's proof is anchored along with the report;
  • the personal data is used only within the scope of the report.

Resolution first resolves and checks the __resolution_link__, then checks that each __digest_data__ matches the corresponding __digest__. The resolved result is then a regular resolution proof.

GDPR​

Off-chain storage lets you honour the right to erasure: your application is responsible for finding the data of the person concerned and deleting it. Once deleted, only the salted digest remains on-chain.

warning

This mechanism does not cover the right to rectification of what is already anchored.