Locker on n8n

Add the FlurryPORT Locker node to the end of a workflow and every execution is kept in Locker, exactly as it happened, with a signed receipt anyone can check. Your execution history stays yours after n8n has pruned it.

Self-hostedWorks today on self-hosted n8n as a community node. Waiting for n8n to verify it for their own list.The node on npm

Before you start

You need a self-hosted n8n where you can install community nodes, and a FlurryPORT account. The node is not available on n8n Cloud. There is no password: you enter your email and we mail you a sign-in link. You can create the account in a minute. Locker starts free with 250 runs or 14 days, whichever comes first.

Set it up

1
In n8n, open Settings, then Community nodes, and choose Install. Enter n8n-nodes-flurryport-locker.
2
In FlurryPORT, open Locker and choose the Any HTTP connector. Give the install a name, such as the name of your n8n server, and choose Create install.
3
Copy the Install key. It starts with fp_. You can show it again later from the install.
4
Back in n8n, create a credential of the type FlurryPORT Locker API. Paste the install key and leave the API base URL as it is. n8n tests the connection when you save.
5
Add the FlurryPORT Locker node as the last step of your workflow and pick the credential. Run the workflow. Each item now carries a receipt.

One install key is one chain of runs. Use one install for a whole n8n server, or one for each client or workflow you want kept apart.

The node

The node has two operations.

OperationWhat it does
ArchiveKeeps this execution with a signed receipt. This is the one you put at the end of a workflow.
Get ReceiptReads one archived run and its receipt by receipt id.

The Archive fields

FieldWhat to put in it
InputsWhat the run started from, as JSON. Optional.
OutputsWhat the run produced. Left alone, it is the items reaching the node, which is what most people want.
Run StatusSuccess, Error or Unknown. Defaults to Success.
LabelsName and value pairs, such as a client name, to find runs by later.

Options

OptionWhat it does
Hash Binary FilesOn by default. Records a SHA-256 fingerprint of each binary file on the items.
Max Section Kilobytes256 by default. Inputs and outputs are each kept up to this size. Past it, the node keeps the whole items that fit, the count, and a fingerprint of the full thing.
Fail Workflow On ErrorOff by default. Turn it on to stop the workflow when a run cannot be archived.

What the node adds to each item

Your items pass through unchanged, with one extra field, locker, for the next node to use:

{
  "orderId": 1001,
  "locker": {
    "archived": true,
    "duplicate": false,
    "receiptId": "2BsKb3W3pL6qkbnuH7yJGg",
    "verifyUrl": "https://flurryport.io/receipts/2BsKb3W3pL6qkbnuH7yJGg",
    "chainPosition": 12,
    "hash": "0f4c6d1c9a7a...",
    "allowance": { "state": "free", "planName": "Free", "runsRemaining": 238, "daysRemaining": 9, "locked": false }
  }
}

verifyUrl is the public receipt page. Send it to anyone who needs proof of the run. duplicate is true when n8n retried an execution that was already archived, in which case the first receipt comes back and nothing new is added.

If something goes wrong

By default the node never stops your workflow. When a run cannot be archived, the items still pass through, and the field says what happened and what to do:

{
  "orderId": 1001,
  "locker": {
    "archived": false,
    "code": "http_401",
    "status": 401,
    "retryable": false,
    "reason": "Locker did not accept the install key. Paste the key from the install page in FlurryPORT into the credential."
  }
}
What happenedWhat the node does
The install key is wrong or was removedPasses the items through with the reason. Fix the credential.
Too many runs at onceWaits as long as it is told to, then tries once more.
The network is downPasses the items through, marked as worth retrying.
The run is larger than 1 MBArchives the fingerprints without the contents, so the run still gets a receipt and the chain has no gap.

Turn on Fail Workflow On Error if a missing archive should stop the workflow instead.

Binary files

Binary files are never uploaded. Each is recorded by name, type, size and SHA-256 fingerprint, so you can show later that a file you still hold is the one the run produced. Files over 25 MB are listed without a fingerprint.

What it cannot tell you

The node archives the executions that reach it. An execution that failed before the last step, or never started, never reaches the node, so it is not in the archive. Locker shows what ran. It does not compare itself against n8n’s own execution list.

Where your runs are

Sign in to FlurryPORT and open Locker. Each install has its own chain of runs. The Locker page covers reading the archive, the receipt page, how long runs are kept, and the API the node calls.

The node is on npm as n8n-nodes-flurryport-locker, under the MIT licence.