Updading the Gmail node to the latest version

Questions
1

Howdy!

I’m seeing a lot of 500 errors when using the Gmail node, even though I’m not changing anything in my flow. It runs fine for a few turns and a few executions, then it breaks at different locations with a 500 error on the Gmail node.

I noticed that my Gmail node is version 2.1 but the newest version says 2.2. I’m at the latest version of n8n but the Gmail node is still at version 2.1. What am I supposed to do to update it to 2.2? And will it address the repetitive 500 errors?

L

2

Hi @DupervalAI Welcome!
I think you should update your GMAIL node and then try to perform actions, and let us know if you get the same errors. I think for the 500 error, you should add the retry on fail, as that error is related to them so the best we could do is retry :slight_smile:

3

hi @DupervalAI

I’d treat a 500 error as an issue coming from the external service, not the node itself. I’d start by reconnecting or recreating the credentials, then check the execution details for the exact Google error message, like invalid_grant or quota exceeded. If it’s intermittent, I’d also add retries and reduce request volume to avoid rate limits. In most cases, fixing auth and checking the real API response is more effective than just updating the node.

PS: Node version and n8n version aren’t exactly the same thing. In n8n, existing workflows usually keep using the node version they were originally saved with, even when a newer node version is available. New workflows load the latest node version by default.

4

Ah.… So I need to delete the current Gmail node and add the new node in my workflow, but with the same instructions?

As for the remote error, I understand that it’s probably a service problem, but the error I get is:


{ "error": 
  { 
    "code": 500, "message": "Unknown Error.", 
    "errors":
      [ 
        { 
          "message": "Unknown Error.",
          "domain": "global",
          "reason": "backendError"
        }
      ],
    "status": "UNKNOWN" 
  }
}

It’s the reason I want to update the node, in case there’s something in the current way it calls the backend that causes the problem.

Thanks,

L

5

@DupervalAI
You don’t necessarily need to replace the node.
but, yes.

6

You probably don’t need to replace the node just to test this. In n8n, existing workflows usually stay on the node version they were saved with, so seeing Gmail 2.1 in an older workflow while newer ones show 2.2 is normal.

I’d check the full Google response in the failed execution first. backendError is usually Google being flaky on their side, not a schema difference between 2.1 and 2.2. If it’s intermittent, I’d add retry-on-fail around the Gmail step and keep request bursts lower.

If you do want to compare behavior, duplicate the workflow, replace just that Gmail node with a fresh one, reconnect the same creds, and run both against the same test input. That’ll tell you pretty quickly whether this is node-version-specific or just Gmail returning transient 500s.