log4js.logger - defaultParseCallStack error appears in my logs

[Schaka]

I see those messages pop up for me once in a while and cannot understand where they're coming from.
They show up in Stackdriver (which observes console output in GCP containers), as well as Sentry which works similarly.

As far as I can tell from the code, this shouldn't be happening at all. Is there any way to turn off those log messages?
They eat into our log budget unfortunately.

I wish I had more info to give you, but the only thing I can see in my logs is the message in the title.
There's nothing else being logged as far as I can tell. Not before and not after.

[lamweili]

Can you paste that entire line of the error? Or a screenshot would be fine as well.

[Schaka]

I should've been more clear. That is literally the full output.
I exclude those messages specifically from Sentry for now.

Here's the JSON Stackdriver returns:

{
  "textPayload": "log4js.logger - defaultParseCallStack error",
  "insertId": "op4yv8zdgzceot5s",
  "resource": {
    "type": "k8s_container",
    "labels": {
      "namespace_name": "company-namespace",
      "cluster_name": "cluster-name",
      "pod_name": "application-service-68d6b88b68-xdqxs",
      "location": "europe-west3",
      "container_name": "browser",
      "project_id": "infrastructure-260414"
    }
  },
  "timestamp": "2024-01-18T09:46:10.785199253Z",
  "severity": "ERROR",
  "labels": {
    "k8s-pod/app_kubernetes_io/instance": "application-name-prod",
    "k8s-pod/adsoul_com/stage": "prod",
    "k8s-pod/pod-template-hash": "68d6b88b68",
    "k8s-pod/app_kubernetes_io/name": "application-name",
    "k8s-pod/app_kubernetes_io/part-of": "the-application",
    "k8s-pod/app_kubernetes_io/component": "pcc-service",
    "compute.googleapis.com/resource_name": "gke-server-name-node-2023020809242-7bb25b77-pqtx",
    "k8s-pod/app_kubernetes_io/managed-by": "Helm",
    "k8s-pod/app_kubernetes_io/version": "2bfe0965"
  },
  "logName": "projects/infrastructure-260414/logs/stderr",
  "receiveTimestamp": "2024-01-18T09:46:12.217156923Z"
}

[lamweili]

What's your log4js version?

Do you have a sample/stripped-down code to replicate the issue?
You can upload the zip file here if you have it.

[Schaka]

Version is 6.9.1

It's quite the complicated setup, I'm afraid I can't really create a stripped down version.
I'm not even able to reproduce the problem locally.

Meaning it happens only in our GCP environment inside a Docker image.
We have a base image that we add some dependencies into, this image inherits from node:18.19-bullseye
I can't speak to the exact dependencies, but it pretty much only installs Chrome and creates a user to use for the image.

Then we create an image from the one I just talked about. I cleaned up that config for you:
Dockerfile.zip

There's not much more I can give you, unfortunately.
We start a NestJS server, use Log4js as a logger and somehow these errors show up, though never locally. This holds true when running it on a local container (MacOS) too.

If my employer gives me more time, I'll debug some more and try to provide extra resources for you. I'm sorry I can't be of too much help.

[xianping-yan]

[xianping-yan]

"log4js": "^6.9.1",
node: v18.17.0

[lamweili]

@xianping-yan, do you have more of the stack trace?
Can you provide the log4js configuration?

Or even better, can you upload a sample.zip for me to replicate?

[xianping-yan]

my-app.zip

I just verified that using it alone in nodejs is not a problem, but when combined with the nextjs framework, this issue will occur. So I think it's a compatibility issue between nextjs and log4js, and the above is the code that caused the problem。

Thank you for your support

[lamweili]

It would seem that they are using the variable location but I would need some time to look into it.

[xianping-yan]

OK， Thank you

[xianping-yan]

And please take a look at this prompt question. Thank you very much