Error [OTLPExporterError]: Unknown error TypeError: fetch failed

[jfsiii]

Steps to reproduce

See https://stackblitz.com/edit/cloudflare-templates-gdw8ku?file=src%2Findex.ts

Given

    const otelSdk = new WorkersSDK(request, ctx, {
      service: 'some-service-name',
      endpoint: 'https://httpbin.org/status/200',
    });

    try {
      console.log('fetching...');
      const upstreamResponse = await otelSdk.fetch('https://httpbin.org/uuid');
      console.log('fetched. now sending');
      return otelSdk.sendResponse(upstreamResponse);
    } catch (ex) {
      console.log('ERROR fetching', ex);
      otelSdk.captureException(ex);
    }

logs

fetching...
fetched. now sending
Failed to flush spans: [
  Error [OTLPExporterError]: Unknown error TypeError: fetch failed
      at new OTLPExporterError2 (/tmp/tmp-7-WtdYhEgYO60S/index.js:3735:24)
      at eval (/tmp/tmp-7-WtdYhEgYO60S/index.js:4050:13) {
    data: undefined,
    code: undefined
  }
]

but with this change there are no errors logged.

- endpoint: 'https://httpbin.org/status/200';
+ endpoint: 'https://httpbin.org/anything',

I'm not sure what the difference is between those two URLs. In both cases, response.ok is true.

I initially ran into this with an endpoint that was returning a 400, but then found this case which seems to trigger it even on a success.

It seems like there may be an error in the error handling/reporting itself. Perhaps an import isn't working as expected and a function's not available?

[RichiCoder1]

I'm not sure why either to be honest. We're doing a pretty bog standard fetch: https://github.com/RichiCoder1/opentelemetry-sdk-workers/blob/main/packages/opentelemetry-sdk-workers/src/exporters/OTLPCloudflareExporterBase.ts#L103.

I could try logging out more details about the specific error to maybe surface the specific issue.

[jfsiii]

Yeah, it seemed straightforward. I think I'm misinterpreting the error message so I think this is more feedback than bug report.

I added some logging to the bundled worker in the Quick Edit panel. Here's the success case

and the error case

The error is from the POST to https://httpbin.org/status/200/v1/traces which 404s. https://httpbin.org/anything/v1/traces is valid.

The "unknown error" and TypeError threw me off. Perhaps there's a clearer way to state that the request to the OLTP endpoint failed. Perhaps I (also) was being a bit dense.

[RichiCoder1]

It's not a full fix since fetch is kind of opaque, but I'm updating the error message to at least hint it might be an export url issue.

[Schachte]

Not directly related, but for those encountering this error, couple things to note while debugging:

• if hosting your own collector endpoint, you can’t target the IP address directly, you need to add a DNS A record
• If you host your own collector endpoint and accessing via non-standard ports outside your zone, the request will not go through. I had to remap the port to expose on 80/443

[jfsiii]

@Schachte I struggled to debug the port issue on a different project as well, but it is documented at https://developers.cloudflare.com/workers/platform/known-issues/#custom-ports

[RichiCoder1]

@Schachte I struggled to debug the port issue on a different project as well, but it is documented at developers.cloudflare.com/workers/platform/known-issues/#custom-ports

Wow, I actually did not know that! I'll see about adding that to the readme, that seems easy to trip over.

[jfsiii]

Yeah, here's where I joined the group who have been surprised by this cloudflare/cloudflare-docs#4299 (comment).

At least it's documented now 🙃 . I believe wrangler also logs an error now