Trace view page stuck in infinite reload — `_expOrImpId` missing from follow-up XHR

Integrator.io
,
1

Symptom: clicking "Open trace view" begins page render, then the page re-mounts and starts loading again, repeating indefinitely. No trace content is ever displayed.

Diagnosis from the Network tab:
The Open trace view page is stuck in an infinite reload loop and never renders. Reproduced in both Chrome and Brave (latest versions), after clearing all site data. The issue started in Chrome a few weeks ago and within the past few days began affecting Brave as well, which suggests a platform-side change rather than a browser/cache issue.

  1. The initial request returns successfully with the trace metadata. Note traceView: false and _expOrImpId populated in steps[0]:
{
    "traceKey": "001-4373-21595624008",
    "duplicateTraceKey": false,
    "traceView": false,
    "steps": [
        {
            "_expOrImpId": "69fc90d54de8d714dccd8d7c",
            "recordId": "PCzJCR",
            "groupId": "AEvqZd",
            "status": "success",
            "timeTaken": 8751,
            "startedAt": "2026-05-14 20:15:51.743",
            "completedAt": "2026-05-14 20:16:00.494"
        }
    ]
}
  1. The follow-up XHR then fails with a 400, because _expOrImpId is not included in the request — even though the value is present in the first response:
{
    "errors": [
        {
            "code": "missing_required_field",
            "message": "_expOrImpId is required."
        }
    ]
}

20260514_integrator.io.zip (145.1 KB)

The component appears to throw on the failed follow-up, React re-mounts, the original request fires again, and the cycle repeats. That's the loop.

HAR file attached with the full request/response sequence captured during a reproduction.

Happy to provide additional captures or hop on a screenshare if helpful. This is blocking debugging on active integration work so any quick acknowledgment of whether this is a known regression would be appreciated.

2

Thanks for logging this @Sean_Ayotte. We saw the same thing yesterday, and we're looking into it as a priority. Logged in Zendesk as 279121 for future reference.

Regards,

Matt