Error rates have returned to normal and HTTP measurement ingestion should work as expected again.
We determined the root cause to be a recent change that was deployed resulting in intermittent errors for around 13 minutes, causing 404 errors to user requests. We rolled back the change and error rates returned to normal.
