Recording intermittent 5xx failures without blaming content quality
An intermittent server failure is an availability problem to reproduce before it becomes a content diagnosis. Preserve successful and failed observations so a developer can investigate the difference.
Capture comparable attempts
Record the exact URL, request time with timezone, status, response body and any visible request identifier. Repeat a small bounded set of requests under comparable conditions. Keep successes as well as failures; a report containing only the failed attempt hides intermittency. Note whether the failure affects one route, a route family or the entire site.
Correlate with operational evidence
Ask the site operator to compare the recorded times with application errors, deployment events and upstream dependencies. Do not expose credentials or private infrastructure details in a public report. After a targeted repair, repeat the same requests and retain the new observations. A single subsequent success is useful evidence, but it is not a complete reliability history.
Source guidance
Use the website analyzer to collect a starting set of observations, then verify the specific evidence described above.Frequently Asked Questions
Should a temporarily failing page be deleted from the site?
Not by default. Establish whether the page remains useful and intended to exist, then investigate the availability failure rather than removing valuable content as a reflex.
What should the handoff record include?
Include the exact URL, observation time, request conditions, captured result and intended behavior. Keep a separate after-change observation so the repair can be checked against the original defect.