The range selector in Monitor Details changes the summary metrics, response-time chart, Uptime Timeline, and incident context shown for one monitor. Choose the shortest range that answers your question, then expand it when you need context.
Range comparison
| View | Best for | Response summary | Uptime Timeline |
|---|---|---|---|
| 24 hours | Recent failures and latency spikes | Five-minute buckets | Up to 24 hourly blocks |
| 7 days | Daily patterns and recent reliability | Hourly buckets | Seven daily rows with hourly blocks |
| 1 month | Week-to-week trends and monthly review | Hourly buckets | Calendar-style daily blocks |
| 1 year | Seasonal patterns and long-term reporting | Daily buckets | Up to 12 monthly cards |
Use 24 hours for active troubleshooting
Choose 24 hours when investigating something that happened today or overnight. It provides the most detailed response chart and makes it easier to compare an alert with nearby checks.
For exact status codes, errors, and request timestamps, continue to Detailed Logs. The chart is still a summary, even at this range.
Use 7 days for recurring patterns
Choose 7 days to answer questions such as:
- Does latency increase at the same hour every day?
- Did failures begin after a weekly deployment?
- Was the problem isolated to one day?
- Is an overnight maintenance window affecting checks?
The timeline shows one row per day and one block per hour, making repeated times easier to compare.
Use 1 month for operational review
Choose 1 month to review reliability across several weeks. The calendar-style timeline helps identify clusters of affected dates, while the hourly response chart preserves more detail than the daily visual blocks.
Use this range for a broad report, then switch to 24 hours or 7 days to investigate a particular event that still falls within raw-check retention.
Use 1 year for long-term trends
Choose 1 year for seasonal changes, infrastructure planning, and a high-level reliability history. Each Uptime Timeline card summarizes a month and shows its availability and check totals.
The one-year view uses daily rollups and is not designed to prove the exact timing of an old individual request. Incident history can provide confirmed outage records after raw checks expire.
Why values change when you switch ranges
Every range recalculates its own start time, totals, availability, average response, p95 response, and incidents. Values can change because:
- The time window includes different checks.
- Longer ranges use larger summary buckets.
- An old outage enters or leaves the selected period.
- A new monitor has only part of the requested history.
- Raw records have expired while retained rollups remain.
Do not expect the 24-hour percentage to match the one-year percentage. They answer different questions.
A practical investigation sequence
- Open 24 hours and inspect the newest behavior.
- Expand to 7 days to look for recurrence.
- Use 1 month to assess broader impact.
- Use 1 year only when long-term context is useful.
- Return to Detailed Logs or an incident page for exact evidence.
Monitor Details keeps a recently loaded range in the browser for one minute. Switching back quickly can display that cached result instead of making a duplicate request.