Baselines
X-ray's memory of what normal looks like on master. For every service it remembers what a
typical Tuesday 14:00 — or any other hour of the week — looks like: how much traffic, how many errors, how slow, how
much Neo4j load. The Anomalies system alert compares live
numbers with this memory and raises a finding when
something is far outside it.
Normal for this hour is about 3.78% (typically 0.00% to 13.81%) — no reading yet this hour.
29 hours compared with their normal; 4 could not be — X-ray has not learnt those hours yet; 2 known from too few samples to judge.
- Mon 18:00 93.81%, normal ≈ 0.00% inf% above normal see it →
- Mon 17:00 85.56%, normal ≈ 0.00% inf% above normal see it →
- Mon 13:00 78.79%, normal ≈ 0.00% inf% above normal see it →
- Mon 09:00 78.24%, normal ≈ 0.00% inf% above normal see it →
- Mon 14:00 72.66%, normal ≈ 0.00% inf% above normal see it →
- Mon 12:00 67.87%, normal ≈ 0.00% inf% above normal see it →
- Tue 04:00 58.71%, normal ≈ 0.00% inf% above normal see it →
- Mon 15:00 64.95%, normal ≈ 0.85% 76.0× normal see it →
- Mon 07:00 44.72%, normal ≈ 0.00% inf% above normal see it →
- Tue 02:00 22.11%, normal ≈ 0.00% inf% above normal see it →
- Mon 08:00 19.25%, normal ≈ 0.00% inf% above normal see it →
- Mon 21:00 18.46%, normal ≈ 0.00% inf% above normal see it →
- Tue 03:00 17.00%, normal ≈ 0.00% inf% above normal see it →
- Tue 01:00 15.31%, normal ≈ 0.00% inf% above normal see it →
- Mon 22:00 9.00%, normal ≈ 0.00% inf% above normal see it →
- Tue 06:00 8.46%, normal ≈ 0.00% inf% above normal see it →
- Mon 16:00 6.56%, normal ≈ 0.00% inf% above normal see it →
- Tue 00:00 6.34%, normal ≈ 0.00% inf% above normal see it →
- Tue 05:00 5.42%, normal ≈ 0.00% inf% above normal see it →
Not learnt yet: 29 scattered hours. An hour counts once two different weeks stand behind it, so the rest fills in as weeks pass.
whole cluster · Errors % of requests failing (4xx + 5xx)
Rows are days, columns are hours of the day, in UTC — baselines are learnt per UTC hour, so the grid cannot shift with your timezone; the cards above give the times in UTC as well. Hover a cell for its numbers. ?How a finding is decided: three of the last four complete five-minute readings must be far outside their hour's normal — at least 4 scaled MADs and 35% away from the median, and above a small absolute floor so a quiet service cannot alarm on noise. An hour needs 20 readings behind it before it may raise anything. The thresholds are editable on the Anomalies system alert.