Tools in Data Science End Term: 13 September 2026 (May 2026 term)

Treat the average as decisive and close the incident because overall latency barely changed.

Scale the database immediately because a tail-latency increase usually indicates database saturation under load.

Inspect slow checkout traces and correlated logs to identify where high- latency requests spend time.

Roll back every service immediately because an increase in p95 means the release slowed most requests.

The health response proves the process and every dependency required by purchases are operational.

The process is live, but dependency readiness and complete purchase behaviour remain unverified.

The health response rules out server-side faults, so the browser is probably responsible.

The invalid credentials are harmless because a responsive process can reconnect automatically.

Reliability worsened because twice as many failed requests were observed on the second day overall.

Reliability cannot be compared accurately until the latency distribution is included in both reports.

Reliability remained unchanged because both reporting periods contained some successful and some failed requests.

The observed failure rate decreased from 10% to 2%, despite the higher absolute failure count.