Drupal Migration — Performance Monitor

Bottom line — Spectre_5 remains healthy just under eight weeks into the Drupal 11 upgrade: average response time is 594 ms on D11, the best of the four phases, and its error rate (0.26%) is the lowest yet, improving on both the D10 (663 ms, 0.36%) and D8 (679 ms, 0.39%) baselines. Apdex is essentially flat across phases (0.77–0.79), so the gains look like real latency/error improvements rather than a shift in user-experience scoring. The last 24 hours and 7 days show normal traffic-following-load behavior with one isolated overnight anomaly.

• D11 early read: with 7.46M requests logged since Aug 1, Drupal 11's sample is now substantial, not thin — response time and error rate are both trending better than D10 and D8, while apdex holds steady around 0.77, a hair below the D8/D10 base of 0.78–0.79.
• Overnight spike: the 02:09 America/Chicago hour in the last 24h shows response time jumping to 1087 ms and apdex dropping to 0.65, against very light overnight traffic (~2.0K requests) — a low-volume outlier rather than a sustained issue, with the surrounding hours back near baseline.
• Elevated error blip: 13:09 today posted a brief error-rate tick to 0.39% (vs a normal 0–0.1%) at normal daytime volume (~19K requests); it did not recur in the following hour and response time stayed in range, consistent with a transient blip rather than a trend.
• Weekly pattern normal: the 7-day view shows the expected weekend traffic dip (Saturday Sep 19 at ~29K requests vs. ~117-211K on weekdays), with response time and error rate staying in the usual 527–634 ms / 0–0.21% range.
• Maintenance window: the Aug 1 05:00–07:00 Chicago maintenance/deploy window is deliberately excluded from both the D10 and D11 phase aggregates so neither version is scored on downtime. It is now well outside both the 24h and 7d series windows and does not appear as a dip in either.

Method & caveats

• Cutovers. Drupal 9 / PHP 8 went live May 23, 2026; Drupal 10 Jul 4, 2026 — marked on every chart that spans them.
• Drupal 10 is an early read (~5 days). Its phase card uses the Jul 4–9 window directly.
• Recent spike. Jul 8 shows an isolated response jump (≈2× the prior day), coinciding with elevated reports/progress-API latency that day — an operational blip, not a migration regression.
• Source & retention. New Relic timeslice metrics (HttpDispatcher, Apdex, Errors/all), ~13-month retention — the full arc back to mid-2025 is covered. Percentiles (p95/p99) live in 8-day event data and are reported in the point-in-time reviews, not here.
• Refresh. Regenerated daily by a scheduled routine that re-queries New Relic and redeploys to this same URL.