Drupal Migration — Performance Monitor

Bottom line — Drupal 11 is performing better than every prior version on live: average response time is down to 592 ms (vs 664 ms on D10 and 679 ms on the D8 baseline), and the error rate has fallen to 0.24% (vs 0.36% on D10 and 0.39% on D8). Apdex on D11 sits at 0.77, essentially flat with D9/D10 and just a touch below D10's 0.78. The Aug 1 upgrade window is excluded from both the D10 and D11 aggregates so neither version is penalized for the maintenance downtime itself.

• D11 early read: D11 now covers roughly eight weeks and ~8.05M requests, so this is no longer a thin sample — it's a solid read, and it's the fastest, lowest-error phase recorded so far, improving on the D10 baseline it just replaced.
• Maintenance window: the 2026-08-01 05:00–07:00 Chicago upgrade window is cut out of the D10/D11 phase numbers, but it is now nearly two months in the past — it no longer shows up as a dip in the 24h or 7-day charts (those only cover the last week), and the weekly "full" series smooths it away too, since a 2-hour gap barely dents a 7-day bucket.
• Recent error blip: the 7-day view shows a small error-rate bump on Sep 23–24 (0.05% and 0.04%, vs near-zero on surrounding days) alongside elevated traffic (~210K–240K req/day) and a slight apdex dip to 0.72–0.73 — worth a quick look if it recurs.
• Traffic pattern: Sep 26 saw unusually light traffic (~56K requests vs 150K–240K on neighboring days), consistent with a normal weekend/low-traffic lull rather than an outage — error rate and apdex stayed healthy that day.
• 24h trend: the last 24 hours show the expected daily ramp (overnight lows around 3K req/hr climbing to ~21K req/hr by mid-morning) with response time rising modestly from ~450 ms to ~630 ms under peak load, and error rate staying at or near 0%.

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.