Drupal Migration — Performance Monitor

Bottom line — Drupal 11 is now the fastest and most reliable phase on record: 604 ms average response (9% faster than D10's 663 ms, 11% faster than the D8 baseline's 679 ms) with the lowest error rate of any phase at 0.30% (vs. 0.36% on D10 and 0.39% on D8). Apdex holds at 0.77, in line with D9/D10 and just under the D8 baseline's 0.79. With roughly six weeks and 5.99M requests since the Aug 1 upgrade, D11 is no longer a thin sample — its numbers look stable, not just an early read. The last 24h and 7d windows show normal day-to-day variation with no sustained degradation.

• D11 vs. baselines: response time and error rate have both improved steadily across D8 → D9 → D10 → D11, with D11 posting the best figures on both metrics of any phase so far — a solid early-to-mid-term result for the Drupal 11 upgrade.
• Overnight response spike: the 24h series shows one hour (02:07 CT) spiking to 1747 ms with apdex dropping to 0.47, against a 24h average of 618 ms — traffic in that bucket was thin (1,467 requests), consistent with a batch/cron job rather than a user-facing regression.
• Error-rate blip: the 19:07 CT hour logged the highest error rate in the last 24h (0.94%, vs. a 24h range of 0.00–0.94%); worth a quick log check but not an outlier trend.
• Traffic dip on Sep 12: the 7-day view shows Sep 12 request volume dropping to ~41K vs. a ~150–220K daily range for the rest of the week, alongside a low apdex-adjacent response time (519 ms) — likely a low-traffic day rather than an incident, since error rate stayed low (0.12%).
• Aug 1 maintenance window: the Aug 1 upgrade maintenance (05:00–07:00 CT) is excluded from the D10/D11 phase aggregates by design, so neither version is penalized for the outage, but it will still show as a visible traffic dip in the 24h/7d/full time-series charts on any date range that includes that morning.

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.