MariaDB Performance Audit

Database performance benchmarks with trend analysis and competitive context.

7/4/2026

45 min read

Note

Competitive comparisons based on publicly available documentation as of June 2026. All SveltyCMS metrics self-measured via bun test tests/benchmarks/.

Important

Console-only by default. Use BENCHMARK_RECORD=1 to write results to this report.

BENCHMARK_RECORD=1 bun test tests/benchmarks/auth-performance.test.ts

The full matrix runner (bun run scripts/benchmark-matrix/index.ts --sql) always records.

📊 Latest Performance Audit

📝 Fix Overlays

Note

Phase notes and remediation entries appear here after matrix or surgical runs.

(21/10/2026, 18:34:03)

✅ PASS · 38/45 tests · 7 skipped · MARIADB · Bun 1.4.2

Dimension health

Dimension	Tests	Status	Worst Δ
Core	13	🟢	+827%
API	9	🟢	-48%
Scale	18	🟢	+118%
Resilience	5	🟢	—

Issues (action required)

✅ All clear — no regressions or budget violations detected.

Executive latency matrix

Scenario	Latency	Trend	Budget	Result
Cold Start	N/A	⚪ (—)	< 5000ms	⚪ N/A
REST (Collections)	N/A	⚪ (—)	< 5ms	⚪ N/A
Middleware Hooks	N/A	⚪	< 2ms	⚪ N/A
GraphQL (Avg)	N/A	⚪ (—)	< 5ms	⚪ N/A
Million-Row Index	N/A	⚪	< 250ms	⚪ N/A
DB Raw (p95)	N/A	⚪	< 50ms	⚪ N/A

Historical pulse

Scope: Sparklines only — full run tables: tests/benchmarks/results/history.sqlite

Metric	Trend	Sparkline	Latest

📈 Latency trend chart (last 10 runs)

xychart-beta title "MARIADB REST p95 (last 10 runs)" x-axis ["R1", "R2", "R3", "R4", "R5", "R6", "R7", "R8", "R9", "R10"] y-axis "Latency (ms)" line "p95" : [0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000]

📊 Regression severity matrix (quadrant)

quadrantChart title "Regression severity (impact vs delta)" x-axis Low latency impact --> High latency impact y-axis Low delta --> High delta quadrant-1 Monitor quadrant-2 Investigate soon quadrant-3 Acceptable variance quadrant-4 Critical — fix first API LAYER LATENCY: [0.09, 0.95] COMPETITIVE 9-WORKLOAD REPLI: [0.05, 0.25] MIDDLEWARE FLAMEGRAPH PROFIL: [0.05, 0.15] Incremental Content Reload: [0.05, 0.30] Middleware & Hooks Performan: [0.05, 0.05] State Machine Transitions: [0.95, 0.05] Database Adapter Raw CRUD: [0.05, 0.05] ACID Transaction Overhead: [0.05, 0.11]

📋 Partial report — run full matrix for complete cross-DB analysis.

🔴 Top Regressions

🔴 seo-performance [Missing Path (404 Log)] (mariadb): +706% | 154.69ms vs 19.18ms baseline 🔴 competitive-workload-replica [Seed Burst (500 docs, HTTP)] (mariadb): +230% | 0.39ms vs 0.12ms baseline 🔴 migration-scale [Bulk Migration (10k docs)] (mariadb): +104% | 30.75ms vs 15.06ms baseline 🔴 scale-spectrum [findByIdRandom @ 1] (mariadb): +97% | 1.55ms vs 0.79ms baseline 🔴 rest-api-performance [Collection Find (List Limit 20)] (mariadb): +79% | 0.77ms vs 0.43ms baseline

🟢 Top Improvements

config-promotion [Config Import (10 cols)] (mariadb): -81% | 12.08ms vs 62.50ms baseline competitive-workload-replica [update (Warm Concurrent 8c)] (mariadb): -66% | 1.98ms vs 5.76ms baseline competitive-workload-replica [listFilterSort (Warm Concurrent 8c)] (mariadb): -65% | 0.33ms vs 0.94ms baseline competitive-workload-replica [findByIdRandom (Warm Concurrent 8c)] (mariadb): -61% | 0.22ms vs 0.57ms baseline competitive-workload-replica [listPlain (Warm Concurrent 8c)] (mariadb): -61% | 0.25ms vs 0.63ms baseline

Current Run Summary (2026-10-17)

Scope: Only tests that ran in THIS invocation.

⚠️ Executive PASS/FAIL reflects the last full matrix run.

Run ID: 01a14a7b-204a-70b5-87d9-dd97e98f0101 · Tests invoked: 1 · Metrics: 2 · DB: mariadb

Test	Metric	Avg (ms)	p95 (ms)	RPS	Trend	Detail
Cache Performance	Cache Miss (Bypass)	0.777	1.021	1234	🟢	⚪ established at 0.777ms (1 run)
Cache Performance	Cache Hit (L1 Warm)	0.176	0.239	5357	🟢	⚪ established at 0.176ms (1 run)

Historical Trends (2026-10-21)

Mode: production-parity matrix runs only (run_mode='matrix' + server_mode='production' — NODE_ENV=production, no TEST_MODE). Standalone/mode-variant samples are recorded but never trended here.

Scope: Sparklines only — full run tables live in history.sqlite (link below). Runs = sparkline bar count.

Detail: tests/benchmarks/results/history.sqlite · SELECT test_id, phase, server_mode, avg_ms, p95_ms, rps, timestamp FROM runs WHERE db_type='mariadb' AND redis=0 ORDER BY timestamp DESC

Series: 54 · DB: mariadb

Metric	Trend	Sparkline (last 7)	Latest
ADMIN UX VITALITY (warm)	🟢	`█▂▆▁`	0.249ms
AI PERFORMANCE (warm)	🟢	`▁▁▁▁`	0.000ms
API LATENCY (warm)	🟢	`▁▇▆▂▁▁█`	4.387ms
AUTH PERFORMANCE (warm)	🟢	`▁▃█▄▄▄`	1.121ms
AUTH SESSION AMR (warm)	🟢	`▃▂▁▇█▅▄`	0.743ms
BUILD ANALYSIS (warm)	🟢	`▁▁`	0.000ms
CACHE HIT RATIO (warm)	🟢	`▅▁█▃▃▁▆`	1.805ms
CACHE PERFORMANCE (warm)	🟢	`▂█▁▇`	0.949ms
CACHE SERVICE (warm)	🟢	`▁▆█▁▄▆`	0.762ms
CLIENT JOURNEY (warm)	🟢	`█▁`	5.501ms
COMPETITIVE WORKLOAD REPLICA (warm)	🟢	`▁▂▁▆▃█▄`	0.815ms
CONCURRENCY MAX (warm)	🟢	`▁█`	32.280ms
CONFIG PROMOTION (warm)	🟢	`▁▆█▅▆▆▆`	11.951ms
CONTENT GUI SAVE (warm)	🟢	`▁█`	2.437ms
CONTENT SCALE STRESS (warm)	🟢	`▇▂█▁`	0.154ms
CONTENT SCAN (warm)	🟢	`█▂▇▁`	0.662ms
DATABASE PERFORMANCE (warm)	🟢	`▂▁▂▁█▂▁`	0.455ms
ENTRY EDIT HYDRATION (warm)	🟢	`▁▁▅█▁▁▆`	0.016ms
GRAPHQL API PERFORMANCE (warm)	🟢	`▁▅▄█▆▁▁`	0.860ms
GRAPHQL CACHE HIT (warm)	🟢	`▂▃▁█▁▃▁`	1.317ms
GRAPHQL LOCAL SDK (warm)	🟢	`▇▁█▁`	0.000ms
GRAPHQL STRESS (warm)	🟢	`█▁▂▃▁▅▆`	13.698ms
HOOKS PERFORMANCE (warm)	🟢	`▃▂▁█▁▄▄`	0.844ms
INDEX PRESSURE (warm)	🟢	`▂█▁▇`	0.892ms
LOCAL API PERFORMANCE (warm)	🟢	`█▄▁▆▁▄`	0.358ms
LOCAL API THROUGHPUT (warm)	🟢	`█▁▄▂`	1.099ms
LOCAL CMS CRUD (warm)	🟢	`▃▃▃▂▁█▅`	1.240ms
MEDIA UPLOAD STRESS (warm)	🟢	`▁█▂▂`	8.640ms
MEMORY COUNTERS (warm)	🟢	`▁▁`	0.000ms
MIDDLEWARE FLAMEGRAPH (warm)	🟢	`▃▂▂█▁▁▇`	1.690ms
MIGRATION SCALE (warm)	🟢	`▄▁▁█▁▁`	0.895ms
MIXED WORKLOAD (warm)	🟢	`█▁`	1.019ms
MULTI TENANT PERFORMANCE (warm)	🟢	`█▁`	0.615ms
OPENAPI PERFORMANCE (warm)	🟢	`▆▁▆▆▁█`	1.980ms
PRODUCTION DAY (warm)	🟢	`▁█`	2.195ms
REALTIME PERFORMANCE (warm)	🟢	`▁▁▁▁▁▁`	0.000ms
REST API PERFORMANCE (warm)	🟢	`▁█▆▂▄▂▅`	0.707ms
REVISION STRESS (warm)	🟢	`▂█▂▁▆▁`	0.291ms
RIGHT TO BE FORGOTTEN AUDIT (warm)	🟢	`█▁`	8.955ms
SCALE SPECTRUM (warm)	🟢	`█▄▄▆▁▁▁`	1.376ms
SECURITY AUDIT (warm)	🟢	`█▁▁▁▁▁▁`	0.001ms
SEO PERFORMANCE (warm)	🟢	`▂▁▁▁█▁▁`	1.823ms
STATE MACHINE TRANSITION (warm)	🟢	`▁▁▇█`	49.378ms
TELEMETRY PERFORMANCE (warm)	🟢	`█▁█▁`	0.000ms
TEMPORAL INTEGRITY (warm)	🟢	`█▁`	3.579ms
TRANSACTION ACID (warm)	🟢	`▁█▁▆`	1.724ms
TRUTH LATENCY (warm)	🟢	`▄▁▁█▃▁▁`	0.000ms
WEBSOCKET BROADCAST (warm)	🟢	`▁█`	0.211ms
WIDGET PERFORMANCE (warm)	🟢	`▇▁▇█▁▇`	0.596ms
CONTENT INCREMENTAL RELOAD (warm)	⚪	`▃▁█▃▂▁█`	0.801ms
LARGE PAYLOAD STREAMING (warm)	⚪	`▃▃▄▁▃▃█`	16.574ms
LOCAL SDK VS DIRECT (warm)	⚪	`▁▄▁▂▅██`	1.846ms
NEGATIVE CACHE (warm)	⚪	`▇█▁▆▆▁`	0.001ms
SHARP VS BUN IMAGE (warm)	⚪	`▅▄▁▁▂█▇`	72.606ms

📊 Infrastructure drill-down (expand for adapter, hooks, memory trends)

📊 Infrastructure drill-down

🏷️ Mixed Workload Performance {#mixedAvg}

Mixed p95: — (not measured in this run) | Trend: ⚪ (—)

Production request mix: 60% Reads, 20% Writes, 15% GraphQL, 5% Media.

🏷️ Database Adapter Performance {#dbRaw}

Raw Adapter p95: — (not measured in this run) | Target: < 50ms | Trend: ⚪ (—)

Direct database access latency (bypass CMS logic).

🏷️ Middleware & Hooks Audit {#hooks}

Middleware p95: — (not measured in this run) | Target: < 2ms | Trend: ⚪ (—)

Total overhead of security, logging, and auth hooks.

🏷️ Memory Stability Audit {#memGrowth}

Memory RSS Δ: — (not measured in this run) | Target: < 60MB | Trend: ⚪ (—)

Peak memory growth during stress workload.

Enterprise scale audit

🏷️ Multi-Tenancy Performance {#tenancyAvg}

Tenancy p95: — (not measured in this run) | Trend: ⚪ (—)

Measures latency across tenant isolation boundaries.

🏷️ Relational Resolver Latency {#relationalAvg}

Relational p95: — (not measured in this run) | Trend: ⚪ (—)

Tests performance of complex JOINs and nested relationship population.

🏷️ Telemetry & Update Performance {#telemetryAvg}

Telemetry p95: — (not measured in this run) | Trend: ⚪ (—)

Measures overhead of telemetry collection and cryptographic signing.

🔬 Full benchmark ledger (41 modules)

Expand dimension groups, then any row, for ASCII truth tables. Executive summary above shows pass/fail and issues only.

Core · 28 tests

🏷️ API LAYER LATENCY · ⚪ 4.387ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 87% — check connection pool saturation or lock contention. **Severe degradation** (+827%) — review recent commits immediately. > **Check**: `src/routes/api/[...path]/+server.ts` · `src/hooks.server.ts`

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                     SVELTYCMS — API LAYER LATENCY                                     ║
║                              File: tests/benchmarks/api-latency.test.ts                               ║
║                                       Ran: 2026-10-06 10:46:41                                        ║
║              Cold full-pipeline findById + warm TURBO-HIT path after responseCache fill.              ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ HTTP: findById @ 8c (Cold)   │ Cold:     2.53 ms │ Warm:     2.601 ms │ p95:     3.252 ms │   2,997 RPS ║
║ HTTP: findById TURBO-HIT @ 8c (Warm) │ Cold:     0.45 ms │ Warm:     0.896 ms │ p95:     1.318 ms │   8,538 RPS ║
║ HTTP: zstd large GET (>64KiB, drained) │ Cold:     2.43 ms │ Warm:     3.604 ms │ p95:     5.131 ms │     503 RPS ║
║ HTTP: findById TURBO-HIT idle (Warm) │        0.203 ms │ p95:        0.269 ms │ RPS:      4,880 ║
║ HTTP: findById TURBO-HIT during zstd (Warm) │        0.277 ms │ p95:        0.478 ms │ RPS:      3,321 ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ PHASED COLD START · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — PHASED COLD START AUDIT ║ ║ File: tests/benchmarks/cold-start-phased.test.ts ║ ║ Ran: 2026-09-18 18:18:21 ║ ║ Measures process spawn, dependency hydration, and first-request TTFB readiness. ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Initial Cold Boot (Fresh Process) │ 2328.690 ms │ p95: 2328.690 ms │ RPS: 0 ║ ║ Steady-State Respawn (Average) │ 1573.369 ms │ p95: 2328.690 ms │ RPS: 0 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ SRE TRUTH AUDIT · ⚪ ⏳ pending · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                      SVELTYCMS — SRE TRUTH AUDIT                                      ║
║                             File: tests/benchmarks/truth-latency.test.ts                              ║
║                                       Ran: 2026-10-06 09:26:42                                        ║
║            Validates performance claims by comparing SDK, Middleware, and Real HTTP Stack             ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Logic Baseline               │ Cold:     0.01 ms │ Warm:     0.000 ms │ p95:     0.001 ms │ 302,024 RPS ║
║ Local SDK (Full)             │ Cold:    16.75 ms │ Warm:     0.005 ms │ p95:     0.009 ms │ 166,380 RPS ║
║ HTTP End-to-End              │ Cold:     4.67 ms │ Warm:     0.282 ms │ p95:     0.354 ms │   3,524 RPS ║
║ HTTP Cold (Cache-Busted)     │ Cold:     2.62 ms │ Warm:     0.885 ms │ p95:     1.221 ms │   1,127 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Incremental Content Reload · ⚪ 0.801ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 79% — check connection pool saturation or lock contention. **Severe degradation** (+247%) — review recent commits immediately. > **Check**: `src/content/engine.server.ts` · `src/routes/api/[...path]/handlers/content.ts`

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                             SVELTYCMS — INCREMENTAL CONTENT RELOAD AUDIT                              ║
║                       File: tests/benchmarks/content-incremental-reload.test.ts                       ║
║                                       Ran: 2026-10-06 09:30:32                                        ║
║      Measures surgical single-file fullReload vs full reconciliation path and batch processing.       ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Incremental fullReload (1 file) │ Cold:     4.84 ms │ Warm:     0.066 ms │ p95:     0.121 ms │  13,745 RPS ║
║ Full Reconciliation Reload   │ Cold:     2.74 ms │ Warm:     0.191 ms │ p95:     0.256 ms │   5,195 RPS ║
║ Sequential 10-File Reload    │ Cold:     6.04 ms │ Warm:     0.801 ms │ p95:     1.193 ms │   1,244 RPS ║
║ Batched 10-File Reload       │ Cold:     1.50 ms │ Warm:     0.231 ms │ p95:     0.389 ms │   3,932 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Middleware & Hooks Performance · ⚪ 0.844ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 16% — check connection pool saturation or lock contention. **Severe degradation** (+33%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                 SVELTYCMS — MIDDLEWARE & HOOKS AUDIT                                  ║
║                           File: tests/benchmarks/hooks-performance.test.ts                            ║
║                                       Ran: 2026-10-06 10:48:35                                        ║
║Measures the cost of the full middleware chain including Turbo, Security, Auth, and Audit via HTTP E2E.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Static Asset (No Middleware) │ Cold:     1.59 ms │ Warm:     0.770 ms │ p95:     1.369 ms │   4,441 RPS ║
║ Turbo Pipeline (Light)       │ Cold:     0.44 ms │ Warm:     0.748 ms │ p95:     1.713 ms │   4,472 RPS ║
║ Full Security + Auth Pipeline │ Cold:     3.46 ms │ Warm:     0.528 ms │ p95:     0.782 ms │   7,209 RPS ║
║ REST with API Caching        │ Cold:     3.67 ms │ Warm:     0.542 ms │ p95:     0.821 ms │   6,907 RPS ║
║ Mutation (audit logging disabled) │ Cold:     6.46 ms │ Warm:     1.494 ms │ p95:     2.139 ms │     638 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ State Machine Transitions · ⚪ 49.378ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS — STATE MACHINE INTEGRITY                                  ║
║                        File: tests/benchmarks/state-machine-transition.test.ts                        ║
║                                       Ran: 2026-10-06 09:30:40                                        ║
║Measures state machine self-healing transition latencies, convergence settling time, and health probe consistency under stress.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ State Transition (Transient) │ Cold:    78.95 ms │ Warm:    49.378 ms │ p95:    63.072 ms │      20 RPS ║
║ Self-Healing Full Convergence │ Cold:    51.69 ms │ Warm:    48.446 ms │ p95:    54.433 ms │      21 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Database Adapter Raw CRUD · ⚪ 0.455ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                          SVELTYCMS — DATABASE ADAPTER PERFORMANCE (MARIADB)                           ║
║                          File: tests/benchmarks/database-performance.test.ts                          ║
║                                       Ran: 2026-10-06 16:24:21                                        ║
║          Measures raw CRUD performance, indexing efficiency, and connection pool resilience.          ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ INSERT                       │ Cold:     2.12 ms │ Warm:     0.443 ms │ p95:     0.752 ms │   2,085 RPS ║
║ FIND ONE                     │ Cold:     1.75 ms │ Warm:     0.376 ms │ p95:     0.544 ms │   2,521 RPS ║
║ FIND MANY (limit 50)         │ Cold:    49.91 ms │ Warm:     0.558 ms │ p95:     0.764 ms │   1,694 RPS ║
║ FILTER (blob field, 1 match) │ Cold:     9.03 ms │ Warm:     1.671 ms │ p95:     1.990 ms │     587 RPS ║
║ QUERY BUILDER LIST (50)      │ Cold:     6.79 ms │ Warm:     0.545 ms │ p95:     0.690 ms │   1,753 RPS ║
║ FIND PAGE (50 hasMore)       │ Cold:     3.40 ms │ Warm:     0.587 ms │ p95:     0.729 ms │   1,684 RPS ║
║ FIND PAGE keyset             │ Cold:     6.27 ms │ Warm:     0.661 ms │ p95:     0.870 ms │   1,442 RPS ║
║ FIND MANY offset 50          │ Cold:     3.47 ms │ Warm:     0.437 ms │ p95:     0.573 ms │   2,182 RPS ║
║ LIST+COUNT (legacy)          │ Cold:     3.23 ms │ Warm:     0.574 ms │ p95:     0.670 ms │   1,678 RPS ║
║ UPDATE                       │ Cold:     4.78 ms │ Warm:     0.711 ms │ p95:     0.955 ms │   1,364 RPS ║
║ UPDATE (no-returning)        │ Cold:     2.21 ms │ Warm:     0.396 ms │ p95:     0.519 ms │   2,474 RPS ║
║ NATIVE UPSERT                │ Cold:     2.38 ms │ Warm:     0.326 ms │ p95:     0.441 ms │   2,927 RPS ║
║ COUNT                        │ Cold:     2.28 ms │ Warm:     0.487 ms │ p95:     0.602 ms │   2,020 RPS ║
║ COUNT ESTIMATE               │ Cold:     2.86 ms │ Warm:     0.310 ms │ p95:     0.427 ms │   3,138 RPS ║
║ COUNT CACHED                 │ Cold:     3.72 ms │ Warm:     0.001 ms │ p95:     0.002 ms │ 723,903 RPS ║
║ DELETE (Active Rows)         │ Cold:    11.38 ms │ Warm:     0.397 ms │ p95:     0.567 ms │   2,282 RPS ║
║ BULK INSERT (100)            │ Cold:     6.07 ms │ Warm:     2.073 ms │ p95:     2.454 ms │     470 RPS ║
║ BULK UPDATE (100 heterogeneous) │ Cold:    22.97 ms │ Warm:     2.294 ms │ p95:     2.895 ms │     420 RPS ║
║ BULK UPDATE (100 homogeneous) │ Cold:    10.03 ms │ Warm:     1.293 ms │ p95:     1.542 ms │     758 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ ACID Transaction Overhead · ⚪ 1.724ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 81% — check connection pool saturation or lock contention. **Severe degradation** (+93%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                   SVELTYCMS — ACID INTEGRITY AUDIT                                    ║
║                            File: tests/benchmarks/transaction-acid.test.ts                            ║
║                                       Ran: 2026-10-06 09:30:17                                        ║
║         Measures transaction commit latencies and rollback overhead across database adapters          ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ TX Commit                    │ Cold:    12.07 ms │ Warm:     0.893 ms │ p95:     1.358 ms │   4,284 RPS ║
║ TX Rollback Integrity        │ Cold:     4.16 ms │ Warm:     1.724 ms │ p95:     2.261 ms │     579 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ System Cache Efficiency · ⚪ 0.949ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 69% — check connection pool saturation or lock contention. **Severe degradation** (+261%) — review recent commits immediately.

╔═════════════════════════════════════════════════════════════════════════════════════════╗
║                            NEGATIVE CACHE PERFORMANCE AUDIT                             ║
║                      File: tests/benchmarks/negative-cache.test.ts                      ║
║                                Ran: 2026-08-23 22:35:48                                 ║
║        Measures Bloom-filter style missing-key cache speedup for repeated misses        ║
╠═════════════════════════════════════════════════════════════════════════════════════════╣
║ First Miss (DB)                │        1.621 ms │ p95:        3.249 ms │ RPS:        612 ║
║ Cached Miss (Negative Cache)   │        0.001 ms │ p95:        0.002 ms │ RPS:    792,591 ║
╚═════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Cache Hit/Miss Ratio & Efficiency · ⚪ 1.805ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 73% — check connection pool saturation or lock contention. **Severe degradation** (+87%) — review recent commits immediately. > **Check**: `src/databases/cache/redis-adapter.ts` · `src/databases/cache/cache-service.ts`

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS — CACHE EFFICIENCY AUDIT                                   ║
║                           File: tests/benchmarks/cache-performance.test.ts                            ║
║                                       Ran: 2026-10-17 15:29:11                                        ║
║         Measures cache hit vs miss latency and efficiency via real HTTP end-to-end requests.          ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Cache Miss (Bypass)          │ Cold:     3.26 ms │ Warm:     0.777 ms │ p95:     1.021 ms │   1,234 RPS ║
║ Cache Hit (L1 Warm)          │ Cold:     0.89 ms │ Warm:     0.176 ms │ p95:     0.239 ms │   5,357 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ unknown · 🟢 0.002ms · ➡️ recorded (awaiting trend) · source

``` ╔═════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — CACHE EVICTION BOUNDARY HARDENING ║ ║ File: unknown ║ ║ Ran: 2026-08-23 22:39:30 ║ ╠═════════════════════════════════════════════════════════════════════════════════════════╣ ║ Cache Flooding & Eviction │ 0.002 ms │ p95: 0.002 ms │ RPS: 460,599 ║ ╚═════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Logic · 🟢 ⏳ pending · ➡️ ⚪ stable at 0ms (7 runs) (2026-09-18 18:18:35) · source

> ⏳ Pending

🏷️ SDK · 🟢 0.445ms · ➡️ recorded (awaiting trend) · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — SDK VS DIRECT ADAPTER (MARIADB) ║ ║ File: tests/benchmarks/local-sdk-vs-direct.test.ts ║ ║ Ran: 2026-10-06 10:53:47 ║ ║Measures the exact SDK tax across Create, Find, and Update pathways with isolated GC sweeps and zero-allocation key pools.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Direct insert │ 0.445 ms │ p95: 0.445 ms │ RPS: 2,246 ║ ║ SDK create (full) │ 0.539 ms │ p95: 0.539 ms │ RPS: 1,856 ║ ║ Direct findOne │ 0.383 ms │ p95: 0.383 ms │ RPS: 2,610 ║ ║ SDK find (bypassCache) │ 0.888 ms │ p95: 0.888 ms │ RPS: 1,126 ║ ║ Direct update │ 0.836 ms │ p95: 0.836 ms │ RPS: 1,196 ║ ║ SDK update (full) │ 1.308 ms │ p95: 1.308 ms │ RPS: 0 ║ ║ SDK create (detached) │ 0.377 ms │ p95: 0.377 ms │ RPS: 2,652 ║ ║ SDK update (detached) │ 1.270 ms │ p95: 1.270 ms │ RPS: 0 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ HTTP · 🟢 0.431ms · ↘️ ⚪ stable at 0.465ms (±0.127ms) (7 runs) (2026-09-18 18:18:35) · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

⏳ Pending

🏷️ Middleware · 🟢 2.589ms · ➡️ ⚪ established at 2.59ms (1 run) (2026-09-30 15:24:57) · source

> ⏳ Pending

🏷️ System · 🟢 0.421ms · ↘️ ⚪ stable at 0.483ms (±0.059ms) (2 runs) (2026-09-18 18:20:41) · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

⏳ Pending

🏷️ Metadata · 🟢 0.765ms · ➡️ ⚪ stable at 0.774ms (±0.100ms) (2 runs) (2026-09-18 18:20:41) · source

> ⏳ Pending

🏷️ CRUD · 🟢 0.535ms · ➡️ ⚪ established at 0.535ms (1 run) (2026-08-25 11:31:30) · source

> ⏳ Pending

🏷️ Content · 🟢 146.190ms · ➡️ recorded (awaiting trend) · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ CONTENT PACKAGE AUDIT (MARIADB) ║ ║ File: tests/benchmarks/content-package.test.ts ║ ║ Ran: 2026-10-06 10:53:10 ║ ║ Measures export/import throughput via ContentPackageService directly, ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Content Export (100 entries) │ Cold: 146.19 ms │ Warm: 101.018 ms │ p95: 116.549 ms │ 10 RPS ║ ║ Content Import Plan (100 entries) │ Cold: 350.37 ms │ Warm: 379.522 ms │ p95: 388.914 ms │ 3 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Analytics · 🟢 3.823ms · ➡️ ⚪ established at 3.82ms (1 run) (2026-08-23 19:46:55) · source

> ⏳ Pending

🏷️ HTTP-Cold · 🟢 0.562ms · ↗️ ⚪ stable at 0.528ms (±0.034ms) (2 runs) (2026-09-18 18:18:35) · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits.

⏳ Pending

🏷️ CRUD (Get) · 🟢 0.758ms · ➡️ ⚪ stable at 0.757ms (±0.086ms) (2 runs) (2026-09-18 18:20:42) · source

> ⏳ Pending

🏷️ CRUD (List) · ⚪ 0.801ms · ↘️ 🟢 faster: 0.889ms → 0.801ms (-10%) (2 runs) (2026-09-18 18:20:42) · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

⏳ Pending

🏷️ CRUD (Search) · 🟢 0.804ms · ➡️ ⚪ stable at 0.845ms (±0.269ms) (2 runs) (2026-09-18 18:20:43) · source

> ⏳ Pending

🏷️ COMPETITIVE 9-WORKLOAD REPLICA · ⚪ 0.815ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Avg degraded but p95 stable — possible cold-start effect, GC pause, or warmup variance. Throughput dropped 90% — check connection pool saturation or lock contention. **Severe degradation** (+205%) — review recent commits immediately.

╔═════════════════════════════════════════════════════════════════════════════════════════╗
║                        COMPETITIVE 9-WORKLOAD REPLICA (MARIADB)                         ║
║               File: tests/benchmarks/competitive-workload-replica.test.ts               ║
║                                Ran: 2026-08-18 20:18:47                                 ║
║Measures the exact 9 workloads tested in the external benchmark harness against the local server.║
╠═════════════════════════════════════════════════════════════════════════════════════════╣
║ findById (Concurrent 8c)       │        1.666 ms │ p95:        3.092 ms │ RPS:      2,770 ║
║ findByIdRandom (Concurrent 8c) │        5.785 ms │ p95:        9.554 ms │ RPS:      1,189 ║
║ listPlain (Concurrent 8c)      │        1.880 ms │ p95:        4.304 ms │ RPS:      2,586 ║
║ listFilterSort (Concurrent 8c) │        1.554 ms │ p95:        2.690 ms │ RPS:      2,875 ║
║ listLarge (Concurrent 8c)      │        1.659 ms │ p95:        2.757 ms │ RPS:      2,998 ║
║ findMissing (Concurrent 8c)    │        1.600 ms │ p95:        3.339 ms │ RPS:      3,038 ║
║ GraphQL Query (Concurrent 8c)  │        5.454 ms │ p95:        6.416 ms │ RPS:      1,366 ║
║ create (Concurrent 8c)         │       16.404 ms │ p95:       20.538 ms │ RPS:        473 ║
║ update (Concurrent 8c)         │       14.945 ms │ p95:       17.009 ms │ RPS:        493 ║
║ mixed (Concurrent 8c 50/50)    │        7.745 ms │ p95:       12.067 ms │ RPS:        716 ║
╚═════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ MIDDLEWARE FLAMEGRAPH PROFILER · ⚪ 1.690ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 54% — check connection pool saturation or lock contention. **Severe degradation** (+125%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                               MIDDLEWARE FLAMEGRAPH PROFILER (MARIADB)                                ║
║                         File: tests/benchmarks/middleware-flamegraph.test.ts                          ║
║                                       Ran: 2026-10-06 09:34:10                                        ║
║Measures isolated microsecond cost and incremental deltas of each middleware stage from edge to database.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Stage 0: Raw Socket Baseline (Static Asset) │ Cold:     2.64 ms │ Warm:     0.967 ms │ p95:     1.551 ms │   3,867 RPS ║
║ Stage 1: System State + Turbo Pipeline │ Cold:     2.62 ms │ Warm:     0.752 ms │ p95:     0.907 ms │   5,290 RPS ║
║ Stage 2: WAF Security Inspection │ Cold:     1.20 ms │ Warm:     0.717 ms │ p95:     0.914 ms │   5,220 RPS ║
║ Stage 3: Token Auth Gating   │ Cold:     6.50 ms │ Warm:     1.945 ms │ p95:     2.523 ms │   1,955 RPS ║
║ Stage 4: Collection Schema Resolution │ Cold:     9.49 ms │ Warm:     0.584 ms │ p95:     0.952 ms │   6,426 RPS ║
║ Stage 5: Full REST Single Document │ Cold:     2.93 ms │ Warm:     0.518 ms │ p95:     0.905 ms │   7,210 RPS ║
║ Stage 6: GraphQL Single Query (JIT) │ Cold:     4.92 ms │ Warm:     1.690 ms │ p95:     3.403 ms │   2,137 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ EVENT LOOP LAG & SATURATION · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ EVENT LOOP LAG & SATURATION (MARIADB) ║ ║ File: tests/benchmarks/event-loop-lag.test.ts ║ ║ Ran: 2026-10-06 10:53:12 ║ ║Measures event loop delay (p50, p95, p99, max) and microtask drain speed under concurrent write bursts.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Idle Baseline Lag │ 0.000 ms │ p95: 0.000 ms │ RPS: 1,454 ║ ║ Write Burst Lag (200 writes @ 16c) │ 0.000 ms │ p95: 0.000 ms │ RPS: 1,218 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Ingestion · 🟢 0.130ms · ↘️ ⚪ stable at 0.159ms (±0.062ms) (4 runs) (2026-10-04 18:01:00) · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

⏳ Pending

API · 9 tests

🏷️ Temporal Integrity Audit · ⚪ 3.579ms · ➡️ — · source

``` ╔═════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — TEMPORAL INTEGRITY ║ ║ File: tests/benchmarks/temporal-integrity.test.ts ║ ║ Ran: 2026-08-23 22:37:06 ║ ║Validates deterministic UTC normalization of ISO date strings across timezone offsets for consistent persistence.║ ╠═════════════════════════════════════════════════════════════════════════════════════════╣ ║ Timezone Ingestion │ 5.756 ms │ p95: 5.756 ms │ RPS: 174 ║ ╚═════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Relational & Nested Queries · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — RELATIONAL RESOLVER AUDIT ║ ║ File: tests/benchmarks/relational-performance.test.ts ║ ║ Ran: 2026-08-29 20:51:37 ║ ║Measures GraphQL resolver performance for shallow listings, deep relational joins, and nested author population across cold/hot execution paths.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Shallow Query (Collections) │ Cold: 2.77 ms │ Warm: 2.937 ms │ p95: 4.628 ms │ 1,945 RPS ║ ║ Deep Join (Cold DB) │ Cold: 2.95 ms │ Warm: 2.350 ms │ p95: 3.543 ms │ 1,568 RPS ║ ║ Deep Join (Warm Cached) │ Cold: 2.93 ms │ Warm: 1.870 ms │ p95: 3.362 ms │ 1,946 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Auth & RBAC Performance · ⚪ 1.121ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — AUTHENTICATION TELEMETRY ║ ║ File: tests/benchmarks/auth-performance.test.ts ║ ║ Ran: 2026-10-06 10:46:48 ║ ║ Evaluates authenticated session resolution, RBAC evaluation, and rejection paths. ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Unauthenticated (401 reject) │ Cold: 2.00 ms │ Warm: 0.874 ms │ p95: 1.212 ms │ 1,089 RPS ║ ║ Reject: unknown session cookie │ Cold: 1.14 ms │ Warm: 0.889 ms │ p95: 1.264 ms │ 1,069 RPS ║ ║ Reject: unknown API key (Bearer sck_) │ Cold: 0.99 ms │ Warm: 0.829 ms │ p95: 1.197 ms │ 1,146 RPS ║ ║ Auth Validation @ 1c │ Cold: 2.34 ms │ Warm: 0.831 ms │ p95: 1.236 ms │ 1,117 RPS ║ ║ HTTP Auth Pipeline @ 8c │ Cold: 2.55 ms │ Warm: 1.862 ms │ p95: 2.797 ms │ 4,118 RPS ║ ║ Auth (full chain) @ 1c │ Cold: 1.84 ms │ Warm: 0.359 ms │ p95: 0.468 ms │ 2,627 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ REST API Performance · ⚪ 0.707ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 26% — check connection pool saturation or lock contention. **Severe degradation** (+36%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                 SVELTYCMS — ENTERPRISE REST API AUDIT                                 ║
║                          File: tests/benchmarks/rest-api-performance.test.ts                          ║
║                                       Ran: 2026-10-06 09:26:54                                        ║
║Measures latency, throughput, and correctness of core REST endpoints: health check, schema, single read, list, and search.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ System Health Check          │ Cold:     1.22 ms │ Warm:     0.707 ms │ p95:     1.583 ms │   4,856 RPS ║
║ Collection Schema Metadata   │ Cold:     7.17 ms │ Warm:     0.518 ms │ p95:     0.796 ms │   7,320 RPS ║
║ Single Document Read         │ Cold:     2.89 ms │ Warm:     0.498 ms │ p95:     0.795 ms │   7,637 RPS ║
║ Collection Find (List Limit 20) │ Cold:     6.44 ms │ Warm:     0.765 ms │ p95:     1.465 ms │   4,651 RPS ║
║ Collection Search Query      │ Cold:     2.55 ms │ Warm:     0.626 ms │ p95:     0.999 ms │   6,007 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ GraphQL API Performance · ⚪ 0.860ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective. Throughput dropped 53% — check connection pool saturation or lock contention.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS — GRAPHQL CACHE HIT AUDIT                                  ║
║                           File: tests/benchmarks/graphql-cache-hit.test.ts                            ║
║                                       Ran: 2026-10-06 09:33:22                                        ║
║        Validates sub-millisecond L1 response cache hits, cache key isolation, and hit headers.        ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Q1: Health (Cold Miss)         │        5.263 ms │ p95:        5.263 ms │ RPS:        190 ║
║ Q1: Health (L1 Hit)            │        1.375 ms │ p95:        1.851 ms │ RPS:        727 ║
║ Q2: Collections (Cold Miss)    │        2.510 ms │ p95:        2.510 ms │ RPS:        398 ║
║ Q2: Collections (L1 Hit)       │        1.317 ms │ p95:        1.668 ms │ RPS:        759 ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Enterprise SEO Suite Audit · ⚪ 1.823ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Avg degraded but p95 stable — possible cold-start effect, GC pause, or warmup variance.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                   SVELTYCMS — ENTERPRISE SEO AUDIT                                    ║
║                            File: tests/benchmarks/seo-performance.test.ts                             ║
║                                       Ran: 2026-10-06 09:28:06                                        ║
║Measures redirect lookup latency, dynamic sitemap.xml generation, robots.txt serving, and 404 logging performance.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Redirect Lookup (301)        │ Cold:     1.87 ms │ Warm:     1.881 ms │ p95:     3.284 ms │   4,207 RPS ║
║ Dynamic Sitemap XML          │ Cold:  1416.60 ms │ Warm:     1.485 ms │ p95:     1.867 ms │   5,323 RPS ║
║ Robots.txt Generation        │ Cold:     2.12 ms │ Warm:     1.823 ms │ p95:     3.138 ms │   4,350 RPS ║
║ Missing Path (404 Log)       │ Cold:   120.30 ms │ Warm:   154.694 ms │ p95:   187.916 ms │      52 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ REAL-TIME BROADCAST AUDIT · ⚪ 0.211ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — YJS COLLABORATION SYNC AUDIT ║ ║ File: tests/benchmarks/websocket-broadcast.test.ts ║ ║ Ran: 2026-10-06 10:52:14 ║ ║Measures WebSocket handshake latency, bidirectional Yjs CRDT update propagation, and synchronization throughput.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ WS Handshake + SyncStep │ 9.630 ms │ p95: 0.000 ms │ RPS: 0 ║ ║ Yjs CRDT Propagation │ Cold: 3.73 ms │ Warm: 0.108 ms │ p95: 0.195 ms │ 8,576 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ WebSocket & Realtime Latency · ⚪ ⏳ pending · ➡️ — · source

> ⏳ **WebSocket & Realtime Latency**: Pending execution. > 🎯 Benchmarks WebSocket connection/broadcast latency.

🏷️ GraphQL Response Cache Hit · ⚪ 1.317ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS — GRAPHQL CACHE HIT AUDIT                                  ║
║                           File: tests/benchmarks/graphql-cache-hit.test.ts                            ║
║                                       Ran: 2026-10-06 10:53:14                                        ║
║        Validates sub-millisecond L1 response cache hits, cache key isolation, and hit headers.        ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Q1: Health (Cold Miss)         │        3.140 ms │ p95:        3.140 ms │ RPS:        318 ║
║ Q1: Health (L1 Hit)            │        1.047 ms │ p95:        1.260 ms │ RPS:        955 ║
║ Q2: Collections (Cold Miss)    │        2.670 ms │ p95:        2.670 ms │ RPS:        375 ║
║ Q2: Collections (L1 Hit)       │        1.073 ms │ p95:        1.562 ms │ RPS:        932 ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

Scale · 18 tests

🏷️ Negative Cache Performance · ⚪ 0.001ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═════════════════════════════════════════════════════════════════════════════════════════╗
║                            NEGATIVE CACHE PERFORMANCE AUDIT                             ║
║                      File: tests/benchmarks/negative-cache.test.ts                      ║
║                                Ran: 2026-08-23 22:35:48                                 ║
║        Measures Bloom-filter style missing-key cache speedup for repeated misses        ║
╠═════════════════════════════════════════════════════════════════════════════════════════╣
║ First Miss (DB)                │        1.621 ms │ p95:        3.249 ms │ RPS:        612 ║
║ Cached Miss (Negative Cache)   │        0.001 ms │ p95:        0.002 ms │ RPS:    792,591 ║
╚═════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Revision & History Growth · ⚪ 0.291ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                   SVELTYCMS — REVISION STRESS AUDIT                                   ║
║                            File: tests/benchmarks/revision-stress.test.ts                             ║
║                                       Ran: 2026-10-06 09:29:23                                        ║
║Measures latest read degradation, history retrieval speed, and storage footprint under heavy revision growth (100 revisions).║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Clean Document Read (Baseline) │ Cold:     3.87 ms │ Warm:     0.299 ms │ p95:     0.460 ms │  11,935 RPS ║
║ Latest Read (100 Revisions)  │ Cold:     2.31 ms │ Warm:     0.291 ms │ p95:     0.486 ms │  12,678 RPS ║
║ History List Retrieval       │ Cold:     8.90 ms │ Warm:     1.542 ms │ p95:     1.987 ms │   1,257 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Memory Stability & GC Behavior · ⚪ ⏳ pending · ➡️ — · source

> ⏳ **Memory Stability & GC Behavior**: Pending execution. > 🎯 Long-running soak test to identify memory leaks and GC pressure.

🏷️ Multi-Tenant Performance & Isolation · ⚪ 0.615ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — MULTI-TENANCY PERFORMANCE AUDIT ║ ║ File: tests/benchmarks/multi-tenant-performance.test.ts ║ ║ Ran: 2026-10-06 09:27:36 ║ ║Measures the overhead of tenant isolation, context switching, and data partitioning across many tenants.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Single Tenant Baseline │ Cold: 2.53 ms │ Warm: 0.639 ms │ p95: 0.885 ms │ 12,061 RPS ║ ║ Multi-Tenant Context Switching │ Cold: 0.78 ms │ Warm: 0.615 ms │ p95: 0.855 ms │ 12,340 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Realistic Mixed Workload · ⚪ 1.019ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — MIXED WORKLOAD AUDIT ║ ║ File: tests/benchmarks/mixed-workload.test.ts ║ ║ Ran: 2026-10-06 09:28:09 ║ ║Simulates real-world traffic with a weighted mix of reads, searches, GraphQL queries, and metadata requests with per-operation telemetry.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Mixed Workload │ Cold: 6.44 ms │ Warm: 1.019 ms │ p95: 2.912 ms │ 6,301 RPS ║ ║ Mix: REST Read │ 1.137 ms │ p95: 1.759 ms │ RPS: 1,476 ║ ║ Mix: REST Search │ 1.185 ms │ p95: 1.941 ms │ RPS: 499 ║ ║ Mix: GraphQL │ 3.104 ms │ p95: 4.402 ms │ RPS: 327 ║ ║ Mix: Metadata │ 1.211 ms │ p95: 2.045 ms │ RPS: 106 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ GraphQL Load Stress · ⚪ 13.698ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. **Severe degradation** (+118%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                SVELTYCMS — GRAPHQL CAPACITY DISCOVERY                                 ║
║                             File: tests/benchmarks/graphql-stress.test.ts                             ║
║                                       Ran: 2026-10-06 10:47:00                                        ║
║                Discovers maximum sustainable GraphQL throughput by ramping concurrency                ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ GQL: 5c Warmup               │ Cold:     2.27 ms │ Warm:     1.340 ms │ p95:     1.805 ms │   3,691 RPS ║
║ GQL: 10c Light               │ Cold:     2.46 ms │ Warm:     1.767 ms │ p95:     2.701 ms │   5,169 RPS ║
║ GQL: 20c Moderate            │ Cold:     2.04 ms │ Warm:     4.658 ms │ p95:     7.774 ms │   4,053 RPS ║
║ GQL: 40c Heavy               │ Cold:     2.09 ms │ Warm:     7.249 ms │ p95:    16.856 ms │   4,713 RPS ║
║ GQL: 60c Stress              │ Cold:     2.38 ms │ Warm:    12.050 ms │ p95:    22.640 ms │   4,827 RPS ║
║ GQL: 80c Extreme             │ Cold:     2.05 ms │ Warm:    15.753 ms │ p95:    23.495 ms │   4,993 RPS ║
║ GQL: 100c Max                │ Cold:     2.08 ms │ Warm:    18.042 ms │ p95:    35.112 ms │   5,074 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Bulk Migration & Scale · ⚪ 0.895ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective. Throughput dropped 39% — check connection pool saturation or lock contention.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS — MIGRATION & SCALE AUDIT                                  ║
║                            File: tests/benchmarks/migration-scale.test.ts                             ║
║                                       Ran: 2026-10-06 09:31:24                                        ║
║Measures bulk ingestion throughput for 10,000 entries, random point-lookups, and indexed range scans across large datasets.║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Bulk Migration (10k docs)    │ Cold:    51.29 ms │ Warm:    30.750 ms │ p95:    65.046 ms │  16,210 RPS ║
║ Random Point Lookup          │ Cold:     5.57 ms │ Warm:     2.526 ms │ p95:     4.326 ms │   2,192 RPS ║
║ Range Scan & Pagination      │ Cold:   102.95 ms │ Warm:     0.895 ms │ p95:     1.660 ms │   6,110 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Million-Row Index Pressure · ⚪ 0.892ms · ↗️ — · source

> [!NOTE] > **Specific to this test**: Both avg and p95 degraded — likely adapter or infrastructure bottleneck. Check DB connection pool, indexes, or recent commits. Throughput dropped 21% — check connection pool saturation or lock contention. **Severe degradation** (+35%) — review recent commits immediately.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                  SVELTYCMS  —  INDEX PRESSURE AUDIT                                   ║
║                             File: tests/benchmarks/index-pressure.test.ts                             ║
║                                       Ran: 2026-10-06 09:29:14                                        ║
║           Measures read performance with sorting and filtering on a large entry collection.           ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Sorted List (25k rows)       │ Cold:    22.82 ms │ Warm:     0.892 ms │ p95:     1.308 ms │   4,454 RPS ║
║ Filtered Query (25k rows)    │ Cold:     7.15 ms │ Warm:     0.575 ms │ p95:     0.951 ms │   6,887 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Content Scale Stress · ⚪ 0.154ms · ↘️ — · source

> [!NOTE] > **Specific to this test**: Performance improved — recent optimizations or cache warming likely effective.

╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗
║                                SVELTYCMS — CONTENT SCALE STRESS AUDIT                                 ║
║                          File: tests/benchmarks/content-scale-stress.test.ts                          ║
║                                       Ran: 2026-10-06 10:50:29                                        ║
║        Measures file-scanning and content discovery performance at scale (1,000+ collections).        ║
╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣
║ Cold Stress Scan (1000 files)  │        0.715 ms │ p95:        0.784 ms │ RPS:      1,399 ║
║ Warm Stress Scan (1000 files) │ Cold:     0.50 ms │ Warm:     0.135 ms │ p95:     0.166 ms │   7,188 RPS ║
╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝

🏷️ Media Upload Stress & Throughput · ⚪ 8.640ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — MEDIA UPLOAD STRESS AUDIT ║ ║ File: tests/benchmarks/media-upload-stress.test.ts ║ ║ Ran: 2026-10-06 09:31:11 ║ ║Measures throughput for large file uploads, concurrent transfers, and streaming efficiency with valid media fixtures.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Single Upload (0.01MB) │ Cold: 23.16 ms │ Warm: 8.336 ms │ p95: 10.365 ms │ 116 RPS ║ ║ Small Upload (0MB @ 4c) │ Cold: 8.79 ms │ Warm: 8.640 ms │ p95: 13.771 ms │ 441 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Full Client Journey Simulation · ⚪ 5.501ms · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — WORLD LIFE DATA ║ ║ File: tests/benchmarks/client-journey.test.ts ║ ║ Ran: 2026-10-06 10:48:54 ║ ║ Measures cumulative latency of a realistic editorial user workflow: Auth → List → View → Edit → Save. ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Full Journey @ 1c │ Cold: 31.55 ms │ Warm: 5.861 ms │ p95: 8.397 ms │ 170 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Concurrency & Race Condition · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — CONCURRENCY AUDIT ║ ║ File: tests/benchmarks/concurrency-race.test.ts ║ ║ Ran: 2026-10-06 09:31:31 ║ ║ Simulates high-concurrency writes to a single document to prove atomic consistency. ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Concurrent PATCH Bomb │ 28.452 ms │ p95: 51.989 ms │ RPS: 826 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Multi-Doc Concurrency Throughput · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — DIRECT ADAPTER THROUGHPUT (MARIADB) ║ ║ File: tests/benchmarks/local-api-throughput.test.ts ║ ║ Ran: 2026-10-06 09:29:25 ║ ║Measures direct adapter write throughput, read throughput, and percentile latencies with zero HTTP overhead.║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ Writes (1000 atomic increments) │ 1.099 ms │ p95: 1.849 ms │ RPS: 6,684 ║ ║ Reads (1000 point lookups) │ 1.174 ms │ p95: 2.564 ms │ RPS: 10,794 ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Max Throughput — No Throttle · ⚪ 32.280ms · ➡️ — · source

> ⏳ **Max Throughput — No Throttle**: Pending execution. > 🎯 1000 concurrent writes across 100 docs — no semaphore.

🏷️ Failure Propagation & Fast-Fail Audit · ⚪ ⏳ pending · ➡️ — · source

``` ╔═══════════════════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — AI INFRASTRUCTURE AUDIT ║ ║ File: tests/benchmarks/ai-performance.test.ts ║ ║ Ran: 2026-10-06 10:48:29 ║ ║ Measures the latency tax of internal CMS logic for AI enrichment and layout generation. ║ ╠═══════════════════════════════════════════════════════════════════════════════════════════════════════╣ ║ AI Text Enrichment │ Cold: 0.06 ms │ Warm: 0.000 ms │ p95: 0.001 ms │ 1,217,039 RPS ║ ║ AI Layout Spec Generation │ Cold: 0.04 ms │ Warm: 0.000 ms │ p95: 0.001 ms │ 1,116,071 RPS ║ ╚═══════════════════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Chaos & Infrastructure Resilience · ⚪ ⏳ pending · ➡️ — · source

``` ╔═════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — CHAOS RESILIENCE AUDIT ║ ║ File: tests/benchmarks/chaos-resilience.test.ts ║ ║ Ran: 2026-07-26 10:34:44 ║ ║Simulates infrastructure brownouts and measures system stability and graceful degradation under stress.║ ╠═════════════════════════════════════════════════════════════════════════════════════════╣ ║ Chaos Resilience (Brownout) │ 197.326 ms │ p95: 734.025 ms │ RPS: 5 ║ ╚═════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Production Day Simulation · ⚪ 2.195ms · ➡️ — · source

> ⏳ **Production Day Simulation**: Pending execution. > 🎯 Full production day with mixed workload, 100% reliability.

Resilience · 5 tests

🏷️ Circuit Breaker Failover · ⚪ ⏳ pending · ➡️ — · source

> ⏳ **Circuit Breaker Failover**: Pending execution. > 🎯 Measures system graceful degradation when external services fail.

🏷️ Throttling & Backoff Stress · ⚪ ⏳ pending · ➡️ — · source

``` ╔═════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — THROTTLING AUDIT ║ ║ File: tests/benchmarks/throttling-backoff-stress.test.ts ║ ║ Ran: 2026-07-21 16:13:32 ║ ║Simulates high-velocity traffic from a single client IP to verify rate-limiting enforcement and backoff consistency.║ ╠═════════════════════════════════════════════════════════════════════════════════════════╣ ║ Throttling Enforcement │ 2.555 ms │ p95: 5.292 ms │ RPS: 391 ║ ╚═════════════════════════════════════════════════════════════════════════════════════════╝ ```

🏷️ Data Residency & Sovereignty · ⚪ ⏳ pending · ➡️ — · source

> ⏳ **Data Residency & Sovereignty**: Pending execution. > 🎯 PII blocking and data sovereignty enforcement with cross-region failover.

🏷️ Database Failover & Reconnection · ⚪ ⏳ pending · ➡️ — · source

``` ╔═════════════════════════════════════════════════════════════════════════════════════════╗ ║ SVELTYCMS — DATA RESIDENCY AUDIT ║ ║ File: tests/benchmarks/data-residency-failover.test.ts ║ ║ Ran: 2026-07-26 10:36:42 ║ ║Simulates geopolitical boundary crossing and verifies PII field blocking for residency law compliance.║ ╠═════════════════════════════════════════════════════════════════════════════════════════╣ ║ PII Field Blocking │ 18.143 ms │ p95: 22.896 ms │ RPS: 55 ║ ╚═════════════════════════════════════════════════════════════════════════════════════════╝ ```

🔬 Host environment

Host	Platform	Arch	Cores	RAM	Runtime
RKroells	win32	x64	24	63GB	Bun 1.4.2