performance-tuning

Optimize Adobe Dispatcher and HTTPD configuration for AEMaaCS cloud deployments with baseline and runtime verification.

1|Updated Jun 22, 2026
One-click install
npx skills add https://github.com/aemgdc/aemdev --skill performance-tuning-aemgdc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: performance-tuning
Source: https://github.com/aemgdc/aemdev/tree/main/.agents/skills/performance-tuning
Command: npx skills add https://github.com/aemgdc/aemdev --skill performance-tuning-aemgdc

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Dispatcher and Apache HTTPD configurations in AEM as a Cloud Service often suffer from low cache hit ratios, redirect chains, and tail-latency hotspots, and tuning them without cloud-specific guardrails risks breaking validator contracts or managed defaults. ## Core Features & Use Cases - Baseline-to-After Verification: Captures metrics and cache evidence before changes, then compares post-change results using the core-7 MCP tools (validate, lint, sdk, trace_request, inspect_cache, monitor_metrics, tail_logs). - Cloud Guardrails Enforcement: Preserves immutable includes, reserved probe paths, and CDN-vs-Dispatcher ownership boundaries required by AEMaaCS validators. - Scenario Playbooks: Provides eight tuning playbooks covering cache hit ratio, invalidation blast radius, redirect latency, static assets, query-parameter fragmentation, tail latency, persisted queries, and cache operations. - Use Case: When a site's cache hit ratio drops below 70%, use this Skill to identify top MISS URLs, apply targeted cache-rule changes, validate them against cloud contracts, and prove the improvement with before/after metrics. ## Quick Start Ask the assistant to analyze the dispatcher configuration for low cache hit ratio and propose verified performance optimizations for the cloud deployment.

Frequently Asked Questions about performance-tuning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I improve the AEM Dispatcher cache hit ratio?▼

Capture baseline metrics with monitor_metrics, identify top MISS URLs via tail_logs and trace_request, then apply targeted cache-rule changes such as TTL extensions or ignoreUrlParams tuning. Validate changes with validate and lint, then confirm improvement with inspect_cache and post-change metrics comparison.

How to reduce redirect chain latency in Apache HTTPD config?▼

Trace redirect-heavy URLs with trace_request, eliminate multi-hop chains by pointing rules directly to final destinations, and use RewriteMap for large redirect sets. Re-validate with validate and lint, then compare latency trends before and after the change.

Does this performance tuning work for AMS or on-premise AEM Dispatcher?▼

No, this Skill is scoped to AEMaaCS cloud deployments only and requires the Dispatcher MCP configured with AEM_DEPLOYMENT_MODE=cloud. Its guardrails, validator contracts, and probe-path rules are specific to the cloud variant.

What AEMaaCS guardrails apply when tuning Dispatcher cache rules?▼

You must preserve managed default includes such as default_rules.any and default_filters.any, keep reserved probe paths like /systemready and /system/probes/* untouched, and route edge or WAF concerns to the CDN layer. Changes require validate, lint, and sdk check-files evidence before sign-off.

Why is my Dispatcher cache fragmented by query parameters?▼

Cache fragmentation happens when marketing or tracking query parameters create unique cache keys for identical content. Tighten the /ignoreUrlParams allow-list so only business-relevant parameters affect caching, and mirror the marketing_query_parameters.any pattern for common tracking params.

Can I enable GraphQL persisted query caching in AEMaaCS Dispatcher?▼

Yes, but first confirm persisted-query traffic is a real hotspot, then align CORS and preflight handling before enabling CACHE_GRAPHQL_PERSISTED_QUERIES. Verify one persisted-query path and one non-cacheable control path with trace_request and inspect_cache, keeping a rollback plan for the toggle.