Page Cache

Enable Caching Engine
Caches WooCommerce catalog and product pages for anonymous visitors, dramatically reducing TTFB. Automatically bypasses dynamic contexts such as Cart, Checkout, My Account, admin, previews, and other non-cacheable requests. Logged-in users are never served from page cache.

Page Cache TTL (seconds)
Defines how long an HTML page remains fresh in the page cache before being regenerated in the background. Longer TTL reduces server load; shorter TTL increases freshness. Default: 7,200 seconds (2 hours).

Browser Cache TTL (seconds)
Sets a fallback browser-side max-age for pages when no cache headers are present. Helps browsers avoid constant re-fetches and makes 304 revalidation more efficient.

Include Accept-Language in Vary Headers
Adds Accept-Language to the response Vary header so multilingual sites using header-based language detection return the correct variants. Disable if your site varies by path or domain to reduce cache fragmentation.

---

Store API Cache

Store API Server Cache TTL (seconds)
How long Store API responses (e.g., product lists, prices, catalog data) are cached server-side. This reduces API load and speeds up headless experiences, Mini-Cart, and Blocks UI. Targeted invalidation keeps data fresh after product or stock changes.

Store API Browser TTL (seconds)
Client-side TTL for Store API responses. Lower values keep the UI more responsive to changes, while higher values reduce network requests. Best used in combination with server TTL for balanced performance.

---

CDN Cache

CDN Cache TTL (seconds)
Sets s-maxage for CDNs and proxies when Mamba emits Cache-Control. Higher values shift more traffic to the edge; use alongside tag-based invalidation to ensure content stays correct. Default: 7,200 seconds (2 hours).

Stale-If-Error TTL (seconds)
Allows CDNs and proxies to serve stale content if the origin is unavailable or returns a 5xx error. Improves uptime during outages. Default: 86,400 seconds (24 hours).

---

Cache Management

Cache Size Limit (MB)
Caps the total on-disk size of the HTML cache and its metadata. When the limit is exceeded, automatic cleanup removes older entries to stay within bounds.

---

Warmup

Warmup Concurrency
Controls the number of parallel requests used by the warmup job. Higher values speed up cache priming after purges but place more immediate load on the origin.

Warmup Batch Size
Sets the number of URLs processed per warmup tick. Larger batches complete faster but use more memory, while smaller batches are gentler on resources.