Glossary

Terms used across Cache Rocket and these docs, plus the general caching vocabulary they build on.

Cache Rocket terms#

API warmer

A warmer for REST or GraphQL endpoints rather than HTML pages. Managed under Account → API Warmers, with its own plan limits.

Auto start interval

How often, in seconds, a warmer begins a new run.

Cache warmer

A saved configuration that crawls your website over HTTP to keep its pages cached. Also called a website warmer or, in older parts of the interface, a crawler.

Config matrix

Dimensions of values on an API endpoint that expand into many warm requests — for example every locale crossed with every tenant.

Cold URL

A URL that stays slow even after being warmed, usually because it is not cacheable.

Crawl

One pass of a warmer over your URLs. Used interchangeably with *warm run*.

Depth

How many link-hops from an entry URL a warmer will travel. Entry URLs are depth 0.

Dimension

One axis of an API config matrix, of kind QUERY, HEADER, COOKIE, or PATH.

Enqueue interval

How often URLs are re-queued within a warming cycle. Distinct from the auto start interval.

Entitlement

A capability your plan does or does not include, as opposed to a numeric limit.

Entry URL

A starting point for a warm run. Must use a verified hostname.

First load

The response time of the first of the two requests Cache Rocket makes to a URL. Represents a cold visit.

Hostname

A domain registered in Cache Rocket, such as www.example.com. Must be verified before it can be warmed.

Idempotent

Safe to call more than once with no side effects. Only idempotent API endpoints should use the cold/warm double fetch.

Priority warm

Warming a specific list of URLs immediately, rather than running a full pass.

Purge → rewarm

Clearing URLs at your CDN and immediately warming them again, so the cache is refreshed instead of merely emptied.

Warm load

The response time of the immediate second request to a URL. Represents what visitors experience after warming.

Warm run

One execution of a warmer, from seeding the queue to finishing or hitting a limit.

Warm schedule window

Hours during which scheduled auto-starts are allowed. Outside them, scheduled runs are skipped.

Warmth score

The share of tracked URLs that look properly cached, on Cache health.

Workspace

An organisation owning its own warmers, hostnames, alerts, and integrations, with its own members and roles.

Managed CDN

Cache Rocket’s asset CDN at assets.cacherocket.com, used to deliver cloud-optimized images, Critical CSS, and LQIP.

LQIP

Low-quality image placeholder — a tiny blurred preview shown while the full image loads.

Critical CSS

The CSS required to render above-the-fold content quickly, generated in the cloud and served from Managed CDN.

Cloud image optimization

Server-side conversion of uploads to modern formats (WebP/AVIF) with delivery via Managed CDN.

General caching terms#

Cache

A stored copy of a response, kept so the same request can be answered without rebuilding it.

Cache hit

A request answered from cache. Fast, and it never touches your application.

Cache miss

A request with no usable cached copy, which must be built by the origin.

Cache key

What a cache uses to decide whether two requests are “the same”. Often the URL, sometimes also headers, cookies, or device class.

CDN

Content Delivery Network. Servers distributed worldwide that cache your content close to visitors.

Cold start

The slow first request after a cached copy expires, while the response is rebuilt.

Edge

The CDN servers nearest your visitors, as opposed to your origin.

Invalidate

Mark a cached copy as no longer valid, so the next request rebuilds it. Similar to purging.

Origin

Your own server — the authoritative source a CDN pulls from when it has no cached copy.

POP

Point of Presence. One geographic location in a CDN's network, with its own cache.

Purge

Explicitly remove cached copies, usually because the content changed.

Reverse proxy

A server in front of your application, such as nginx or Varnish, often caching responses.

SSR

Server-Side Rendering. Generating a page's HTML on the server for each request.

Stale

A cached copy that no longer matches the current content.

TTFB

Time To First Byte. How long before the first byte of a response arrives — the number cold caches hurt most.

TTL

Time To Live. How long a cached copy stays valid before it expires. Your warming frequency should relate to this.