Revalidation and request collapsing
Revalidation is a caching mechanism that checks the freshness of cached data before serving it to users. If a cached object is no longer fresh and Cloudflare receives a request for it, the system makes a request to the origin to revalidate the object in the Cloudflare cache. By using headers like If-Modified-Since and ETag, Cloudflare validates content without fully re-fetching it. When these headers are missing, Smart Edge Revalidation generates a Last-Modified header, ensuring efficient updates and delivery of fresh content while reducing origin traffic.
For stale (expired TTL) content, Cloudflare will send a revalidation request to the origin. If the stale content is still valid Cloudflare will set a new TTL. If the content is expired, then the origin will provide new fresh content to replace the old.
Consider the following example scenarios.
One-thousand (1,000) requests arrive simultaneously at Cloudflare's network, and the requested resource was in Cloudflare but cannot be served because its TTL configuration indicates it is no longer fresh. This means that the resource needs to be served from the origin server. In this case, one request will go to the origin to be revalidated, while the other 999 requests will be served from cache with the status of UPDATING. This means that the resource, although expired, is served stale from Cloudflare's cache, while the origin server is updating it. This behavior is defined by the stale-while-revalidate directive in cache-control. If you do not wish to serve stale content, set the directive to zero seconds, stale-while-revalidate=0.
One-thousand (1,000) requests arrive simultaneously at a single Cloudflare data center, and the requested asset is not in Cloudflare's cache (a cache miss). These requests will use a cache lock to communicate with your origin. This means that only the first request will go to origin to fetch the asset. The remaining 999 requests wait for the first request to fetch the data, after which the response is streamed ↗ to all the waiting requests. The cache lock ensures that Cloudflare only sends the origin one request at a time for a given asset from a location in Cloudflare's network, preventing the origin from getting too much traffic.
When both Last-Modified ↗ and Etag ↗ headers are absent from the origin server response, Smart Edge Revalidation will use the time the object was cached on Cloudflare's global network as the Last-Modified header value. When a browser sends a revalidation request to Cloudflare using If-Modified-Since or If-None-Match, our global network can answer those revalidation questions using the Last-Modified header generated from Smart Edge Revalidation. In this way, our global network can ensure efficient revalidation even if the headers are not sent from the origin.