Cache by status code
Customers can set cache time-to-live (TTL) based on the response status from the origin web server. Cache TTL refers to the duration of a resource in the Cloudflare network before being marked as STALE
or discarded from cache. Status codes are returned by a resource's origin.
Setting cache TTL based on response status overrides the default cache behavior (standard caching) for static files and overrides cache instructions sent by the origin web server. To cache non-static assets, set a Cache Level of Cache Everything using a Cache Rule. Setting no-store
Cache-Control or a low TTL (using max-age
/s-maxage
) increases requests to origin web servers and decreases performance.
The maximum caching limit for Free, Pro, and Business customers is 512 MB per file, and the maximum caching limit for Enterprise customers is 5 GB per file. If you need to raise the limits, contact your Customer Success Manager.
By default, Cloudflare caches certain HTTP response codes with the following Edge Cache TTL when a cache-control
directive or expires
response header are not present.
HTTP status code | Default TTL |
---|---|
200, 206, 301 | 120m |
302, 303 | 20m |
404, 410 | 3m |
All other status codes are not cached by default.
To set cache TTL by response status, create a Cache Rule for Cache TTL by status code.
curl --request PUT \"https://api.cloudflare.com/client/v4/zones/{zone_id}/rulesets/{ruleset_id}" \--header "Authorization: Bearer <API_TOKEN>" \--header "Content-Type: application/json" \--data '{ "rules": [ { "expression": "(http.host eq \"www.example.com\")", "description": "set cache TTL by response status", "action": "set_cache_settings", "action_parameters": { "cache": true, "edge_ttl": { "status_code_ttl": [ { "status_code_range": { "to": 299 }, "value": 86400 }, { "status_code_range": { "from": 300, "to": 499 }, "value": 0 // no-cache }, { "status_code_range": { "from": 500 }, "value": -1 // no-store } ], "mode": "respect_origin" } } } ]}'
Provide a JSON object containing status codes and their corresponding TTLs. Each key-value pair in the cache TTL by status cache rule has the following syntax:
status_code
: An integer value such as 200 or 500.status_code
matches the exact status code from the origin web server. Valid status codes are between 100-999.status_code_range
: Integer values forfrom
andto
.status_code_range
matches any status code from the origin web server within the specified range.value
: An integer value that defines the duration an asset is valid in seconds or one of the following strings:no-store
(equivalent to-1
),no-cache
(equivalent to0
).
The cacheTtlByStatus option is a version of the cacheTtl feature that designates a cache TTL for a request’s response status code (for example, { "200-299": 86400, 404: 1, "500-599": 0 }
).
-
If a TTL is not explicitly set for status code
304
, we automatically set it to match the TTL of status code200
(if the user has defined one for200
). -
If a user explicitly sets a different TTL for
304
than for200
, the following behavior will occur:
- When a
200
response is received, the asset is cached with the TTL specified for status200
. - Once the asset expires and we revalidate with the origin, if the origin returns a
304
, the cache TTL is updated to the value set for304
.
For example, if a user specifies a TTL of one hour for status 200
and 0 seconds (cache and always revalidate) for status 304
, the asset will be cached for 1 hour. After it expires, we revalidate with the origin. If the origin returns a 304
, each subsequent request will trigger revalidation. If the origin continues to return 304
, this cycle will persist.
This behavior is likely undesirable unless the user has a specific use case. Therefore, users should ensure that the TTL for 304
matches the TTL for 200
unless they intentionally require this behavior.