After you add a website to Edge Security Acceleration (ESA), ESA points of presence (POPs) determine whether to cache resources that are requested by clients based on configured cache rules. After a POP caches a file, when clients request the file, the POP responds the file to clients without retrieving it from the origin server over a long route. This reduces latency and improves load times. If the requested file does not exist on the POP or has expired, the POP asks the origin server for the most recent file.
Features
You can configure the following caching features for a website.
Default cache rules
Feature | Description | |
When you use ESA to accelerate the delivery of static resources on the origin server, ESA caches the resources on the ESA point of presence (POP) that is nearest to the client. Then, you can access and obtain the static resources from the POP instead of the origin server. This prevents origin pulls over a long route and speeds up content delivery. All POPs of ESA are equipped with caching systems. When a user or an origin server interacts with a ESA POP, the caching system caches resources from the origin server and specifies a Time-to-Live (TTL) for the cached resources. | ||
Global cache configurations
Feature | Description |
After you enable the development mode, all requests are redirected to the origin server. This allows requests to your website to temporarily bypass the caching components of Edge Security Acceleration (ESA) so that you can verify changes to the cached content. This feature is useful when you want to view changes in real time. | |
You can configure browser cache time-to-live (TTL) without modifying the configuration of the origin server. During the TTL, the browser loads the files from the cache for subsequent requests. This speeds up page loading. | |
The edge cache time-to-live (TTL) is the period of time during which origin resources are cached on Edge Security Acceleration (ESA) points of presence (POPs). When the TTL ends, resources that are cached on POPs are marked as expired. If the requested resource has expired on a POP, the POP retrieves the most recent resource from the origin server and caches it. You can configure a cache TTL for static resources based on file directories or file name extensions. | |
You can specify whether points of presence (POPs) ignore the question mark ( | |
After you turn on Sort Query Strings, Edge Security Acceleration (ESA) automatically sorts the query strings in the URLs when processing requests. Then, ESA returns the requested resources from the cache or redirects the requests to the origin server based on the sorted query strings. This way, POPs return the same file for requests that contain the same parameters and values, regardless of the order of query strings in the request URLs. This feature improves the cache hit ratio. |
Cache rules
Feature | Description |
If you want to use different browser cache TTLs in different business scenarios, you can create multiple cache rules and specify different cache policies for the rules. | |
If you want to use different edge cache TTLs in various business scenarios, you can create multiple cache rules and specify different cache settings in the rules. | |
When a point of presence (POP) retrieves a resource from the origin server, the origin server returns an HTTP status code. You can configure a cache time-to-live (TTL) for the HTTP status code in the Edge Security Acceleration (ESA) console. When a client requests the same resource, the POP returns the HTTP status code without triggering origin fetch. This reduces loads on the origin server. After a cached HTTP status code expires, requests that trigger the code are redirected to the origin server. | |
You can create custom cache keys based on different parts of requests, such as the query string, HTTP request headers, or cookies. This way, you can set a cache key for a type of requests that are destined for the same file to prevent the file from being cached as different files. This increases the cache hit ratio and reduces origin requests, response time, and bandwidth usage. |
Purge and prefetch
Feature | Description |
You can use the purge cache feature to clear cached resources from points of presence (POPs) and retrieve the most recent ones from the origin server. This feature is useful when you update content to or remove illicit content from your origin server. Purging a large number of cached resources in a short period of time results in a significant increase in origin requests, which, in turn, increases loads on the origin server. | |
The prefetch feature of Edge Security Acceleration (ESA) proactively fetches resources from an origin server to points of presence (POPs). This way, ESA can directly serve the resources from the POPs even when the resources are requested for the first time, without the need to retrieve them from the origin server. The prefetch feature increases the cache hit ratio. |
Cache reserve
Feature | Description | |
ESA provides the tiered cache feature to ensure that requests are served as fast as possible from POPs closest to clients, improving the performance and efficiency of the overall content delivery system. | ||
Edge Security Acceleration (ESA) provides the cache reserve feature to converge origin requests. | ||
Feature availability
Cache | Entrance | Pro | Premium | Enterprise |
Cache analytics | Not supported | Supported | Supported | Supported |
Browser cache TTL | 2 hours | 1 hour | 1 second | 1 second |
Edge cache TTL | 2 hours | 1 hour | 1 second | 1 second |
Purge by URL (times) | 1,000 | 50,000 | 100,000 | 500,000 |
Purge all cache | 5 | 80 | 125 | 200 |
Purge by directory (times) | Not supported | 20 | 50 | 2,000 |
Purge by hostname | Not supported | Not supported | Not supported | 2,000 |
Custom cache key | Not supported | Not supported | Not supported | Shares quota with URL purge. |
Purge resources with parameters ignored | Not supported | Not supported | Not supported | 2,000 |
Purge by tag | Not supported | Not supported | Not supported | 2,000 |
URL prefetch | Not supported | 50,000 | 100,000 | 500,000 |
Edge tiered cache | Supported | Supported | Supported | Supported |
Edge tiered cache + regional tiered cache | Not supported | Supported | Supported | Supported |
Edge tiered cache + smart tiered cache | Not supported | Supported | Supported | Supported |
All cache layers | Not supported | Not supported | Not supported | |
Cache reserve | Not supported | Not supported | Not supported | Supported |