Methods and apparatus for serving content to customer devices based on dynamic content popularity
Methods and apparatus for processing requests for content received from customer devices are described. A decision on how to respond to a request for content is made at an edge node based on a locally maintained popularity list. The local popularity list reflects the local popularity of individual pieces of content at the edge node. A white list of content to be cached and served irrespective of popularity is sometimes used in combination with the local popularity list to make decisions as to how to respond to individual requests for content. In some embodiments the edge node decides on one of the following responses to a content request: i) cache and serve the requested content; ii) serve but don't cache the requested content; or iii) redirect the content request to another node, e.g., an alternate serving node, which can respond to the content request.
1. A method of operating a first edge node in a communications network, the method comprising:
receiving, by the first edge node, a request for content;
determining that requested content is not available in a cache which is accessible to the first edge node;
determining, by the first edge node, a response to be taken in response to said request for content based on local popularity of the requested content, said step of determining a response to be taken including determining that requested content is below a serve but don't cache threshold on a local popularity list but above a redirect threshold, said response being a serve but don't cache response action; and
implementing, by the first edge node, the determined serve but don't cache response action by obtaining the requested content from another node and serving the requested content without caching the requested content in said cache.
2. The method of claim 1 , further comprising:
receiving general content popularity information, prior to receiving said request for content; and
initializing local content popularity information using content popularity values obtained from said general content popularity information, said content popularity values being content popularity ratings.
3. The method of claim 2 , further comprising:
updating local content popularity information based on requests for content received by said first edge node.
4. The method of claim 3 , further comprising:
receiving updated general content popularity information; and
updating the local content popularity information based on information included in the updated general content popularity information.
5. The method of claim 4 , wherein the update of the local content popularity information based on requests for content received by said first edge node is performed at a faster rate than the updating of the local content popularity information based on updated general content popularity information.
6. The method of claim 1 , wherein determining the response to be taken is based on the redirect threshold, requests for content corresponding to content having a popularity on a local content popularity list below the redirect threshold being redirected.
7. The method of claim 1 , wherein determining the response to be taken is based on a serve and cache threshold, requests for content corresponding to content having a popularity on a local content popularity list above the serve and cache threshold being responded to by caching and serving the requested content.
8. The method of claim 6 , wherein determining the response to be taken is also based on a serve and cache threshold, requests for content corresponding to content having a popularity on the local content popularity list at or below the serve and cache threshold and above the redirect threshold being responded to by serving but not caching the requested content.
9. The method of claim 1 , further comprising:
generating a cache efficiency metric based on a number of received content requests in a time period which can be served by already cached content relative to a number of content requests received in the time period which can not be served by already cached content; and
making a decision whether to modify a local popularity threshold used in making the response determination based on the generated cache efficiency metric.
10. The method of claim 1 , further comprising:
prior to determining the response to be taken in response to a request for content based on the local popularity of the requested content, determining that the content requested by said request for content is not on a white list of content which is to be served and cached irrespective of local popularity.
11. A first edge node in a communications network, comprising:
a receiver configured to receive a request for content;
a processor configured to:
determine a response to be taken in response to said request for content based on local popularity of the requested content, said step of determining a response to be taken including determining that requested content is below a serve but don't cache threshold on a local popularity list but above a redirect threshold, said response being a serve but don't cache response action;
and implement the determined serve but don't cache response action by obtaining the requested content from another node and serving the requested content without caching the requested content in said cache.
12. The first edge node of claim 11 , wherein the receiver is further configured to receive general content popularity information, prior to receiving said request for content; and
wherein said processor is further configured to initialize local content popularity information using content popularity values obtained from said general content popularity information, said content popularity values being content popularity ratings.
13. The first edge node of claim 12 , wherein said processor is further configured to update local content popularity information based on requests for content received by said first edge node.
14. The first edge node of claim 13 ,
wherein said receiver is further configured to receive updated general content popularity information; and
wherein said processor is further configured to update the local content popularity information based on information included in the updated general content popularity information.
15. The first edge node of claim 14 , wherein said processor is further configured to update the local content popularity information based on requests for content received by said first edge node at a faster rate than a rate of updating the local content popularity information based on updated general content popularity information.
16. The first edge node of claim 11 , wherein said processor is further configured to determine the response to be taken based on a redirect threshold, said processor being configured to control the first edge node to redirect requests for content corresponding to content having a popularity on a local content popularity list below the redirect threshold.
17. The first edge node of claim 11 , wherein said processor is further configured to determine the response to be taken based on a serve and cache threshold, said processor being configured to respond to requests for content corresponding to content having a popularity on a local content popularity list above the serve and cache threshold by controlling the first edge node to cache and serve the requested content.
18. The first edge node of claim 16 , wherein said processor is further configured to determine the response to be taken based on a serve and cache threshold, said processor being configured to respond to requests for content corresponding to content having a popularity on the local content popularity list at or below the serve and cache threshold and above the redirect threshold by controlling the first edge node to serve but not cache the requested content.
19. The first edge node of claim 18 , wherein said processor is further configured to:
generate a cache efficiency metric based on a number of received content requests in a time period which can be served by already cached content relative to a number of content requests received in the time period which can not be served by already cached content; and
make a decision whether to modify a local popularity threshold used in making the response determination based on the generated cache efficiency metric.
20. The first edge node of claim 11 , wherein said processor is further configured to determine that the content requested by said request for content is not on a white list of content which is to be served and cached irrespective of local popularity, prior to determining the response to be taken in response to a request for content based on the local popularity of the requested content.