Method and apparatus for performing predictive caching of DNS requests by correlating IP addresses
View Patent ↗Content on the World Wide Web is made available for downloading by distributing it to cache providers. The content provider is a distinct entity from the cache providers and enters into contractual arrangements in order to benefit from the services provided by the cache providers. The contractual arrangements specify certain rights in cache resources that are bought and sold as commodities, exemplarily through a broker or exchange. An exemplary such right is the right to a stated amount of bandwidth, in a stated time interval, for servicing download requests. Responsive to such contractual arrangements, a redirective element such as a DNS server is programmed to redirect, to one or more cache providers, download requests initially addressed to the content provider.
1. A method for use in an IP packet-based data network, the method comprising the steps of:
receiving first and second IP packets and extracting corresponding first and second IP addresses therefrom;
determining a first set of one or more domain names associated with said first IP address and a second set of one or more domain names associated with said second IP address;
correlating the first set of domain names with the second set of domain names based upon a corresponding correlation of a time of receipt of said first IP packet with a time of receipt of said second IP packet;
receiving a third IP packet, extracting a third IP address therefrom, and determining a third set of one or more domain names associated with said third IP address;
identifying said first set of domain names as including one or more domain names included in said third set of domain names; and
performing one or more DNS requests relating to one or more of said domain names included in said second set of domain names based on said correlation of said second set of domain names with said identified first set of domain names.
2. The method of claim 1 wherein said IP packet-based data network comprises the Internet.
3. The method of claim 1 wherein each of said IP packets are received at a firewall.
4. The method of claim 1 wherein said first, second and third IP packets are all associated with a given browser.
5. The method of claim 4 wherein the given browser has a browser address, wherein each of said first and second IP addresses and each of said determined first and second sets of domain names are tagged with said browser address, and wherein said step of correlating the first set of domain names with the second set of domain names is based on both sets having been tagged with said browser address.
6. The method of claim 1 wherein each of said sets of domain names are determined based on correspondingly derived domain name association trees which include said corresponding associated IP address.
7. The method of claim 6 wherein said derivation of each of said domain name association trees comprises performing a reverse A-record lookup based on said corresponding associated IP address to identify one or more domain names to be included in said domain name association tree.
8. The method of claim 7 wherein said derivation of each of said domain name association trees further comprises identifying one or more CNAME-records related to said identified one or more domain names to identify one or more additional domain names to be included in said domain name association tree.
9. The method of claim 8 wherein said identifying one or more CNAME-records related to said identified one or more domain names comprises searching a DNS cache for said related CNAME-records.
10. The method of claim 7 wherein said derivation of each of said domain name association trees further comprises heuristically generating and checking one or more domain names based on said identified one or more domain names to identify one or more additional domain names to be included in said domain name association tree.