Optimizing encrypted wide area network traffic

Abstract

Optimization of encrypted traffic flowing over a WAN is provided by an arrangement in which WAN compression is distributed between endpoints (i.e., client machines or servers) in a subnet of a hub and branch network and a WAN compression server in the subnet. A client portion of the WAN compression running on each of one or more endpoints interfaces with a disposable local cache of data seen by endpoints in the subnet that is used for compressing and decompressing traffic using dictionary-based compression techniques. The local WAN compression server in a subnet stores a shared central database of all the WAN traffic in the subnet which is used to populate local disposable caches in the endpoints.

1 . A method for distributing WAN optimization between a server and an endpoint, the server and endpoint being located in a subnet of a hub and branch network, the method comprising the steps of:
 storing, in a traffic database on the server, data indicative of traffic observed flowing in to and out of the subnet;   storing, in a central dictionary store on the server, dictionaries usable by the endpoint to respectively compress and decompress traffic flowing out of and in to the subnet; and   sending data from one of the traffic database or central dictionary store to the endpoint, the endpoint being arranged to compress outbound traffic before being encrypted and further arranged to decompress inbound traffic after being decrypted.   
     
     
         2 . The method of  claim 1  including a further step of creating a tunnel to a remote server for non-encrypted traffic. 
     
     
         3 . The method of  claim 1  in which the sending is performed responsively to a request from the endpoint, or is performed as an update from the server to the endpoint. 
     
     
         4 . A method performed by devices, including an endpoint and a central server in a subnet of a hub and branch network, to optimize performance of a WAN link between the hub and branch, the method comprising the steps of:
 storing in a database on the central server traffic data representative of traffic flowing in to and out of the subnet;   caching traffic data in a cache at the endpoint, the cached traffic data being a subset of the traffic data stored in the database;   periodically refreshing the traffic data in the cache with traffic data from the database; and   optimizing compression and decompression, respectively, of outbound and inbound traffic at the endpoint using the cached traffic data.   
     
     
         5 . The method of  claim 4  including further steps of storing in a dictionary store on the central server one or more dictionaries usable for compressing and decompressing, respectively, outbound and inbound traffic at the endpoint, caching a subset of the dictionaries in a dictionary cache in the endpoint, and optimizing compression and decompression, respectively, of outbound and inbound traffic at the endpoint using the cached subset of dictionaries. 
     
     
         6 . The method of  claim 5  including a further step of periodically refreshing the cached subset of dictionaries with dictionaries from the dictionary store. 
     
     
         7 . The method of  claim 5  including a further step of periodically wiping out dictionaries in the cached subset of dictionaries, the wiped out dictionaries being unused for a period of time meeting a predetermined threshold.