Bandwidth allocation for a data path

Abstract

A compute engine allocates data path bandwidth among different classes of packets. The compute engine identifies a packet's class and determines whether to transmit the packet based on the class' available bandwidth. If the class has available bandwidth, the compute engine grants the packet access to the data path. Otherwise, the compute engine only grants the packet access to the data path if none of the other packets waiting for data path access have a class with available bandwidth. After a packet is provided to the data path, the compute engine decrements a bandwidth allocation count for the packet's class. Once the bandwidth count for each class is exhausted, the compute engine sets each count to a respective starting value-reflecting the amount of bandwidth available to a class relative to the other classes. A compute engine employing the above-described bandwidth allocation can be employed to perform different networking services, including but not limited to: 1) virtual private networking; 2) secure sockets layer processing; 3) web caching; 4) hypertext mark-up language compression; 5) virus checking; 6) firewall support; and 7) web switching.

1 . A method for allocating bandwidth of a data path among packets, said method comprising the steps of: 
 (a) identifying a class value for a packet, wherein said class value corresponds to a first class in a set of classes; and    (b) determining whether to grant said packet access to said data path, based on said class value identified in said step (a), wherein said step (b) includes the steps of: 
 (1) determining whether allocated bandwidth for said first class has been exhausted, and  
 (2) determining whether access to said data path is requested for any packets with a class value having available allocated bandwidth and corresponding to a class in said set of classes, other than said first class.