Cisco Technology, Inc.

Other

Ketan Arun Padwekar

Dipankar Bhattacharya

Rina Panigrahy

Yiren Huang

Raymond Kloth

Yung-Chin Chen

Abhijit V Warkhedi

Various systems and methods that synchronize local copies of a variable used in a distributed algorithm, such as that used in distributed policing, are disclosed. For example, one system includes several distributed algorithm participants that are coupled to communicate with each other. Each of the each of the distributed algorithm participants maintains a shared variable in order to implement the distributed algorithm. One of the distributed algorithm participants is a synchronizer. The synchronizer is configured to calculate a new value of the shared variable and to communicate the new value to each of the other distributed algorithm participants. The distributed algorithm can be a distributed policing algorithm, and the shared variable can be a global bucket.

System and method for performing distributed policing

A method is disclosed for preventing compressed keys from being aliased across multiple hash tables. A special hashing technique is used to produce hash keys and compressed keys that make such aliasing avoidable. Given a first input key K, a second and different input key K', a first hash function Hthat produces a first hash key as an index into a first hash table, a second hash function Hthat produces a second hash key as an index into a second hash table, a first compression function Lthat produces a first compressed key for the first hash table, and a second compression function Lthat produces a second compressed key for the second hash table, functions HHLand Lare selected in such a way that H(K)=H(K&#x2032;) and L(K)=L(K&#x2032;) if and only if H(K)=H(K&#x2032;) and L(K)=L(K&#x2032;).

Preventing aliasing of compressed keys across multiple hash tables

Various systems and methods that synchronize local copies of a variable used in a distributed algorithm, such as that used in distributed policing. For example, one system includes several distributed algorithm participants that are coupled to communicate with each other. Each of the each of the distributed algorithm participants maintains a shared variable in order to implement the distributed algorithm. One of the distributed algorithm participants is a synchronizer. The synchronizer is configured to calculate a new value of the shared variable and to communicate the new value to each of the other distributed algorithm participants. The distributed algorithm can be a distributed policing algorithm, and the shared variable can be a global bucket.

An apparatus and method for implementing comprehensive QOS independent of the fabric system is disclosed. According to one embodiment, a system is provided comprising a switching medium, a plurality of line cards, a first scheduler associated with an ingress line card, and a second scheduler associated with an egress line card and communicatively coupled with the first scheduler. According to another embodiment, a line card is provided comprising a plurality of input buffers, a plurality of output buffers, output buffer status logic coupled to the output buffers operable to produce output buffer status information corresponding to each output buffer, a first scheduler coupled to the input buffers and operable to select data from the input buffers based upon the output buffer status, and a second scheduler coupled to the output buffers and operable to select data from the output buffers in accordance with Quality of Service requirements.

Apparatus and method for implementing comprehensive QoS independent of the fabric system

The present invention defines a method of unbiased policing of data flow in a network device. According to an embodiment of the present invention, the token bucket policer of the network device 'permits' (forwards) incoming packets even when the size of the token bucket is less than the size of the incoming packets. Permitting incoming packets that are larger than the token bucket ensures that incoming packets are not dropped because of the size of the incoming packets. Incoming packets are policed by TBP when the magnitude comparison of the token bucket and a predetermined constant value does not comply with the policing scheme defined for the incoming packets. When a packet is 'permitted' (forwarded), the size of the token bucket is reduced by an amount equal to the size of the packet.

Unbiased token bucket

A method for policing traffic on a computer communications network having a multitude of nodes interconnected by various communications media. An individual policer is established at each node for monitoring and/or policing the traffic incoming to that node. Traffic policy parameters are established for traffic-classes and the policy is implemented at each individual policer. Thresholds may be established and when the thresholds are met or exceeded the individual policer will export the traffic conditions at the respective node. The other individual policers or a master policer will receive the exported information. -The individual policers police the traffic incoming to its associated node depending on the traffic condition information received from all the nodes. Several classes may be handled by each individual policer. Leaky bucket algorithms may be used in some instances.

Growing community of inventors

Saratoga, CA, United States of America

Ketan Arun Padwekar