Verizon Patent and Licensing Inc.

Mindspeed Technologies, Inc.

Michael A Tsukernik

Paul Michael Curtis

Joseph B Tompkins

Bruce Burns

Chitz Shung

David A Tassie

Jamie Mulderig

A device may store a data structure containing entries and times associated with the entries. The device may determine a fullness level associated with the data structure. The device may determine an expiration threshold based on the fullness level of the data structure. The device may identify a first group of entries, of the entries in the data structure, based on times associated with the first group of entries satisfying the expiration threshold. The device may remove the first group of entries from the data structure. The device may randomly select, based on a probability associated with the fullness level, a second group of entries of the entries in the data structure. The device may remove the second group of entries from the data structure.

Dynamic removal of MAC table entries based on a MAC table fullness level

A source network device of a cloud computing network receives a packet destined for a destination virtual machine provided in a destination cloud computing device. The packet is received from a source virtual machine provided in a source cloud computing device. The source network device associates, to the packet, a source shaping resource of the source network device, where the source shaping resource includes a bandwidth that matches or exceeds a bandwidth associated with the source virtual machine. The source network device adds, to the packet, a header that identifies an address of the source virtual machine, an identifier associated with the source shaping resource, and the bandwidth associated with the source virtual machine. The source network device provides, via the source shaping resource, the packet and the header to the cloud computing network for transmission to the destination virtual machine.

Learning information associated with shaping resources and virtual machines of a cloud computing environment

A device stores shaping resource information in a data structure, where the shaping resource information includes identifiers associated with shaping resources of the device, information indicating whether the shaping resources are available, and information indicating transfer speeds associated with the shaping resources. The device defines a group of available shaping resources based on the shaping resource information stored in the data structure, and receives a packet requesting a particular transfer speed associated with transmitting the packet. The device performs a search of the group of available shaping resources based on the particular transfer speed requested by the packet, and identifies, based on the search, a list of one or more available shaping resources with transfer speeds that match the particular transfer speed requested by the packet. The device transmits the packet toward a destination via one of the one or more available shaping resources provided in the list.

Allocating shaping resources for a cloud computing environment based on transfer speeds associated with the shaping resources

A device may receive a packet, via a first port, that identifies a source network address. The device may determine a second port, identified in a data structure, that is associated with the source network address. The second port may be different from the first port. The device may determine an age indicator, stored in the data structure, associated with the source network address and the second port. The device may compare the age indicator to a threshold, and may selectively perform a first action or a second action based on comparing the age indicator to the threshold. The first action may include maintaining, in the data structure, a first indication of a first relationship between the source network address and the second port. The second action may include storing, in the data structure, a second indication of a second relationship between the source network address and the first port.

Determining when to relearn a relationship between a network address and a port

Various systems and methods for buffer memory management are disclosed. In one embodiment a buffer memory includes at least one queue configured to store a number of buffer access tasks. Buffer reclamation logic is executed to free at least one segment of the buffer memory holding an amount of stale data. Buffer reclamation logic is also included that enables the buffer reclamation logic to submit a buffer access task to the buffer memory based upon a total number of the buffer access tasks stored in the at least one queue.

Growing community of inventors

Newton, MA, United States of America

Michael A Tsukernik