Intel Corporation

Google Inc.

Alok Kumar

Uday Ramakrishna Naik

Sushant Jain

Prashant R Chandra

Raj S Yavatkar

Aspi Siganporia

Ned M Smith

Vincent R Scarlata

Faraz A Siddiqi

Anand Raghuraman

Tasneem Brutch

Michael E Kounavis

Willard M Wiseman

Michael Alex Frumkin

Anshuman Gupta

Ameya S Varde

Eswar Eduri

Sridhar Lakshmanamurthy

Chen-Chi Kuo

Murari Kumar

Pierre-Emmanuel Ettori

Parveen Patel

Harrick M Vin

Sonny Ben-Shimon

Noam Lampert

Idan Brown

Ittai Balaban

Daniel Berkovitch

Uri Goren

Kalpana M Roge

David W Grawrock

Amin Vahdat

Donald F Hooper

Junlan Zhou

Arjun Singh

Gargi Adhav

Harshawardhan Vipat

Subhasree Mandal

Emilie Jeanne Anne Danna

Madhukar Tallam

Steven Padgett

Glen Anderson

Kuo-Lang Tseng

Sreedhar Mukkamalla

Avinatan Hassidim

Yuquan Jiang

James Wanderer

Ramesh Thomas

Jonathan Zolla

David Romano

Dan Raz

David Champagne

Shashank Ashok Bapat

Michal Segalov

Yossi Richter

Chang-Ming P Lin

Subbaiah Naidu Kotla Venkata

Pavlin Radoslavov

Minal B Patel

Marco Leogrande

Salomon Sonny Ben-Shimon

Mukta Gupta

Shijeesh Nharappadath Sankaranathan

Kannan Sattainathan

Ujjwal Jain

Aakash Bhushan Arora

Benjamin John Benardos

Anjin Guo

Vijay Chandramohan

Willard M (Monty) Wiseman

Aneet Chopra

Harsha L Narayan

Matthew Jonathan Holliman

David Chou

Rui Zhang-Shen

Benjamin C Hardekopf

Subramanian Anand

Uday Ramakrishma Naik

Aspi Homi Siganporia

Aspects of the disclosure are directed to a software defined network (SDN) having a sharded control plane. The SDN may include a host device and a sharded control plane. The sharded control plane may include a first controller and a second controller sharded by one or more dimensions. The first controller and the second controller may be configured to process requests received from the first host device based on their respective sharded one or more dimensions. The one or more dimensions may be networks or functions.

Sharded SDN control plane with authorization

The present disclosure provides for full mesh connectivity between all endpoints in a VPC, including virtual machines, load balancers, routers, interconnects, virtual private networks, etc. Connectivity may be extended to on-premises devices, such as those connected via VPN and Interconnects. The connectivity is high performance, reliable, and secure.

Full mesh proxyless connectivity between networks

A method for a proxyless protocol includes intercepting, from a client, a first Transmission Control Protocol (TCP) connection request requesting to establish a TCP connection between the client and a proxy for routing data to a destination server. The request includes client information and a first sequence number. The method also includes transmitting, to the destination server, a second TCP connection request to establish a TCP connection between the client and the destination server. The second request includes a second sequence number less than the first sequence number. The method also includes intercepting, from the destination server, an acknowledgment message indicating acknowledgment of the second TCP connection request. The method also includes transmitting, to the destination server, a proxy protocol header message with the client information. The method also includes generating a new acknowledgment message using the first TCP connection request and transmitting, to the client, the new acknowledgment message.

Proxyless protocol

Aspects of the disclosure provide for a proxyless NAT infrastructure with dynamic port allocation. A proxyless NAT infrastructure is configured to perform NAT between a network of virtual machines (VMs) and a device external to the network, without a device, such as a NAT server or a router, acting as a proxy. A system can include a control plane for provisioning VMs of a network, including configuring each VM to perform NAT and initially assigning a number of ports for communicating with other devices. The control plane maintains a feedback loop-receiving data characterizing port usage and network traffic at ports allocated to the various VMs and scaling the port allocation for each VM based on the received data. The control plane can allocate additional ports as determined to be needed by a VM, and later retrieve the ports to be reused for other VMs.

Proxyless network address translation with dynamic port allocation

The present invention pertains to the allocation of bandwidth on a global large scale network. Systems and methods are provided to base bandwidth allocation on the predicted bandwidth demands of the network users. In allocating bandwidth, each user is assigned a user weight value, and is to be allocated bandwidth in proportion to that value. In addition, the bandwidth allocation is enforced so as to ensure that bandwidth allocation policies are maintained.

System to share network bandwidth among competing applications

Systems and methods for achieving high utilization of a network link are provided. A first communication protocol can be selected for transmitting network flows of a first type. A first quality of service can be assigned to network flows of the first type. A second communication protocol can be selected for transmitting network flows of a second type. A second quality of service, lower than the first quality of service, can be assigned to network flows of the second type. A first percentage of available bandwidth can be allocated to the network flows of both the first and second types. The remaining bandwidth, plus a second percentage of available bandwidth, can be allocated to the network flows of the second type, such that the total allocated bandwidth exceeds the available bandwidth of the network link.

Systems and methods for achieving high network link utilization

A system and method are provided that allocate bandwidth on a global large scale network. Bandwidth allocation is based on the predicted bandwidth demands of the network users. Each user may be assigned one of a plurality of different values that determines the amount of bandwidth allocated to that user. In instances where an application runs on behalf of a plurality of remote clients, a system and method is provided that allows for the allocation of bandwidth based each individual remote client.

Sharing bandwidth among multiple users of network applications

A traffic engineering process seeks to optimize network utilization and creating paths in the network such that the network is used efficiently. The exemplary embodiments seek to maintain fairness across allocation of flows while honoring constraints on the number of paths used per flow and the granularity of splits allowed. Exemplary embodiments provide a solution in a speedy manner and get close to optimal solutions for allocation of the network traffic.

Traffic engineering for network usage optimization

Systems and methods are provided for allocating resources of a network among a plurality of traffic demands to optimize fairness and network utilization. Methods based on flow-increase dynamics converge toward an upward max-min fair (UMMF) allocation, in which the value of each traffic demand cannot be increased, along any of its paths, even if larger traffic demands are removed from the network. An efficient iterative algorithm that converges to a UMMF solution is also provided. The described methods and systems can be implemented efficiently, distributively, and asynchronously.

Iterative max-min fairness algorithms

Systems and methods are provided for allocating network resources across network flow aggregates to apply fairness at different levels of flow aggregation. Utility functions of member flows in each aggregate may be adjusted based on a target aggregate utility function. Processes for allocating network resources may then be performed using the adjusted member utility functions.

Hierarchical fairness across arbitrary network flow aggregates

Exemplary traffic engineering solutions are described herein. According to exemplary embodiments, multiple tiered thresholds representing allocatable capacities of links in the network may be provided. The tiered thresholds may effectively limit the amount of data that a traffic engineering system is permitted to allocate on links in the network. The traffic engineering system may attempt to allocate data flows in the network according to the smallest threshold. If the network cannot accommodate the data flows according to the smallest threshold, then the next-smallest threshold may be attempted. The thresholds may be sequentially tested in increasing order until the largest threshold is attempted. If a workable threshold is found, then the data flow may be allocated to a path in the network. If the traffic engineering system is unable to accommodate the data flows at any of the tiered thresholds, then the traffic engineering system may report a failure.

Traffic engineering system for preventing demand deadlock and achieving uniform link utilization

Load balancing in a server pool may be achieved by rebalancing subsets of backend servers allocated to frontend servers. The rebalancing may be achieved by determining a max and min subset of servers and transferring backend servers between the subsets.

Systems and methods for proxy-less load balancing

Systems and methods are described herein for validating operation of a centrally managed network. A central controller is configured to manage a network, and an emulator configured to emulate one or more nodes in the network is connected to the central controller. The emulator determines a state value for one or more nodes in the network, and the state value represents whether the respective node is in error. The emulator receives a first plurality of policies including instructions for the nodes in the network, and whether the policies are consistent with one another is identified.

Systems and methods for testing a central controller in a centrally managed network

Growing community of inventors

Santa Clara, CA, United States of America

Alok Kumar