Vmware, Inc.

Jin Heo

Jui-Ting Weng

Boon Seong Ang

Wenyi Jiang

Guolin Yang

Yong Wang

Lenin Singaravelu

Ayyappan Veeraiyan

Srividya Murali

Shilpi Agarwal

Daniel Michael Hecht

Kalyan Saladi

Chien-Chia Chen

Reza Taheri

Jeffrey Buell

Dongdong Deng

Some embodiments provide a method for a first smart NIC of multiple smart NICs of a host computer. Each of the smart NICs executes a smart NIC operating system that performs virtual networking operations for a set of data compute machines executing on the host computer. The method receives a data message sent by one of the data compute machines executing on the host computer. The method performs virtual networking operations on the data message to determine that the data message is to be transmitted from a port of a second smart NIC of the multiple smart NICs. The method passes the data message to the second smart NIC via a private communication channel connecting the plurality of smart NICs.

Teaming of smart NICs

Some embodiments of the invention provide a method for configuring a physical network card or physical network controller (pNIC) to provide flow processing offload (FPO) for a host computer connected to the pNIC. The host computers host a set of compute nodes in a virtual network. The set of compute nodes are each associated with a set of interfaces that are each assigned a locally-unique virtual port identifier (VPID) by a flow processing and action generator. The pNIC includes a set of interfaces that are assigned physical port identifiers (PPIDs) by the pNIC. The method includes providing the pNIC with a set of mappings between VPIDs and PPIDs. The method also includes sending updates to the mappings as compute nodes migrate, connect to different interfaces of the pNIC, are assigned different VPIDs, etc. In some embodiments, the flow processing and action generator executes on processing units of the host computer, while in other embodiments, the flow processing and action generator executes on a set of processing units of a pNIC that includes flow processing hardware and a set of programmable processing units.

Configuring pNIC to perform flow processing offload using virtual port identifiers

Some embodiments provide a method for synchronizing state between multiple smart NICs of a host computer that perform operations using dynamic state information. At a first smart NIC of the plurality of smart NICs, the method stores a set of dynamic state information. The method synchronizes the set of dynamic state information across a communication channel that connects the smart NICs so that each of the smart NICs also stores the set of dynamic state information.

State sharing between smart NICs

Some embodiments provide a method for a first smart NIC of multiple smart NICs of a host computer. Each of the smart NICs is for performing virtual networking operations for a set of data compute machines executing on the host computer. The method determines that the first smart NIC is elected to communicate with a network management and control system that configures the virtual networking operations. The method receives a set of configuration data for the virtual networking operations from the network management and control system. The method provides the received set of configuration data to the other smart NICs of the host computer.

Smart NIC leader election

Configuring PNIC to perform flow processing offload using virtual port identifiers

A method of optimizing network processing in a system comprising a physical host and a set of physical network interface controllers (PNICs) is provided. The physical host includes a forwarding element. The method includes determining that a set of conditions is satisfied to bypass the forwarding element for exchanging packets between a particular data compute node (DCN) and a particular PNIC. The set of conditions includes the particular DCN being the only DCN connected to the forwarding element and the particular PNIC being the only PNIC connected to the forwarding element. The method exchanges packets between the particular DCN and the particular PNIC bypassing the forwarding element. The method determines that at least one condition in said set of conditions is not satisfied. The method utilizes the forwarding element to exchange packets between the particular DCN and the particular PNIC.

Specializing virtual network device processing to avoid interrupt processing for high packet rate applications

Some embodiments of the invention provide a method for providing flow processing offload (FPO) for a host computer at a physical network interface card (pNIC) connected to the host computer. A set of compute nodes executing on the host computer are each associated with a set of interfaces that are each assigned a locally-unique virtual port identifier (VPID) by a flow processing and action generator. The pNIC includes a set of interfaces that are assigned physical port identifiers (PPIDs) by the pNIC. The method includes receiving a data message at an interface of the pNIC and matching the data message to a stored flow entry that specifies a destination using a VPID. The method also includes identifying, using the VPID, a PPID as a destination of the received data message by performing a lookup in a mapping table storing a set of VPIDs and a corresponding set of PPIDs and forwarding the data message to an interface of the pNIC associated with the identified PPID.

Flow processing offload using virtual port identifiers

In one embodiment, a method selects a percentage of a plurality hosts that are coupled together via a network fabric and calculates a number of workloads needed for the percentage of hosts based on a benchmark test to run. A plurality of data compute nodes are configured on one or more host pairs in the percentage of the plurality of hosts to send and receive the number of workloads through the network fabric to perform the benchmark test. A set of measurements is received for sending and receiving the workloads through the network fabric using the plurality of data compute nodes. The method increases the percentage of the plurality of hosts until the set of measurements fails a criteria or the percentage of the plurality of hosts is all of the plurality of hosts.

Control of a computing system to perform network fabric benchmark measurements

In a system that includes a physical forwarding element (PFE), the mapping of each possible destination MAC address to a destination port number of the PFE is stored in a mapping table. A destination port number on the PFE is identified for a first packet in a sequence of packets arrived at a source port of the PFE by searching the mapping table and stores the destination port number in a caching table. For each packet after the first packet, when the destination MAC address of the packet matches the destination MAC address stored in the caching table, the destination port number stored in the caching table is used as the destination port number of the packet. Otherwise, a destination port number for the packet is identified by searching the mapping table and the destination port number stored in the caching table is replaced with the identified destination port number.

Fast software L2 switching using a caching technique

Specializing virtual network device processing to bypass forwarding elements for high packet rate applications

The current document is directed to virtualized PMUs provided by virtualization layers. The currently disclosed virtualized PMUs are decoupled from the underlying PMU hardware features of processors on which the virtualization layer executes. The decoupling is achieved, in part, by time multiplexing the underlying hardware PMU registers to provide a greater number of virtualized PMU registers than the number of hardware-PMU registers provided by at least some of the underlying hardware PMUs. The decoupling is also achieved by providing for monitoring, by the virtualized PMU registers, of computed processor events and approximated processor events in addition to the processor events monitored by the underlying hardware PMUs. In addition, the virtualized PMU registers are implemented, in certain implementations, to support a variety of different monitoring modes, including monitoring of processor events that occur only during execution of the virtualization layer and monitoring of hardware-thread-specific processor events.

Virtual performance monitoring decoupled from hardware performance-monitoring units

A method of high packet rate network processing in a system that includes a physical host and a set of physical network interface controllers (PNICs). The physical host is hosting a set of data compute nodes (DCNs). Each DCN includes a virtual network interface controller (VNIC) for communicating with one or more PNICs to exchange packets. The method determines that a rate of packets received from a particular DCN at the VNIC of the particular DCN exceeds a predetermined threshold. The method performs polling to determine the availability of packets received at the VNIC from the particular DCN while the rate of packets received from the DCN at the VNIC is exceeding the threshold. The method utilizes interrupts to determine the availability of packets received at the VNIC from the particular DCN while the rate of packets received from the DCN at the VNIC does not exceed the threshold.

Exemplary methods, apparatuses, and systems determine a virtual processing unit utilization value representing utilization for a first virtual machine of a plurality of virtual machines running on a host computer. A host processing unit utilization value representing utilization for the host computer including the plurality of virtual machines running on the host computer is also determined. A target coalescing rate is selected based upon the virtual processing unit utilization and host processing unit utilization values. A coalescing rate or a coalescing depth for the first virtual machine is updated based upon the selected target coalescing rate.

Growing community of inventors

Mountain View, CA, United States of America

Jin Heo