Vmware, Inc.

Nicira, Inc.

Chidambareswaran Raman

Subin Cyriac Mathew

Raju Koganty

Dileep Kumar Devireddy

Senthilkumar Karunakaran

Hongwei Zhu

Sreenivas Duvvuri

Alexander Tessmer

Sami Boutros

Subrahmanyam Manuguri

Rishi Kanth Alapati

Akhila Naveen

Stephen Tan

Fenil Kavathia

Ganesh Sadasivan

Parasuramji Rajendran

Anirban Sengupta

Rahul Mishra

Margaret Angeline Petrus

Anuprem Chalvadi

Aditi Vutukuri

Feng Gu

Weiming Xu

Shireesh Kumar Singh

Ganesan Chandrashekhar

Todd Sabin

Meenakshi Sundaram Selvaraj

Prerit Rodney

Mansi Babbar

Hua Wang

Xin Wang

Donghai Han

Jianjun Shen

Wenyi Jiang

Shadab Shah

Jingmin Zhou

Yusheng Wang

Xin Qi

Jayant Jain

Mani Kancherla

Nithin Bangalore Raju

Abhishek Mishra

Yashika Narang

Philip Kippen

Dharmaraja Rajan

Pallavi Moghe

Srinivas Nimmagadda

Mayank Agarwal

Jerome Catrouillet

Aditya Gajanana Holla

Ankit Parmar

Li Sun

Jeremy Olmsted-Thompson

Nishant Jain

Neha Repal

Amit Aneja

Naveen Vijayan Kodakkal Puthiyaveettil

Sichen Song

Examples described herein include efficient data packet transmission between virtual machines ('VMs') on different hosts. An example method includes generating a large data packet at a source VM and determining a modified maximum segment size for efficient transmission. This modified size replaces the default maximum segment size through a TSO MSS override. Segmentation occurs based on the modified size, and the data segments are transmitted to the destination VM, even if on a different host. Dynamic determination of the modified size optimizes data transmission efficiency and network performance. It accounts for network headers and enables efficient transmission with or without large receiving offload (“LRO”) support. Additionally, non-transitory computer-readable media and servers implementing the method are disclosed. These systems and methods achieve streamlined data transmission, improving network performance and reducing processing overhead.

Underlay-based TCP segmentation offload in overlay networks

Some embodiments provide a novel method for forwarding data messages between first and second host computers. To send, to a first machine executing on the first host computer, a flow from a second machine executing on the second host computer, the method identifies a destination network address of the flow. The method uses the identified destination network address to identify a particular tunnel endpoint group (TEPG) including a particular set of one or more tunnel endpoints (TEPs) associated with a particular non-uniform memory access (NUMA) node of a set of NUMA nodes of the first host computer. The particular NUMA node executes the first machine. The method selects, from the particular TEPG, a particular TEP as a destination TEP of the flow. The method sends the flow to the particular TEP of the particular NUMA node of the first host computer to send the flow to the first machine.

Bringing NUMA awareness to load balancing in overlay networks

Some embodiments provide a method of multicasting data in a segregated logical network with multiple network segments. The network segments each have at least one router. The multicast originates from a multicast source in a first network segment and goes to multicast receivers in several other network segments. In the method, the router of the first network segment receives encoding data for a set of virtual tunnel endpoints (VTEPs) of the receivers. The router then receives the multicast data from the multicast source. For each VTEP, the router then encodes the data for the receiver using the encoding data for the VTEP of the receiver. The router then sends (e.g., as a unicast) the encoded data to a router of the network segment of the receiver. The router performing the method may be a domain router of the network segment of the receiver.

Distributing multicast receiver information across multi-tier edge gateways

Example methods and systems for virtual tunnel endpoint (VTEP) mapping for overlay networking are described. One example may involve a computer system monitoring multiple VTEPs that are configured for overlay networking. In response to detecting a state transition associated with a first VTEP from a healthy state to an unhealthy state, the computer system may identify mapping information that associates a virtualized computing instance with the first VTEP in the unhealthy state; and update the mapping information to associate the virtualized computing instance with a second VTEP in the healthy state. In response to detecting an egress packet from the virtualized computing instance to a destination, an encapsulated packet may be generated and sent towards the destination based on the updated mapping information. The encapsulated packet may include the egress packet and an outer header identifying the second VTEP to be a source VTEP.

Virtual tunnel endpoint (VTEP) mapping for overlay networking

Some embodiments provide a method for forwarding multicast data messages at a forwarding element on a host computer. The method receives a multicast data message from a routing element executing on the host computer along with metadata appended to the multicast data message by the routing element. Based on a destination address of the multicast data message, the method identifies a set of recipient ports for a multicast group with which the multicast data message is associated. For each recipient port, the method uses the metadata appended to the multicast data message by the routing element to determine whether to deliver a copy of the multicast data message to the recipient port.

Optimized processing of multicast data messages in a host

A method for offloading multicast replication from multiple tiers of edge nodes implemented by multiple host machines to a physical switch is provided. Each of the multiple host machines implements a provider edge node and a tenant edge node. One host machine among the multiple host machines receives a packet having an overlay multicast group identifier. The host machine maps the overlay multicast group identifier to an underlay multicast group identifier. The host machine encapsulates the packet with an encapsulation header that includes the underlay multicast group identifier to create an encapsulated packet. The host machine forwards the encapsulated packet to a physical switch of the network segment. The physical switch forwards copies of the encapsulated packet to tenant edge nodes at one or more ports that are determined to be interested in the underlay multicast group identifier.

Scalable overlay multicast routing in multi-tier edge gateways

A method of transmitting multicast traffic to workloads of tenants communicating over overlay networks provisioned on top of a physical network includes the steps of: detecting the multicast traffic; determining that the multicast traffic is bound for workloads of a first tenant and workloads of a second tenant; encapsulating one instance of the multicast traffic using a Layer 2 (L2) over Layer 3 (L3) encapsulation protocol to generate encapsulated traffic, wherein the encapsulated traffic includes an identifier of a first backplane network corresponding to the first tenant and an identifier of a second backplane network corresponding to the second tenant in a header portion of each packet of the encapsulated traffic; and transmitting, to a first host computing device, the encapsulated traffic with the identifiers of the first and second overlay networks.

Multicast routing through multi-tier edge gateways

Some embodiments provide a method for clustering a set of data compute nodes (DCNs), which communicate with each other more frequently, on one or more host machines. The method groups together guest DCNs (GDCNs) that (1) execute on different host machines and (2) exchange network data among themselves more frequently, in order to reduce interhost network traffic. The more frequently-communicating GDCNs can be a set of GDCNs that implement a distributed application, GDCNs of a particular tier in a multi-tier network architecture (e.g., a web tier in a three-tier architecture), GDCNs that are dedicated to a particular tenant in a hosting system, or any other set of GDCNs that exchange data among each other regularly for a particular purpose.

Monitoring and optimizing interhost network traffic

In some embodiments, a method adds a specific route for an IP address that is associated with a first workload into a routing table for a first network device in a first site in response to the first workload being migrated from a second site to the first site. The first network device receives a packet from a second workload for the first workload and determines that a destination of the packet matches the specific route in the routing table. The method routes the packet from the second workload to the first workload using the specific route in the routing table without sending the packet to the second site.

Routing using specific routes

In some embodiments, a method determines when a packet is fragmented into multiple fragmented packets in a flow between a first workload and a second workload. The method switches from generating an outer source port in the outer header using layer 4 information from the inner header to using layer 3 information from the inner header. A fragmented packet is encapsulated with the outer header that includes an outer source port value that is generated using the layer 3 information. The method initiates a process to determine when to switch back to using layer 4 information from the inner header to generate the outer source port. When it is determined to switch back to using layer 4 information, the method switches back to using layer 4 information from the inner header to generate the source port in the outer header of a packet from the first workload.

Adaptive source port generation in overlay networks

The disclosure provides an approach for reducing congestion within a network, the network comprising a plurality of subnets, the plurality of subnets comprising a plurality of host machines and a plurality of virtual computing instances (VCIs) running on the plurality of host machines. Embodiments include receiving, by an edge services gateway (ESG) of a first subnet of the plurality of subnets, membership information for a group identifying a subset of the plurality of host machines. Embodiments include receiving a multicast packet directed to the group and selecting from the plurality of host machines, a replicator host machine for the multicast packet. Embodiments include sending, to the replicator host machine, the multicast packet along with metadata indicating that the replicator host machine is to replicate the multicast packet to remaining host machines of the subset of the plurality of host machines identified in the membership information for the group.

Scalable overlay multicast routing

Described herein are systems and methods for allocating to tunnel endpoints to virtual machines on a host. In one example, a host identifies when a new virtual machine is coupled to a virtual switch and determines a tenant associated with the new virtual machine. The host further determines whether a tunnel endpoint is allocated to the tenant and, when a tunnel endpoint has not been allocated, allocates a new tunnel endpoint to the tenant and maps the new virtual machine to the new tunnel endpoint.

Allocation and management of tunnel endpoints in software defined networks

Some embodiments provide a method for deploying edge forwarding elements in a public or private software defined datacenter (SDDC). For an entity, the method deploys a default first edge forwarding element to process data message flows between machines of the entity in a first network of the SDDC and machines external to the first network of the SDDC. The method subsequently receives a request to allocate more bandwidth to a first set of the data message flows entering or exiting the first network of the SDDC. In response, the method deploys a second edge forwarding element to process the first set of data message flows of the entity in order to allocate more bandwidth to the first set of the data message flows, while continuing to process a second set of data message flows of the entity through the default first edge node. The method in some embodiments receives the request for more bandwidth by first receiving a request to create a traffic group and then receiving a list of network addresses that are associated with the traffic group. In some embodiments, the method receives the list of network addresses associated with the traffic group by receiving a prefix of network addresses and receiving a request to associate the prefix of network addresses with the traffic group. Based on this request, the method then creates an association between the traffic group and the received prefix of network addresses.

Allocating additional bandwidth to resources in a datacenter through deployment of dedicated gateways

In an embodiment, a computer-implemented method provides mechanisms for identifying a source location in a service chaining topology. In an embodiment, a method comprises: determining, at an egress interface of a host that hosts a virtual machine ('VM'), whether a service plane MAC address ('spmac') in a packet header of a packet, provided to the egress interface, is the same as an inner destination MAC address in the packet; in response to determining that the spmac in the packet header of the packet, provided to the egress interface, is the same as the inner destination MAC address in the packet: encapsulating the packet with a destination virtual tunnel endpoint ('VTEP') address retrieved from a mapping of VTEP-labels onto VTEP addresses; and causing providing the packet from the egress interface of the host that hosts the VM to a source host that hosts a source guest virtual machine (&#x201c;GVM&#x201d;).

Label-based methods for identifying a source location in service chaining topology

The technology disclosed herein enables a dynamic chain of service functions for processing network traffic. In a particular embodiment, a method includes, in a logical router for a logical network connecting service functions, receiving a network packet from a service function over the logical network after the network packet has been processed by the service function. The method further includes determining a new classification of the network packet and determining a next service function based on application of a service chain policy to the new classification. The method also includes directing the network packet to the next service function over the logical network.

Dynamic chain of service functions for processing network traffic

Growing community of inventors

Sunnyvale, CA, United States of America

Chidambareswaran Raman