Cisco Technology, Inc.

Mouli Vytla

Samar Sharma

Rajendra Kumar Thirumurthi

Nikhil Dhar

Ashish Purushottam Attarde

Avni Baveja

Mitali Parthasarathy

Sachin Vijay Vishwarupe

Puneet Kumar

Giovanni Meo

Chickayya Naik

Kushal A Patel

Rahul BrajeshKumar Jain

Praveen Parthasarathy Iyengar

Mitesh Mukesh Kanjariya

Aditi Srivastava

Senthilkumar Krishnamurthy

Karthikeyan Gurusamy

Hemanth Basappa Mallikarjunaswamy

Souvik Ghosh

Tejas Nagarmat

In an example, a network switch is configured to natively act as a high-speed load balancer. Numerous load-balancing techniques may be used, including one that bases the traffic 'bucket' on a source IP address of an incoming packet. This particular technique provides a network administrator a powerful tool for shaping network traffic. For example, by assigning certain classes of computers on the network particular IP addresses, the network administrator can ensure that the traffic is load balanced in a desirable fashion. To further increase flexibility, the network administrator may apply a bit mask to the IP address, and expose only a portion, selected from a desired octet of the address.

Adjustable bit mask for high-speed native load balancing on a switch

In one embodiment, load balancing criteria and an indication of a plurality of network nodes is received. A plurality of forwarding entries are created based on the load balancing criteria and the indication of the plurality of nodes. A content addressable memory of a network element is programmed with the plurality of forwarding entries. The network element selectively load balances network traffic by applying the plurality of forwarding entries to the network traffic, wherein network traffic meeting the load balancing criteria is load balanced among the plurality of network nodes.

Selective load balancing of network traffic

In an example, a network switch is configured to operate natively as a load balancer. The switch receives incoming traffic on a first interface communicatively coupled to a first network, and assigns the traffic to one of a plurality of traffic buckets. This may include looking up a destination IP of an incoming packet in a fast memory such as a ternary content-addressable memory (TCAM) to determine whether the packet is directed to a virtual IP (VIP) address that is to be load balanced. If so, part of the source destination IP address may be used as a search tag in the TCAM to assign the incoming packet to a traffic bucket or IP address of a service node.

TCAM-based load balancing on a switch

In one embodiment, a maximum transmission unit (MTU) mismatch assessment service receives a notification of a mismatch between a packet size of a packet sent by a source to a destination in a network and an MTU of an intermediate router between the source and destination in the network. The service determines, using a machine learning-based model, that the mismatch represents a persistent MTU mismatch condition at the intermediate router. The service identifies a target router in the network to receive a configuration adjustment instruction, based on the persistent MTU mismatch condition. The service sends the configuration adjustment instruction to the target router, to alleviate the persistent MTU mismatch condition at the intermediate router.

Discovering and mitigating MTU/fragmentation issues in a computer network

Techniques for monitoring packet telemetry are provided. A policy is received at a first node from a controller, where the policy includes an indication of a first flow. A first packet belonging to the first flow is received at the first node. A second node in a network path for the first packet is determined. A first header is added to the first packet based on the policy, wherein the first header includes an indication of the controller. The first packet is transmitted to the second node. Finally, telemetry data associated with the first node is transmitted to the controller based on the policy.

Packet telemetry data via first hop node configuration

In an example, there is disclosed a network apparatus for providing native load balancing within a switch or router, including a first network interface operable to communicatively couple to a first network; a plurality of second network interfaces operable to communicatively couple to a second network; one or more logic elements comprising a switching engine operable for providing network switching or routing; and one or more logic elements forming a load balancing engine operable for receiving incoming network traffic via the first network, the incoming traffic having a destination internet protocol address (IP) corresponding to a virtual IP (VIP) designated for load balancing; assigning the incoming traffic to a traffic bucket associated with the second network; and logging the assigning; wherein the switching engine and load balancing engine are configured to be provided on the same hardware as each other and as the first network interface and plurality of second network interface.

Statistical collection in a network switch natively configured as a load balancer

An approach enables networking functionality to be configured at network switches for containers hosted by computing devices. A container supervisor on the network switch is configured to obtain, via a container network manager, container attribute data for containers hosted by the computing devices. The container network manager is configured to interact with and manage container orchestration engines on behalf of the containers. This allows the network switch to be container orchestration agnostic. The network switch enables networking functionality for the containers by updating container configuration data on the networking switch based upon the container attribute data obtained from the container network manager. This may include, for example, updating hardware tables on the network switch with the container attribute data to allow the network switch, and applications executing on the network switch, to perform networking functionality with respect to the containers.

Configuring container attribute data on network switches to enable networking functionality

Embodiments include receiving configuration information including a match criterion for packets received at a network device in a network and a pool of layer 3 addresses associated with a set of servers in the network, resolving layer 2 destination addresses based on the layer 3 addresses of the servers, and programming a hardware layer of the network device based, at least in part, on the match criterion, the pool of layer 3 addresses, and the layer 2 destination addresses. Specific embodiments include configuring a policy to indicate that packets from an external source are to be forwarded to a server of the set of servers. Further embodiments include receiving a packet at the network device, and matching the packet to the pool of layer 3 addresses and the resolved layer 2 addresses based, at least in part, on the match criterion programmed in the hardware layer.

Scalable network address translation at high speed in a network environment

In an example, there is disclosed a network apparatus, comprising: one or more logic elements, including at least one hardware logic element, to provide a network manager engine to: provide a switched fabric management function; communicatively couple to at least one network switch, the network switch configured to provide optional native hardware-based load balancing; monitor one or more load balancing factors; and at least partly responsive to the one or more load balancing factors, configure native hardware-based load balancing on the at least one network switch.

Load balancing mobility with automated fabric architecture

A network apparatus for providing native load balancing, having: a first network interface to communicatively couple to a first network; a plurality of second network interfaces to communicatively couple to a second network; one or more logic elements providing a switching engine to provide network switching or routing; and one or more logic elements, including at least one hardware logic element, providing a load balancing engine to: load balance network traffic among a plurality of service nodes; probe a first service node with a first probe; and probe a second service node with a second probe, the second probe different in kind from the first probe.

Node-specific probes in a native load balancer

In an example, there is provided a network apparatus for providing native load balancing within a switch, including a first network interface operable to communicatively couple to a first network; a plurality of second network interfaces operable to communicatively couple to a second network, the second network comprising a service pool of service nodes; one or more logic elements providing a switching engine operable for providing network switching; and one or more logic elements comprising a load balancing engine operable for: load balancing incoming network traffic to the service pool via native hardware according to a load balancing configuration; detecting a new service node added to the service pool; and adjusting the load balancing configuration to account for the new service node; wherein the switching engine and load balancing engine are configured to be provided on the same hardware as each other and as the first network interface and plurality of second network interfaces.

Growing community of inventors

San Jose, CA, United States of America

Mouli Vytla