An ACL applies rules to switch ports or (Internet Protocol) IP addresses available to a host or switch. The rules determine whether or not packets can flow through those ports or IPs.

A communications protocol used by IPv4 to map an IP address to a physical machine address.

Free software that installs and removes software packages from some Linux operating systems, such as Debian.

A multicast where hosts can transmit to/from groups without any restrictions on the receivers.

An integrated circuit customized for a specific use.

A physical network switch without a proprietary operating system. For example, a switch purchased directly from Lenovo that you can install on NVIDIA® Cumulus Linux™ or another open operating system (OS).

Unnumbered interfaces don't have unique IP addresses. For Border Gateway Protocol (BGP), you configure unnumbered interfaces using extended next-hop encoding (ENHE).

An exterior gateway protocol that exchanges routing and reachability information among autonomous systems.

Three types of network traffic. Unicast involves communication from one single point to another single point. Broadcast is when information is sent from a single point to multiple points. Multicast is communication of information from multiple points to other points.

The funds that a company spends on physical assets, such as equipment.

The proprietary version of Link Layer Discovery Protocol (LLDP).

Another term for a leaf/spine network.

The interface through which one interacts with a computer program.

The process of making continual improvement rather than a configure-and-forget mentality to IT configuration management.

Enables a server or switch with a two-port bond (such as a link aggregation group [LAG], EtherChannel, port group, or trunk) to connect those ports to different switches and operate as if they’re connected to a single, logical switch.

NVIDIA Cumulus VX™ is a virtual appliance that runs on many popular hypervisors to help you preview NVIDIA technology.

A computer program that tells the OS how to communicate with a piece of hardware.

A protocol that translates domain names into IP addresses.

A way to automatically assign IP addresses and settings to a network device.

A load-sharing routing strategy where a packet being forwarded to the next hop on the way to a destination can travel over multiple best paths.

A standards-based control plane for virtual extensible local area network (VXLAN) that enables building VXLANs at scale. EVPN allows legacy layer-2 applications to operate over next-generation layer-3 networks.

An extension to the IP and the Transmission Control Protocol (TCP) that allows end-to-end notification of network congestion without dropping packets.

One of two branches of NVIDIA Cumulus® code. The ESR branch focuses on stability, so only critical-impact security fixes and urgent bug fixes are added to the code.

A modern open-source routing application that evolved from the original Quagga project.

A list of all hardware that’s compatible with the NVIDIA Cumulus Linux OS.

Used to generate broadcast, unknown-unicast, and multicast (BUM) traffic in hardware to multiple VXLAN destinations, also called virtual tunnel end points (VTEPs).

A form of rapid trading that’s based on highly sophisticated programs and algorithms. They’re usually short-term investments, characterized by high speed and sensitivity to latency.

An updated Linux interface manager.

A communications protocol that enables an internet computer to report its multicast group membership to adjacent routers.

The layer of the Open Systems Interconnection (OSI) stack concerned with IP addresses; the routing layer.

An open, vendor-neutral, industry-standard application protocol for accessing and maintaining distributed directory information services over an IP network.

A technique for deploying VXLANs without a central controller on bare-metal switches.

A method to control the bundling of several physical ports together to form a single logical channel.

Used by network devices to advertise their identity, capabilities, and neighbors on an Ethernet network.

A computer network that interconnects computers within a limited geographic area.

The largest size packet or frame that can be communicated in a network transaction.

Enables a server or switch with a two-port bond (such as a link aggregation group [LAG], EtherChannel, port group, or trunk) to connect those ports to different switches and operate as if they’re connected to a single, logical switch.

A set of protocols to define and implement a policy that describes how to secure access to network nodes by devices when they initially attempt to access the network.

A rich and simple Cumulus Linux CLI for easy network configuration and operation.

The operating system that runs on network switches. Cumulus Linux is an NOS.

Network operations are actions completed within an organization’s network that are comprised of tasks such as monitoring, troubleshooting, capturing packets, provisioning network resources, responding to critical issues that arise, and more.

A collaborative community seeking to improve hardware technology and infrastructure design.

A real-time, policy-driven tool for orchestration and automation of physical and virtual network functions.

An Open Compute Project open-source initiative contributed to by Cumulus Networks (now NVIDIA) that defines an open “install environment” for bare-metal network switches. ONIE is a bootloader for NOSes.

A link-state routing protocol for IPv4 (OSPFv2) and IPv6 (OSPFv3).

Funds that a company spends on operations.

Interfaces without unique IP addresses. OSPF unnumbered interfaces reduce the links between routers into pure topological elements, which dramatically simplifies network configuration and reconfiguration.

Refers to data that’s transferred outside of the data plane network and is used for management traffic.

A network protocol that enables Ethernet cables to carry electric power for devices.

A dynamic cabling verification tool developed by Cumulus Networks (now NVIDIA). With PTM, you can program your data center to verify network connections and resolve issues faster.

A link-level flow control mechanism that can be controlled independently for each class of service (CoS) with the intention to ensure no data frames are lost when congestion occurs in a bridged network.

A multicast control plane protocol that advertises multicast sources and receivers over a routed layer-3 network. Has sparse and dense modes of operation.

Original open-source routing application in NVIDIA Cumulus Linux, replaced by FRRouting.

An NVIDIA solution that provides an out-of-band management switch preinstalled with an NOS.

A protocol for writing to compute or storage elements using remote direct-memory access (RDMA) over an Ethernet network instead of using host CPUs.

A mechanism for IP subnets to span racks without forcing the end hosts to run a routing protocol.

A protocol that provides centralized authentication, authorization, and accounting management (AAA) for users connected to a network device or service.

The ability for an ASIC to do VXLAN routing.

A network routing component for BGP that acts as an alternative to the full-mesh of internal BGP (IBGP).

A troubleshooting tool for monitoring computer networks.

As opposed to traditional data centers, where infrastructure is defined by hardware, SDDCs are comprised of virtualized infrastructure (networking, storage, CPU, and security) that’s delivered as a service.

A cloud computing strategy that facilitates network management and enables programmatically efficient network configuration to improve network performance and monitoring.

A secure method of delivering multicast packets, where the receiver only receives packets from the specific source address they requested.

An open-source framework that enables ASICs to be represented in software, specifically in the user space; a Microsoft contribution to the OCP.

An authentication protocol used for remote authentication and related services for networked access control through a centralized server.

A network architecture where servers within a rack are connected to one or two network switches within the same rack.

The combined cost of CapEx and OpEx.

A standard overlay protocol that abstracts logical virtual networks from the physical network underneath. You can deploy simple and scalable layer-3 Clos architectures while extending layer-2 segments over that layer-3 network.

Separate virtual networks on the same physical network; network packets are tagged so traffic flows to the appropriate VLAN. This lets you keep network applications separate despite being connected to the same physical network and doesn’t require multiple sets of cabling and networking devices to be deployed.

An isolated or abstracted network within a larger network.

Enables hosts to communicate with any redundant router without reconfiguration, running dynamic router protocols or running router redundancy protocols. Thus, redundant routers will respond to ARP requests from hosts.

Lets you run multiple network paths without the need for multiple switches, giving you traffic isolation and network segmentation for multiple devices. It's achieved through the presence of multiple independent routing tables working simultaneously on the same router or switch.

Physical or virtual switch that’s the termination point for one member of a multi-tenant network. A VTEP connects to other VTEPs in other tenants via VXLAN tunnels.

The ID used by VXLAN for a VXLAN tunnel.

Sometimes referred to as inter-VXLAN routing, it provides IP routing between VXLAN VNIs in overlay networks. The routing of traffic is based on the inner header or the overlay tenant IP address.

A modern architectural approach that enables networks to scale to the level of cloud giants, such as AWS. Key constructs for this architecture include open and modular, intelligence in software, scalable, and efficient.

Same as a bare-metal switch: a physical network switch without an operating system. For example, a switch purchased directly from Edge Core without Cumulus Linux.

A switch feature that enables automation of provisioning and configuration.