Aruba Data Center Professional key terms

A two-tier data center fabric in which every leaf switch connects to every spine switch, giving predictable latency, high bandwidth, and many equal-cost paths between servers.

The fabric tier that connects servers, storage, and firewalls to the network. Each leaf uplinks to every spine and often serves as the tunnel endpoint for its attached hosts.

The backbone tier of a spine-leaf fabric. Spines interconnect leaves and provide high-speed, low-latency transit; they do not normally connect to servers or endpoint devices directly.

The relationship between the total bandwidth offered to endpoints and the bandwidth of the uplinks leaving a switch or tier. Lower ratios mean more capacity and higher cost.

Traffic that moves between servers inside the data center, such as application-to-database flows. It dominates modern designs and drives the need for flat, high-bandwidth fabrics.

A switch mounted at the top of a server rack that aggregates the servers in that rack and uplinks into the fabric. It is often the same device as a leaf switch.

Aruba's modern network operating system for CX switches. It provides a REST API, a built-in time-series database, and a modular architecture across the data center switching portfolio.

An Aruba technology that pairs two CX switches so they act as one logical redundancy unit. Both peers stay independently operational and synchronize state over an inter-switch link for fast failover.

An Aruba feature that stacks multiple physical switches into one logical device with a single management address, typically used at access and small aggregation layers.

A data center switch family that embeds a stateful distributed firewall and telemetry engine directly in the switch, enabling line-rate policy enforcement at the rack edge.

The physical, IP-routed fabric that carries traffic between switches and provides loop-free, equal-cost paths. It transports the encapsulated packets used by overlay networks.

A logical network built on top of the physical underlay using tunnels. It carries tenant and segment traffic and is largely independent of the underlying routing topology.

A link-state interior gateway routing protocol that floods topology information within an area and computes shortest paths. It can serve as a data center underlay, though large fabrics often favor BGP.

A path-vector routing protocol used to exchange reachability between autonomous systems, and widely used inside data centers as the underlay because it scales well and supports multipath.

Forwarding traffic across several equally good next hops at once, spreading load and improving resilience. In spine-leaf fabrics it lets a leaf use all of its spine uplinks simultaneously.

A lightweight protocol that sends frequent keepalives between two devices to detect link or path failure in sub-second time, allowing routing protocols to converge faster.

A configuration style where BGP peers on point-to-point interfaces discover each other's link-local IPv6 addresses, avoiding per-link IP addressing and simplifying fabric provisioning.

A tunneling technology that encapsulates Layer 2 frames inside IP and UDP packets, letting a Layer 2 segment stretch across a Layer 3 network by adding a 24-bit segment identifier.

The device that encapsulates and decapsulates VXLAN traffic, mapping local VLANs to VNIs and sending frames through the underlay to remote tunnel endpoints.

A BGP control plane for VXLAN and other data planes that distributes MAC and IP reachability, enabling efficient flooding, multihoming, and host mobility in the overlay.

A 24-bit value in the VXLAN header that identifies a virtual network segment. It supports roughly 16 million segments, far beyond the VLAN limit.

A design where every leaf uses the same IP and MAC address as the default gateway for a segment, so hosts always find a local gateway and traffic never detours.

An EVPN feature that lets a server or device connect to several leaf switches using the same address, providing active-active redundancy without relying on spanning tree.

A way to run multiple independent routing tables on one device, keeping address spaces separate. It is commonly used in data centers to isolate tenants or functions.

A logical customer or workload group that receives its own isolated network segments, routing table, and policy while sharing the same physical fabric.

Running several isolated logical networks for different customers or applications over one shared physical infrastructure, using overlays and separate routing instances to prevent address overlap.

Applying security policy to individual workloads or workload groups rather than to subnets, so east-west traffic between servers can be filtered and inspected at fine granularity.

Granting network access according to assigned roles rather than static addresses, so policy follows a user or device. Roles are often derived from authentication or group membership.

An IEEE standard for port-based network access control. A supplicant authenticates to an authenticator, typically a switch, which then applies a role or VLAN to the port.

An Aruba feature that applies stateful, role-based rules to traffic, controlling which users and devices may reach which services regardless of their IP address.

An Aruba approach that assigns users and devices to roles and maps their traffic to the correct segment automatically, instead of configuring VLANs one port at a time.

A switch boot process in which the device contacts a server, downloads its configuration and software image, and configures itself without manual setup, using DHCP options to locate the server.

Managing network configuration through version-controlled, machine-readable files rather than manual command entry, so changes are repeatable, reviewable, and testable.

An agentless automation tool that pushes configuration to devices over SSH or APIs using declarative playbooks and reusable modules, including collections written for AOS-CX.

An HTTPS interface on CX switches that lets programs read and change configuration and state using standard HTTP methods, enabling automation and integration with external tools.

Aruba's GUI-based configuration and compliance tool for CX switches, letting teams build plans, validate them before deployment, and push changes to many devices at once.

A built-in AOS-CX capability that runs scripts against streaming telemetry and time-series data, raising alerts or taking automated action when conditions match a rule.

Continuously pushing structured operational data from network devices to collectors on a schedule or when values change, rather than polling with SNMP, giving near real-time visibility.

A packet-sampling standard in which switches send periodic samples of traffic to a collector, giving a view of top talkers, protocol mix, and interface utilization.