CCNP Data Center key terms

An overlay encapsulation that carries Layer 2 frames inside UDP/IP packets across a Layer 3 network, using a 24-bit VXLAN Network Identifier to separate up to roughly 16 million logical segments.

A BGP address family used as the control plane for VXLAN and other overlays; it distributes MAC and IP reachability so endpoints are learned from the protocol instead of data-plane flooding.

A Cisco Nexus feature that lets two switches form one logical PortChannel toward servers or downstream switches, giving active-active forwarding and link redundancy without Spanning Tree blocking.

A heartbeat path, typically a routed out-of-band link, that the two vPC peers use to detect whether their partner is alive before deciding to keep or suspend vPC member ports.

A remote line-card device connected to a parent Nexus switch over fiber, extending ports so many access racks appear as one managed switch with a single control plane.

A Layer 2 overlay for connecting data centers across a Layer 3 transport; it encapsulates frames in IP and uses a control plane to advertise MAC reachability while suppressing unknown-unicast flooding.

Cisco's policy-driven data center fabric in which spine and leaf switches are programmed by a centralized controller, and connectivity is defined by endpoint groups and contracts rather than per-switch CLI.

An ACI object that groups endpoints sharing the same policy requirements, such as a VLAN or a set of virtual machines; communication between EPGs is permitted or denied by contracts.

Firmware embedded in the UCS Fabric Interconnects that provides the GUI, CLI, and API used to discover, configure, and monitor all attached chassis, blades, and rack servers.

A UCS object holding a complete server configuration - identity, firmware, network, and storage settings - that is associated with a blade or rack server to make it operational.

Top-of-fabric UCS switches that run UCS Manager, carry LAN and SAN traffic from servers, and act as the single management point for a UCS domain.

The I/O module inside a UCS chassis that multiplexes blade adapter traffic onto the fabric interconnects and is discovered and managed by UCS Manager as part of the chassis.

Cisco's converged adapter for UCS servers, presenting many virtual NICs and HBAs to the operating system and offloading features such as VM-FEX and traffic monitoring.

A design in which server identity (MAC addresses, WWPNs, UUID, firmware, and connectivity) lives in a service profile rather than on hardware, so any server can take over that profile.

Cisco's cloud-based management platform that monitors and configures UCS, HyperFlex, and other infrastructure through a SaaS portal or an on-premises virtual appliance.

Cisco's hyperconverged infrastructure, combining UCS compute with a distributed storage controller so compute and storage scale together inside the same virtualization cluster.

A logically isolated Fibre Channel fabric inside a physical SAN; each VSAN keeps its own domain IDs, zoning database, and fabric services, letting one switch carry multiple independent fabrics.

A 64-bit identifier assigned to a specific port on a Fibre Channel device; switch and array configurations use it in zoning and LUN mapping so each port is uniquely identified on the fabric.

A Fibre Channel fabric service that restricts which ports may communicate by grouping them into zones; a device can log in and exchange traffic only with ports in the same zone.

The fabric login process in which a node port obtains a fabric-assigned FC ID and registers its parameters with the switch, making it a valid member of the fabric.

N_Port ID Virtualization; a standard that lets one physical HBA port log in to the fabric multiple times, so each virtual machine gets its own FC ID and worldwide name.

Fibre Channel frames encapsulated in Ethernet so a converged adapter can send storage and LAN traffic over the same lossless 10/25 Gb link.

An IP tunneling protocol that connects Fibre Channel SAN islands over IP WAN links by wrapping FC frames in TCP, typically used for remote replication between sites.

A protocol that carries SCSI commands over TCP/IP, allowing hosts to access block storage across a standard Ethernet network instead of a Fibre Channel fabric.

An interswitch link: the connection between two Fibre Channel switches that carries fabric traffic and is normally a member of a SAN port channel for redundancy.

A unique number assigned to each switch in a Fibre Channel fabric; it forms the upper byte of every FC ID and must be unique for logins and fabric merges to succeed.

A REST and RESTCONF interface on NX-OS switches that lets scripts read and change configuration and state over HTTPS instead of running SSH CLI sessions.

The Data Management Engine is the model-based data store behind the NX-OS API; objects, properties, and methods are addressed by distinguished names and exposed through NX-API and the CLI.

A modeling language that defines the structure and data types of a device's configuration and state, used by NETCONF, RESTCONF, and model-driven telemetry.

An agentless automation tool that runs playbooks over SSH or API to push configuration to devices; modules such as cisco.nxos make tasks repeatable and idempotent.

A Linux container that runs on NX-OS devices, giving automation scripts, Python libraries, and tools direct access to the device CLI and file system.

A push model in which a switch periodically sends structured performance and state data to a collector, replacing polling with near-real-time, model-driven updates.

Role-Based Access Control; users are mapped to roles that grant permission for specific commands or objects, so privileges follow the assigned role rather than shared credentials.

A AAA protocol that separates authentication, authorization, and accounting, letting a central server decide which commands each administrator is permitted to run on a device.

A port-based network access control standard in which a supplicant authenticates to an authenticator switch before the port is authorized to pass normal traffic.

IEEE 802.1AE link-layer encryption and integrity checking between directly connected devices, protecting a point-to-point Ethernet segment from eavesdropping and tampering.

Control Plane Policing; a policy applied to traffic destined for the switch CPU that rate-limits and filters it so floods or attacks cannot overwhelm routing and management protocols.

A switch feature that inspects DHCP messages, builds a trusted binding table of client MAC, IP, port, and VLAN, and drops server replies that arrive on untrusted ports.