Inference
Marketplace
Introduction
This page defines key terms used across ZEDEDA Edge Intelligence Platform documentation. The first section covers ZEDEDA-specific terminology, including items renamed in the current console. The second section covers general industry terms that appear in ZEDEDA documentation.ZEDEDA Terminology
| Term | Description | Previous name (if renamed) |
|---|---|---|
| Edge Intelligence Platform | The unified ZEDEDA platform combining infrastructure orchestration, AI model inference, and a cross-service marketplace. | ZEDEDA Edge Intelligence Platform (unchanged) |
| Edge Inference Services | The service layer for AI model management, evaluation, validation, and inference operations at the edge. Provides tools for deploying and monitoring AI models across edge infrastructure. | — |
| Edge Infrastructure Services | The service layer that provides EVE-OS-based node management, workload orchestration, and Kubernetes operations at the edge. Previously referred to as ZEDEDA Cloud or ZedCloud. | ZEDEDA Cloud / ZedControl / ZedCloud |
| EVE-OS (Edge Virtualization Engine) | The open-source operating system that runs on nodes. EVE-OS abstracts hardware resources and enables remote orchestration of containerized and virtual machine workloads. | EVE-OS (unchanged) |
| EVE-k | A k3s-based EVE-OS variant that enables node clustering and ZEDEDA Edge Kubernetes Service. The standard EVE-OS variant remains the recommended option for deployments that do not use clustering or Kubernetes. | EVE-Kubevirt |
| Marketplace | The cross-service catalog for browsing and deploying edge applications, Kubernetes applications, AI applications, models, and hardware configurations. | Marketplace (unchanged) |
| API-first | A design approach where all functionality available through the console GUI is also accessible programmatically through the ZEDEDA API, enabling integration with existing IT and OT systems and automation of onboarding, deployment, and policy management. | — |
| App Images | Container or virtual machine image artifacts used as the source for deploying applications. Images can be stored in ZEDEDA-managed storage, customer-owned data stores, or public registries such as Docker Hub. | Edge App Images |
| Applications | Deployed workload instances running on nodes, including containers and virtual machines. An application is a running deployment of an edge application image on a specific node. | Edge App Instances |
| configuration drift | The condition where a node’s actual configuration deviates from its intended state, often due to manual changes or unmanaged updates. Edge Infrastructure Services combats drift through centralized policies and the Configuration Lock feature. | — |
| edge container | A lightweight containerized workload running on a node. Used as the generic term for workloads, which may be containers, virtual machines, or unikernels. | VM, container, unikernel |
| edge gap | The disparity between the perceived potential of edge computing and the practical complexities that hinder its adoption, including limited connectivity, diverse hardware, and security requirements. | — |
| node | A physical or virtual compute device running EVE-OS and managed by Edge Infrastructure Services. Before onboarding, the device is referred to as an edge device. After onboarding, it is referred to as a node. | edge device (used before onboarding) |
| Edge Sync | A tool for managing nodes that do not have internet access. Edge Sync acts as a local controller, managing configurations and collecting metrics from nodes in air-gapped or intermittently connected environments. Previously known as local operator console (LOC). | local operator console (LOC) |
| Edge View | A remote access tool that provides secure terminal access to nodes and their applications for troubleshooting and maintenance. Implemented as a Docker container running on both the remote device and the user’s machine. | — |
| enterprise | A complete, discrete set of nodes, applications, users, and an associated Edge Infrastructure Services configuration. Each ZEDEDA customer has one enterprise, which can be subdivided into projects. | — |
| Full Disk Encryption (FDE) | A security feature that encrypts all data on an edge device’s storage, including the operating system, making it unreadable to unauthorized parties if the device is lost or stolen. | — |
| jobs | A group of tasks triggered by a bulk operation in Edge Infrastructure Services, such as deploying multiple app instances or updating EVE-OS across a fleet. Jobs are tracked so you can monitor progress across all tasks. | — |
| Local UI | An on-device interface for monitoring and basic management of nodes, accessed by physically connecting a keyboard and monitor to the device. Useful for diagnosing issues when a node is disconnected from Edge Infrastructure Services. | — |
| measured boot | A security process that cryptographically measures each component of the boot process before it executes, from firmware to operating system. Measurements are stored in the TPM, creating an unchangeable record used in remote attestation. | — |
| Clusters | Groups of nodes managed together as a unit to provide high availability and data replication. If one node fails, another in the cluster automatically takes over. | Clusters |
| Nodes | The managed inventory of nodes in a project or enterprise. | Nodes |
| Offline Profile Server (OPS) | A feature that allows you to manage application instances on a node even when the node is offline and cannot reach Edge Infrastructure Services. Only application instances assigned to a profile can be managed through OPS. | — |
| onboarding | The process of integrating a new edge device into a ZEDEDA enterprise. During onboarding, the node receives a project association, security credentials, management policies, and network settings. EVE-OS must be installed on the device before onboarding. | — |
| project | A logical grouping of nodes, applications, and deployments within an enterprise. Projects help organize assets by geography, department, or business function. Each node belongs to exactly one project. | — |
| remote attestation | A security mechanism that allows a trusted authority to verify that a remote node is in an expected, untampered state. The process uses cryptographically signed evidence of the node’s hardware configuration, firmware, operating system, and running applications, validated using the Trusted Platform Module (TPM). | attestation |
| Runtime Bundles | Packaged runtime resources, configurations, or binary artifacts that support application execution at the edge. | Patch Envelopes / Binary Artifacts |
| truck roll | The act of dispatching a technician to a physical edge site for maintenance, troubleshooting, or upgrades. ZEDEDA’s zero-touch remote management reduces the need for truck rolls by enabling centralized control and updates without on-site presence. | — |
| Trusted Platform Module (TPM) | A hardware security chip on an edge device that stores cryptographic keys and performs security tasks such as generating keys, encrypting data, authenticating devices, and supporting remote attestation. | — |
| user agent | An interface used to access and control resources in a ZEDEDA deployment. The three supported user agents are the console GUI, ZCLI, and the Terraform provider. All platform functionality is also available through direct API access. | — |
| Virtual Networks | Network configurations that control how applications attach to networks on a node, including routing, DHCP, NAT, and segmentation. Virtual networks are virtual subnets maintained by EVE-OS and can be internal to the node or attached to external interfaces. | Network Instances |
| ZCLI | The ZEDEDA command-line interface for interacting with Edge Infrastructure Services. Requires Docker Desktop to run. Provides the same management capabilities as the console GUI. | — |
| ZedCloud | Deprecated. The previous name for Edge Infrastructure Services. | — |
| ZedControl | Deprecated. The previous name for the ZEDEDA console GUI. | — |
| ZEDEDA Cloud Terraform Provider | A Terraform provider that allows you to automate and manage nodes and resources in Edge Infrastructure Services using infrastructure-as-code. Used for configuration management only, not for monitoring. | — |
| ZEDEDA Edge Kubernetes App Flows | The GitOps and fleet management layer of the ZEDEDA Kubernetes solution, used for managing application deployments across Kubernetes clusters at the edge using GitOps repositories. | — |
| ZEDEDA Edge Kubernetes Service | The Kubernetes features that Edge Infrastructure Services exposes on a cluster when Kubernetes access is enabled: GitOps repositories, Helm charts, the Kubernetes dashboard, and the cluster shell. Every cluster runs Kubernetes on its nodes; enabling Kubernetes access makes that Kubernetes available to you in addition to application instances. | Edge Kubernetes |
| Zero-Touch Deployments | An automated deployment approach that uses tagging to match nodes with their corresponding applications and policies, enabling deployments with minimal administrative overhead. | — |
| zero trust edge security policy | A security model that treats every access request as untrusted regardless of origin. Zero trust continuously verifies user and device identities, validates device health, restricts access to the minimum required, segments networks, and encrypts all data end-to-end. | — |
General Industry Terms
| Term | Definition |
|---|---|
| air gapped | A network that is physically isolated from all other networks, including the internet. Nodes can operate in air-gapped environments using Edge Sync to receive configurations and policies without a live internet connection. |
| AMD64 | A 64-bit processor architecture (also called x86-64) developed by AMD. Widely used in edge servers, industrial automation, and high-performance edge computing. Supports applications requiring higher computational power such as data processing, analytics, and AI inference. See also: x86. |
| ARM64 | A 64-bit processor architecture (also called AArch64) developed by ARM Holdings. Designed for high performance and energy efficiency, making it well-suited for IoT devices, edge gateways, and mobile devices operating under space, power, or heat constraints. |
| Bash | A popular command-line shell used on Linux and Unix-like operating systems. Used on nodes for low-level operations, troubleshooting, and custom scripting. |
| cluster | A group of interconnected computers that work together as a single system, distributing workloads and providing redundancy. In ZEDEDA, a cluster is a group of EVE-k nodes that runs application instances with high availability and storage replication. A cluster can also have Kubernetes access enabled, which exposes its Kubernetes to you for GitOps, Helm, and kubectl. An imported cluster is an external Kubernetes cluster adopted for application delivery only. |
| container | A lightweight, standalone software package that bundles an application with all its dependencies, enabling it to run consistently across different environments. Containers share the host OS kernel and are isolated from other applications. |
| CRUD | Create, Read, Update, Delete. The four fundamental operations for managing data and resources. In the Edge Infrastructure Services console, these operations are labeled Add, Monitor, Operate, and Remove. |
| Docker | A platform for building, packaging, and running applications in containers. Docker is the de facto standard for containerization and provides consistent deployment from cloud to edge environments. |
| Eastbound / Westbound | Terms describing horizontal, peer-to-peer communication between components at the same hierarchical level, such as service-to-service or server-to-server communication within the same tier. Often used interchangeably. |
| edge computing | The practice of processing data closer to where it is generated, rather than sending it to a centralized data center. Edge computing reduces latency, improves response times, and supports real-time processing for applications such as industrial IoT, manufacturing, and AI inference. |
| GitOps | An operational framework that uses a Git repository as the single source of truth for infrastructure and application state. An automated process continuously reconciles the live system state with the desired state defined in the repository, providing a complete audit trail and enabling consistent rollbacks. |
| Helm Chart | The packaging format used by Helm, the Kubernetes package manager. A Helm chart bundles all Kubernetes resource definitions, configurations, and dependencies needed to deploy an application into a single manageable package. |
| immutable | An image or volume that cannot be changed after creation. To apply changes, a new version is deployed. Immutability improves security and reliability by preventing tampering and ensuring consistent behavior across deployments. |
| instance | A single running copy of an application or operating system. In edge computing, an application instance is a deployed, running application on a specific node. |
| k3s | A lightweight, production-ready Kubernetes distribution designed for resource-constrained and remote environments. Used as the foundation for ZEDEDA Edge Kubernetes Service clusters through the EVE-k OS variant. |
| k8s | A common abbreviation for Kubernetes, derived from the eight letters between the K and the S. |
| kubectl | The primary command-line tool for interacting with Kubernetes clusters. Used to deploy applications, inspect nodes, and manage the overall state of a cluster. |
| Kubernetes | An open-source container orchestration system that automates the deployment, scaling, and management of containerized applications across a fleet of nodes. |
| LF Edge | An open-source framework governed by the Linux Foundation that provides edge computing infrastructure and tools. EVE-OS is an LF Edge project. |
| network | In Edge Infrastructure Services, a network handles traffic routing for node hardware, enabling the node to communicate with Edge Infrastructure Services for management. Distinct from a Virtual Network, which provides the virtual network environment for edge applications. |
| Northbound / Southbound | Terms describing vertical communication in a network hierarchy. Northbound refers to data or commands flowing up from edge devices to a management system or cloud platform. Southbound refers to commands or configurations flowing down from the management system to edge devices. |
| OCI | The Open Container Initiative (OCI) maintains a set of Docker standards for image spec, runtime spec, and distribution spec. |
| SBC | Single Board Computer. A fully functional computer built onto a single circuit board, such as a Raspberry Pi. Commonly used in IoT, education, and edge computing for low-cost, compact deployments. |
| shell | A command-line interface program that serves as an intermediary between the user and the OS kernel. Provides a text-based environment for running commands, managing processes, and automating tasks. Examples include Bash, Zsh, and PowerShell. |
| SSH | Secure Shell. A cryptographic network protocol for secure remote command-line access and file transfers over an unsecured network. Used to access nodes for troubleshooting and maintenance. |
| storage | In the context of nodes, storage refers to how data is saved and managed on the device. This includes persistent storage, which retains data when applications are not running, and transient storage, which is tied to the lifecycle of a running application. |
| systemd | A system and service manager for Linux that initializes and controls processes after the kernel boots. Ensures that EVE-OS services and edge applications start reliably, manage their lifecycles, and recover from failures. |
| TKG | Tanzu Kubernetes Grid. A VMware Kubernetes solution designed for hybrid and multi-cloud environments. Can be managed through ZEDEDA Edge Kubernetes Service alongside k3s-based clusters. |
| virtualization | The process of creating a software-based representation of a computing resource such as a server, storage device, or operating system, enabling multiple virtual instances to run on a single physical hardware platform. |
| volume instance | A data storage resource attached to a node. Volume instances can be persistent, retaining data after the associated application stops, or transient, storing data only while the application is running. |
| x86 | See AMD64. |
Next Steps
Navigate the ZEDEDA Console
Orient yourself in the sidebar, service areas, and key navigation patterns.
ZEDEDA Edge Intelligence Platform Overview
Introduction to the full ZEDEDA Edge Intelligence Platform.
Permissions and Visibility
Understand how roles control access and what users can see.