Skip to main content
Migration state: source-mapped. This page was imported from the ZEDEDA Help Center and needs product, support, and editorial review before production launch.
Article Published Date: 6/9/2026

​
Edge Infrastructure Services Versions

  • ZEDEDA GUI: 19.2.9
  • ZEDEDA API: 19.2.9

​
New GA Features

​
802.1x Port-based Network Access Control with Simple Certificate Enrollment Protocol

802.1x Port-based Network Access Control (PNAC) with Simple Certificate Enrollment Protocol (SCEP) support enables nodes to authenticate with enterprise networks using certificate-based identity before being granted full network access. This feature requires EVE-OS version 17.x-LTS or greater. See the following:
  1. Network Overview
  2. 802.1x Port-based Network Access Control Overview
  3. Configure 802.1x Port-based Network Access Control with Simple Certificate Enrollment Protocol

​
Flexible PCR Policy

The Flexible PCR Policy allows you to customize which Platform Configuration Registers (PCRs) EVE-OS (Edge Virtualization Engine) uses when sealing the vault encryption key into the Trusted Platform Module (TPM).

​
OpenTelemetry Metrics Export

The OpenTelemetry Metrics Export provides a standardized, streaming mechanism to send ZEDEDA platform observability data directly to your preferred monitoring and alerting tools.

​
AI Assistant

The AI Assistant that was introduced as an early access feature in Edge Infrastructure Services Release 18.x is now generally available. LAG/Bond interface support enables you to create and manage link aggregation groups (bonds) on nodes, combining multiple physical network interfaces into a single logical interface for increased bandwidth and redundancy. Bond interfaces can be configured through the ZEDEDA API or Terraform provider. The ZEDEDA Console displays bond interface status, member ports, and interface usage types for monitoring and troubleshooting.

​
Enhanced GA Features

​
ZEDEDA Edge Kubernetes Load Balancer Service

The Load Balancer Service allows ZEDEDA Edge Kubernetes Service clusters to assign external IP addresses to Kubernetes services. When enabled, you specify the network interface and an IP address range that the cluster uses to allocate IPs for load-balanced services.

​
Edge View GUI Improvements

Improved the following GUI elements in Edge View:
  • Revised the default size of the SSH and VNC consoles, optimizing the usability of the control characters (tab key, escape key, arrow keys, etc.)
  • Revised the behavior of the minimize and maximize icon
  • Ability to customize the size of the terminal
  • Support for EVE-k
  • Better session control

​
EVE-k Remote Access VNC

EVE-k now supports Remote Console VNC. Previously Remote Console access to Virtual Machines (VMs) was not supported in EVE-K 16.0-LTS. You can still use the Virtual Network Computing (VNC) alternative from your laptop.

​
New for POC & Demo Only

​
Onboard an Edge Device

The new onboarding process is a streamlined method for registering an edge device with Edge Infrastructure Services without requiring hardware information upfront. Instead of pre-configuring hardware identity details before connecting the device, you provide only an onboarding key and a serial number. Modelless onboarding via single EVE OS Installer requires EVE-OS version 17.x-LTS or greater.

​
Known Issues

Known behavior: grub.cfg overrides must be mounted when building an installer ISO for EVE-k Install-time storage parameters configured in grub.cfg (for example, ZFS RAID level, install disk, and persist disk settings) are only applied during installation if the configuration file is passed to the build container using a volume mount. If you build the installer ISO without mounting an overrides directory, EVE-k uses default values, which may result in unexpected filesystem behavior such as ext4 being used instead of ZFS on the install disk. Workaround: Place your grub.cfg in a local overrides directory and mount it when running the installer ISO build command:
Known behavior: Node app instances stuck in init state after off-boarding and re-onboarding with a different device name (EVE-k) On EVE-k deployments, if a node is off-boarded and re-onboarded using a different node name, app instances might become stuck in the init state indefinitely. This occurs because Kubernetes does not support node name changes: renaming the node causes Longhorn storage to enter an invalid state, and app instances fail to start with the error “failed to build kube config: Failed to register with Kubevirt.” When off-boarding and re-onboarding an EVE-k node, use the same device name for both operations. Known limitation: Internet access must be available on the lowest-cost (highest-priority) management interface for EVE-k deployment and upgrade. EVE-k requires online download of its artifacts during deployment or upgrade, and only the lowest-cost management interface is used for this process. The download does not probe all available management interfaces.
Source article: Release 19.2