NativeEdge

This document provides step-by-step instructions for installing Ubuntu 24.04 LTS Server from ISO media and configuring it as a Dell NativeEdge Witness Node. The installation uses minimal packages to optimize performance and security. The usual and recommended method for installing a Dell NativeEdge Witness Node is to use the free blueprint provided in the NativeEdge blueprint catalog. This manual installation guide is intended for users who, for specific reasons, must create a witness node manually instead of using the automated blueprint deployment.

The Dell Automation Platform orchestrator displays a critical alert for the modification to the security setup of Orchestrator Microsegmentation. Alert code 0x20010002 is shown.

When using the legacy Dell NativeEdge v1 Gateway API POST /rest/v1/jobs to add, update, or delete a host network (endpoint:network:hostNw:*), the request may fail withHTTP 400 and the message property "nwSeg" is missing. This happens when the hostNw payload objects do not include the endpoint id field. The v2 Gateway API is recommended; if v1 must be used, include the endpoint UUID as the id in each hostNw object.

Dell Distributed Private Cloud protection jobs for many assets can fail with a 3600-second timeout when Data Domain server connectivity is lost mid-backup, leaving stale backup processes on DDPC endpoints. Reboot the affected ECE nodes to clear the stale jobs.

Dell Distributed Private Cloud (DDPC) single-node cluster upgrades can force-kill running VMs when VM Shutdown procedure skips cluster VMs because maintenance mode is not enabled. Manually request maintenance mode before the upgrade to migrate VMs.

On Dell Distributed Private Cloud (DDPC) High-Availability clusters, Cluster Storage goes into Degraded State when self-encrypting drive (SED) does not auto rejoin after the disk is removed and reinserted. The LVM activation step is skipped because the mount step fails. Reboot the affected node to recover.

After multiple rapid create and destroy operations using the Gateway API /rest/v1/endpoints, an Edge Compute Engine (ECE) may lose all datastores. This occurs when the destroy request is sent before cluster creation completes, causing concurrent operations that leave drives in an unusable state.

After a VM failover during degraded shared-storage conditions, scheduled synthetic-full backups consistently fail with "Domain checkpoint not found" error. This occurs when a failed synthetic-full operation leaves a stale checkpoint/bitmap reference without the corresponding shared checkpoint XML file.

After completing an Instant Restore of a VM with Instant Access enabled, the restored VM cannot be added to a protection policy in PowerProtect Data Manager (PPDM). This occurs because the DDPC compute-svc retains stale metadata, setting deployment_type to "instant_access" instead of "restore" after successful Instant Restore completion. The PPDM UI filters out VMs with deployment_type "instant_access" as ineligible for backup.

In Dell Distributed Private Cloud, creating a High-Availability cluster through a sequence of API calls can leave one node unhealthy. The root cause is a stale cleanup request from a previously failed or interrupted cluster operation that destroys Ceph components on a node that is now part of an active cluster. The ECE agent does not serializecluster lifecycle operations, so a delayed cleanup can corrupt the live cluster.

In large-scale Dell Automation Platform (DAP) deployments, endpoint and virtual machine (VM) metrics can stop appearing in dashboards, reports, and the /api/v2/datapoints API while endpoint operations continue. This occurs when the VictoriaMetrics vmstorage database becomes full and switches to read-only mode. New metrics cannot be ingested and recent data cannot be queried until storage is expanded or data is removed. This article describes symptoms, root cause, workarounds, and sizing guidance.

Cluster upgrade fails with "Update denied by vm-handler" error when attempting to upgrade a Dell NativeEdge cluster from version 4.2.0.0 to 4.3.0.0. The failure occurs when the cluster contains imported VMs that have incorrect Raft runstate records, causing the system to attempt live migration on stopped VMs.

The Dashboard status count does not include clusters with Unknown, Error, or Warning states. This causes a discrepancy between the Dashboard count and the actual number of clusters shown in Infrastructure navigation. This behavior is by design; the Dashboard uses a platform-generic status enum for consistency across providers.

NativeEdge endpoints lose complete network connectivity after VM deployment on Broadcom bnxt_en NICs when VLAN filtering is enabled. The cause of the issue is still under investigation. A workaround is available.

After Dell Automation Platform (DAP) OVA upgrade from 2.0 to 2.1.0.1, where the Dell Distributed Private Cloud (DDPC) service version is already 4.3.0.0, iDRAC IP address is not displayed and "Get Credentials" button is disabled for PowerEdge 17G devices. The issue is caused by Redfish API failures.

This article explains the procedure to open a Service Request (SR) using the Dell support page. Get help fast for your Dell device. No more waiting on hold. No more tech headaches. Dell makes it simple to fix problems, schedule repairs, or get support—right from your device or our support website.

A NativeEdge device shows “No Data Available” in the Dell Automation Platform (DAP) Metrics tab because the telegraf handler depends on a machineID message from the Orchestrator (EO), which may be missing or delayed.

During a software bundle deployment, a host firewall may be configured to block traffic on Kubernetes pod network interfaces. This prevents the Container Network Interface (CNI) plugin from communicating. This issue is observed with McAfee Endpoint Security firewall but may occur with any host firewall solution.

The NativeEdge Outcome node firmware bundle is showing Unknown for PowerEdge systems which contain a HBA355i.

This article addresses an issue in NativeEdge Orchestrator (NEO) version 3.1 and also Dell Automation Platform versions 1.0 - 2.1. The system becomes unavailable due to expired mTLS certificates. The issue affects systems that have been running for longer than one year. The internal mTLS certificate rotation mechanism fails to renew certificates in time, causing widespread pod startup failures and making the NEO user interface inaccessible. This issue was identified as affecting all systems with expiring mTLS certificates approximately one year after initial deployment.

In NativeEdge environments, attempting to enable USB controller passthrough for a virtual machine (VM) on Gateway 3200 (EGW3200) results in VM update failure. This behavior is not a defect but a platform limitation, as USB controller passthrough is only supported on EGW5200 devices and not on EGW3200.

Expected storage volumes (virtual disks) are not created immediately when a DisableRAID job run through the ECE Agent completes with a SUCCESS status on Dell PowerEdge devices with a RAID controller.

Firmware upgrade process for XR5610 systems fails because the systems have firmware versions that do not match the expected firmware lock versions. The Native Edge Orchestrator firmware update service cannot find a valid upgrade path for the devices because the current versions on the systems are not recognized in the available catalog.

This article describes issues seen during an attempted default TLS certificate with a custom certificate using Portal UI.

Dell Native Edge Orchestrator (EO) remediation is available for multiple security vulnerabilities that could be exploited by malicious users to compromise the affected system.

A NativeEdge Asset reports an incorrect IP address for its management address.

This article explains an issue where the kernel drivers fail to detect one of the NVIDIA Graphics Processing Units (GPU) in a Unified Extensible Firmware Interface (UEFI) Virtual Machine (VM) configured with multiple GPU passthroughs.

When updating a deployment the Deployment_id input update requires manual change of the deployment to allow the update to complete.

Upload of Dell Private Cloud Zip bundle to Orchestrator fails with "Invalid blueprint" error.

A NativeEdge Virtual Machine (VM) which is imported from VMware vCenter with multiple disks can fail to boot.