Core Fabric Status Guide#

Prerequisites#

Access to the VergeOS interface with node management privileges
Basic understanding of VergeOS core network architecture
Physical console or IPMI access to nodes (for troubleshooting)
Knowledge of your core VLAN assignments (Core 1 and Core 2) and expected NIC link speed

What is the Core Fabric?#

The fabric is the backbone of your VergeOS system, utilizing the core network to manage all node-to-node communications, including vSAN traffic, peer discovery, management operations, cross-node network traffic, VM and network migrations, and other functions.

A typical VergeOS deployment uses two independent physical core networks ("Core 1 Switch", "Core 2 Switch") for redundancy. Each node should have two independent physical paths to every other node in the cluster.

Zero Switch Hops Required

All nodes must be connected to the same switching fabric with zero switch hops between them. Target latency between nodes on core fabric networks is less than 0.05ms. Adding switch hops introduces latency that degrades fabric scores and cluster performance.

This core fabric redundancy is vital to maintain system resiliency and uninterrupted operation, even during a node or drive failure, and allows maintenance operations without downtime.

Core Fabric MTU Requirements:

Component	MTU
Physical switch port	>= 9216
Physical NIC	9192 (typical)
VXLAN overlay	NIC MTU minus 50 bytes overhead

How Core Fabric Redundancy Works

The core fabric handles redundancy at a low level, creating a mesh where every node maintains redundant paths to every other node in the system. Because of this built-in redundancy, physical LAG or port bonding should not be used on core fabric networks — doing so will interfere with the fabric's own mechanisms.

The VergeOS core fabric provides more comprehensive detection and resiliency than traditional link aggregation. LAG only detects and protects against link-level failures, while the VergeOS fabric operates at the application layer, detecting a much wider range of problems including dropped packets, MTU mismatches, NIC lockups, and bad firmware — in addition to simple disconnected links.

Accessing Fabric Status (UI)#

Fabric status is available in the VergeOS UI at multiple levels of detail.

Method	Detail Level	Use Case
Alarms	Summary	Day-to-day monitoring — alerts when paths are degraded or lost
Node NICs List	Per-NIC	Quick status check of all Node NICs
Node Dashboard	Per-NIC (selected node)	Quick status check of individual NICs and their connections to other nodes
Node Diagnostics	Full JSON report	Advanced troubleshooting — complete path, score, and peer details

Alarms#

On a day-to-day basis, monitoring fabric status can be handled through the same alarm system used for the rest of your VergeOS environment. A warning alarm is raised when bidirectional communication is unavailable on a core network path.

Tip

Clicking an alarm in the list will navigate directly to the affected Node Dashboard, where further detail is available.

For more information on viewing and managing alarms, see the Alarms Guide.

Address Core Network Alarms Immediately

Core network alarms indicate that your system may not have full fabric redundancy. Resolve these promptly to ensure your cluster can tolerate a failure without disruption. Event triggers can be configured to send notifications via email, text alerting systems, monitored Slack channels, and more, ensuring administrators are notified immediately. See the Task Engine Product Guide for more information about creating automated tasks; this Automation Example KB article provides an example of setting up event-driven notifications.

Node NICs List#

This is a quick way to view fabric status on all core network NICs from a single page.

Navigate to Infrastructure > Nodes.
Click NICs on the left menu.
A list of all NICs from all nodes is displayed. The Fabric Status column shows the status for core network NICs (e.g. 'Confirmed', 'No Path', 'Degraded'). A Fabric Status of 'None' is shown for NICs that do not participate in the core fabric (e.g. external networks).

Node Dashboards#

Status information is available per NIC from each Node Dashboard.

Navigate to Infrastructure > Nodes.
Double-click the desired node from the list.
Scroll down to the NICs section on the Node Dashboard. Each core fabric NIC displays either a Confirmed status indicator or a problem status message (e.g. No Path, Degraded).
For more detailed information, click the globe icon on the right. This provides a popup showing NIC details:
- Vendor, Model, Interface, and Driver
- Confirmed / No Path / Degraded status per connection to each other node in the system
- Score per connection to each other node (see Score Values below)
Each path should show Confirmed status. Any path showing No Path or Degraded indicates a connectivity issue that should be investigated and resolved.

Node Diagnostics#

More extensive fabric status details (useful for advanced troubleshooting) are available through Node Diagnostics. This returns a full JSON report of fabric status as seen by the selected node, including all discovered peers, their paths, scores, and confirmation status.

Navigate to Infrastructure > Nodes.
Select the desired node from the list.
Click Diagnostics in the left menu.
Select Fabric Configuration from the Query dropdown.
Click Send to execute.
Review the output. Key fields to check first: paths[].confirmed and paths[].score for each peer node.

Field Reference#

The following fields appear in the fabric status JSON output.

Field	Description
`$sysid`	SHA-1 hash identifying this VergeOS system (sourced from `/.system_id`)
`$last_update`	Timestamp of the most recent fabric status refresh
`syncing_time`	Top-level field indicating whether the node is currently synchronizing its clock with the cluster. This must be `false` before the node fully joins. During initial node join, it is normal for the value to be `true`.
`paths`	Array of network paths to this peer node
`paths[].ip`	IP address of the remote node on the core network
`paths[].iface`	Local network interface used to reach this path
`paths[].score`	Numeric connectivity quality score (higher is better). The maximum depends on NIC link speed — see Score Values below.
`paths[].confirmed`	Whether this path has been verified as active and reachable (`true` / `false`)
`vxlans`	VXLAN tunnel endpoints programmed for this peer. These are the overlay tunnels used for cross-node virtual network traffic.

Confirmed Status#

Value	Meaning
`true`	The path has been verified — bidirectional communication is working
`false`	The path could not be verified — connectivity is lost or never established

Score Values#

The score field represents the quality of the connection to a peer node through a specific path. The maximum score corresponds to the link speed of the core NIC — a higher score indicates a faster, healthier connection.

NIC Link Speed	Maximum Score
100 Gbps	200
50 Gbps	100
25 Gbps	50
10 Gbps	20

Interpreting Scores

A "perfect" score means the value matches the expected maximum for your NIC speed. For example, a score of 50 on a 25Gbps NIC is healthy, while a score of 50 on a 100Gbps NIC indicates degradation. Always compare the score against the maximum for your link speed.

A score significantly below the expected maximum indicates degradation - possible causes include network latency, packet loss, or suboptimal routing. A score of 0 indicates a complete loss of bidirectional communication.

Confirmed vs Score

confirmed indicates whether the path is reachable, while score reflects the quality of that path.

Healthy vs. Unhealthy Fabric Examples#

Healthy Fabric (2-Node System)#

All nodes visible, two paths each, scores at maximum for NIC speed, all confirmed:

{
    "$sysid": "68e1925057aa7c6afaf9a255dcfc623794a6398e",
    "$last_update": "03/24/2026 13:31:46",
    "syncing_time": false,
    "node2": {
        "paths": [
            { "ip": "172.16.1.2", "iface": "enp148s0f0np0", "score": 200, "confirmed": true },
            { "ip": "172.16.2.2", "iface": "enp148s0f1np1", "score": 200, "confirmed": true }
        ],
        "vxlans": ["vx2 via 172.16.1.2", "vx1 via 172.16.2.2"]
    },
    "node1": {
        "paths": [
            { "ip": "172.16.1.1", "iface": "enp148s0f0np0", "score": 200, "confirmed": true },
            { "ip": "172.16.2.1", "iface": "enp148s0f1np1", "score": 200, "confirmed": true }
        ],
        "vxlans": ["vx2 via 172.16.1.1", "vx1 via 172.16.2.1"]
    }
}

What to Look For

Every node in the cluster appears in the output (in a 4-node cluster, you should see all 4 node entries)
Each node has two paths (one per core network)
All paths show "confirmed": true
"syncing_time": false at the top level
Scores match the expected maximum for your NIC speed (e.g., 200 for 100Gbps, 50 for 25Gbps)

Degraded Fabric — Lost Redundancy#

One path missing for a node (single core network failure):

{
    "node2": {
        "paths": [
            { "ip": "172.16.1.2", "score": 200, "confirmed": true }
        ]
    }
}

Impact

The node is reachable through only one core network. If the remaining path fails, the node will lose cluster connectivity entirely. Investigate immediately.

Degraded Fabric — Low Score#

Both paths present but one showing reduced quality:

{
    "node2": {
        "paths": [
            { "ip": "172.16.1.2", "score": 200, "confirmed": true },
            { "ip": "172.16.2.2", "score": 120, "confirmed": true }
        ]
    }
}

Impact

A score below the expected maximum for your NIC speed indicates network degradation on that path. vSAN performance may be affected. Check for latency, packet loss, or switch issues on the affected core network.

Critical Fabric — Path Not Confirmed#

A path exists but cannot be verified:

{
    "node2": {
        "paths": [
            { "ip": "172.16.1.2", "score": 200, "confirmed": true },
            { "ip": "172.16.2.2", "score": 0, "confirmed": false }
        ]
    }
}

Impact

The node has lost communication on one core network. If both paths show "confirmed": false, the node is isolated from the cluster, which will cause vSAN and workload disruption.

Critical Fabric — Missing Node#

A node that should be in the cluster does not appear in the fabric output at all.

Impact

The missing node is completely unreachable. It may be powered off, have both core NICs down, or be on a different VLAN. Check physical connectivity and node power status immediately.

Pre-Maintenance Fabric Verification#

VergeOS maintenance operations — including system updates, vSAN scale-ups, and scale-outs — require a healthy fabric as a prerequisite. Do not proceed with maintenance if the fabric is unhealthy. Resolve any issues first using the Troubleshooting section below.

A healthy fabric means:

All peer nodes are visible in the output
Each peer has two paths (one per core network)
All paths show "confirmed": true
Scores match the expected maximum for your NIC link speed
"syncing_time": false at the top level

Quick Verification

From any node, run Node Diagnostics > Fabric Configuration and confirm every peer meets the criteria above before proceeding with maintenance.