Network Operations Centre

Fleet-wide host visibility with investigation actions through QuoxBastion.

Overview

The Network Operations Centre (NOC) is a QuoxCORE plugin that provides real-time fleet monitoring from within the dashboard. It connects to QuoxBastion to retrieve host status, exposes seven investigation actions for diagnosing individual hosts, and renders conditional monitoring panels for infrastructure tooling you already run.

NOC ships with every QuoxCORE instance. There is no separate purchase or activation step.

Tier: Free Price: Included with every QuoxCORE instance

Fleet Status

The primary view is a fleet status board populated by QuoxBastion. Each host in your fleet appears as a row with a health indicator derived from the most recent liveness check.

Health Indicators

Each host displays one of three states:

Health data is fetched from QuoxBastion's fleet_status tool. The NOC plugin does not open its own SSH connections or send traffic directly to fleet hosts. All communication routes through the bastion.

Host Actions

Selecting a host reveals seven investigation actions. Each action has an assigned severity level and requires confirmation before execution.

1. Health Check

Severity: Low

Runs a liveness and connectivity probe against the selected host. Returns round-trip latency and confirms whether the host is reachable from the bastion.

2. Disk Usage

Severity: Low

Reports storage capacity and utilisation across all mounted volumes. Output includes filesystem type, total size, used space, and percentage full for each mount point.

3. CPU and Memory

Severity: Low

Returns real-time CPU load averages and memory utilisation. Values are point-in-time readings taken at the moment of execution, not historical averages.

4. Network Status

Severity: Low

Lists network interfaces with their operational state, assigned addresses, and basic connectivity metrics. Useful for confirming whether a host has lost its uplink or has an interface in a down state.

5. Backup Health

Severity: Low

Checks the timestamps of the most recent backups for the host. Reports the last successful backup time and flags any backup older than the expected schedule window.

6. Running Services

Severity: Medium

Enumerates active services on the host. Returns service names, current state (running, stopped, failed), and uptime. This action is classified as medium severity because it queries the service manager, which may briefly increase load on a struggling host.

7. Error Logs

Severity: Medium

Extracts recent error-level entries from system logs. The output is filtered to the most recent window (typically the last hour) and limited in volume to avoid transferring large payloads through the bastion.

Confirmation Flow

Every action displays a confirmation dialog before execution. The dialog shows the action name, the target host, and the severity level. You must explicitly confirm to proceed. There is no way to bypass confirmation or batch-execute actions across multiple hosts without individual approval.

All actions execute through QuoxBastion via the fleet_status tool. Every invocation is recorded in the audit log with the requesting user, target host, action type, and timestamp.

AI Analysis

After an action completes, its output is available to CommanderQ for interpretation. You can send the results directly to CommanderQ, which will analyse the data, highlight anything unusual, and suggest next steps.

This is particularly useful for error log output or resource utilisation data where the raw numbers benefit from contextual interpretation. CommanderQ can correlate findings across multiple actions -- for example, connecting high memory usage from action 3 with a failing service from action 6.

Monitoring Panels

The NOC includes a Monitoring tab that renders panels for external monitoring tools. Panels are conditional: they only appear if the corresponding integration is configured in your QuoxCORE instance.

Supported Panels

If none of these integrations are configured, the Monitoring tab displays a message indicating no panels are available, with links to the relevant configuration pages.

Auto-Refresh

All monitoring panels refresh their data every 30 seconds. The refresh cycle is synchronised across panels so that all data updates at the same time, reducing visual inconsistency between panels showing related metrics.

Multi-Monitor

NOC supports popping out the monitoring view into a separate browser window. This is designed for operators running multi-monitor setups who want the fleet status on one screen and monitoring panels on another.

Popout

Click the popout control on any NOC view to open it in a dedicated browser window. The popout window receives the same data and updates as the inline view.

Cross-Tab Synchronisation

Popout windows stay in sync with the main dashboard using the browser's BroadcastChannel API. Actions triggered in the main window (such as running a host investigation) are reflected in the popout, and vice versa. If you close the main dashboard tab, the popout continues to operate independently and will resynchronise when you reopen the dashboard.

Security

NOC is designed around a single principle: no direct access from the browser to fleet hosts.

Bastion-Only Execution

All host actions route through QuoxBastion. The browser sends a request to QuoxCORE, which forwards it to QuoxBastion, which executes the action on the target host. At no point does the browser open an SSH connection or communicate directly with a fleet host.

Audit Logging

Every action is logged with:

• The user who initiated the action
• The target host
• The action type and severity
• A timestamp
• The execution result (success or failure)

Audit logs are stored within QuoxCORE and are accessible to administrators. They provide a full trail of who ran what, against which host, and when.

No Stored Credentials

NOC does not store SSH keys, passwords, or host credentials. Authentication to fleet hosts is handled entirely by QuoxBastion, which manages its own credential store independently of the dashboard.