{
"meta": {
"title": "Proxmox Network Management — Bridges, Bonds, Diagnostics, Repair | ProxMenux",
"description": "Read-only diagnostics, analyze-then-suggest workflows and guided repairs for the Debian / Proxmox network stack (/etc/network/interfaces). Inspect bridges and bonds, run live monitoring tools, persist interface names, back up and safely restart networking — with mandatory backups and step-by-step previews.",
"ogTitle": "Proxmox Network Management — Bridges, Bonds, Diagnostics, Repair",
"ogDescription": "Diagnostics, analysis and guided repairs for the Proxmox network stack with mandatory backups and step-by-step previews.",
"twitterTitle": "Proxmox Network Management | ProxMenux",
"twitterDescription": "Bridges, bonds, diagnostics and guided repairs for the Proxmox network stack with mandatory backups."
},
"header": {
"title": "Network Management",
"description": "Read-only diagnostics, analyze-then-suggest reports and guided repairs for the classic Debian/Proxmox network stack. Every destructive flow takes a backup first and previews the exact changes before applying them.",
"section": "Network"
},
"intro": {
"title": "What this menu is for",
"body": "Inspect, diagnose and (when needed) repair the Proxmox host network configuration without losing remote access. The tools are designed around one principle: read first, propose, then apply only with explicit consent and a safety backup. There is no \"auto-fix everything\" button."
},
"openingMenu": {
"heading": "Opening the menu",
"intro": "From ProxMenux's main menu, select Network. You will see this:",
"imageAlt": "Network Management menu with diagnostics, monitoring, analysis, repair, persistent names and backup options"
},
"safety": {
"heading": "The safety model",
"body": "Editing network configuration on a remote Proxmox host is one of the easiest ways to lock yourself out. ProxMenux treats every action accordingly. The tools fall into three behavioural tiers — pick the card that matches your intent:"
},
"tiers": {
"readOnly": {
"title": "Read-only",
"body": "Pure inspection. Cannot modify the system under any circumstance.",
"items": [
"Routing table, connectivity tests, advanced statistics",
"Live traffic monitoring (iftop, iptraf-ng)",
"Bandwidth test (iperf3)"
]
},
"analyze": {
"title": "Analyze, then suggest",
"body": "Detects issues, prints a report with proposed shell commands, and stops. You decide whether to enter the guided repair afterwards.",
"items": [
"Bridge configuration analysis",
"Network configuration analysis"
]
},
"apply": {
"title": "Apply with backup",
"body": "Modifies the system. Always takes a timestamped backup of the affected file first and shows a preview before writing.",
"items": [
"Persistent interface names (.link files)",
"Manual backup, restore and service restart"
]
}
},
"classicTitle": "Classic stack only",
"classicBody": "Every analysis and repair function checks the active network manager before touching anything. If the host runs netplan, systemd-networkd or NetworkManager, the tool aborts immediately with a clear message — the menu only supports the classic Debian/Proxmox stack at /etc/network/interfaces. This is intentional: editing a netplan file with rules written for ifupdown would silently corrupt the configuration.",
"backups": {
"heading": "Where backups go",
"intro": "Every guided repair, restore or manual backup writes a timestamped copy of /etc/network/interfaces to /var/backups/proxmenux/:",
"rollbackIntro": "To roll back manually from a console:"
},
"readOnlySection": {
"heading": "Read-only inspection",
"body": "The starting point when something feels off. Pure inspection — never writes to /etc/network/interfaces and never runs a modifying command (with one explicit, opt-in exception for purging NetworkManager when detected). Safe to use over SSH at any time.",
"options": [
{
"title": "Diagnostics",
"description": "Three one-shot read-only checks: Show Routing Table, Test Connectivity and Advanced Diagnostics. Pure inspection — never writes to the system."
},
{
"title": "Live monitoring tools",
"description": "Three interactive launchers: iftop (real-time bandwidth per host pair), iptraf-ng (multi-mode traffic monitor) and iperf3 (bandwidth test, server / client mode)."
}
]
},
"analyzeSection": {
"heading": "Analyze, then suggest",
"body": "Used when an inspection (or a real outage) points at a configuration issue. Each tool walks the relevant part of /etc/network/interfaces, prints a detailed report with the exact shell command that would fix each finding, and stops. If you accept the optional guided repair afterwards, every change is backed up and previewed before being written.",
"options": [
{
"title": "Bridge analysis & guided repair",
"description": "Detects vmbrX bridges with missing or invalid ports (typical after PCI re-enumeration). Shows a report first; only repairs when you accept the 5-step guided flow."
},
{
"title": "Config analysis & guided cleanup",
"description": "Finds physical interfaces declared in /etc/network/interfaces that no longer exist (orphan configs left behind by hardware changes). Reports them and offers a guided removal."
}
]
},
"applySection": {
"heading": "Apply with backup",
"body": "Tools that write to disk by design. Each one takes a timestamped backup of the affected file before writing, and the destructive options (restore, restart) require an explicit yes/no confirmation. Persistent interface names takes effect at the next reboot, not immediately, so it is safe to schedule even on a remote host.",
"options": [
{
"title": "Persistent interface names",
"description": "Pins interface names (eno1, enp3s0, …) to MAC addresses via systemd .link files. Names survive PCI slot changes, kernel upgrades and adding / removing other NICs."
},
{
"title": "Interfaces backup & restart",
"description": "Manual snapshot of /etc/network/interfaces, browse and restore previous backups, view the live config, and restart the networking service when needed."
}
]
},
"consoleTitle": "Have console access ready",
"consoleSubTitle": "Before any repair on a remote host",
"consoleBody": "If you are connected over SSH and only have one path to the host, have a fallback before applying network changes: physical / IPMI / iKVM console, or another machine on the same LAN. The guided repairs are safe and always offer a roll-back command, but a misconfigured bridge or a dropped link can still leave you locked out until you can reach the console."
}