Diagnostics

Network~3 minView script

Three one-shot read-only checks that inspect the live network state: routing table, connectivity test and advanced statistics. None of them write to /etc/network/interfaces — safe to use at any time, including over SSH. For interactive live monitoring, see Live monitoring tools.

What this does

Reports on routing, reachability and aggregate statistics without modifying anything. Use it to confirm the network is healthy or identify duplicate IPs. For continuous traffic observation see Live monitoring tools; for viewing the raw config file see Interfaces backup & restart.

Show Routing Table

Lists every route the kernel currently uses (ip route show) and highlights the default gateway. Useful for confirming that traffic to the internet leaves through the expected interface and that no leftover routes are taking precedence.

Total routes: 4

➡  default via 192.168.1.1 dev vmbr0 onlink
   • 10.10.10.0/24 dev vmbr1 proto kernel scope link src 10.10.10.1
   • 169.254.0.0/16 dev vmbr0 scope link metric 1000
   • 192.168.1.0/24 dev vmbr0 proto kernel scope link src 192.168.1.10

🌍 Default Gateway: 192.168.1.1

Test Connectivity

Sends two ICMP probes to three targets in sequence and reports each one. Then runs an nslookup google.com to verify DNS resolution independently of ICMP.

TestTargetWhat it confirms
Google DNS8.8.8.8External reachability via well-known anycast IP
Cloudflare DNS1.1.1.1Independent second-opinion if Google blocks ICMP
Gateway(default route)L2 / L3 connectivity to the local router
DNS Resolutiongoogle.comDNS server is reachable and answering

Reading the result

If the gateway test passes but the public DNS targets fail, you have local connectivity but no internet — check the gateway's upstream. If DNS resolution fails but pings to 8.8.8.8 succeed, the issue is the resolver, not the network: check /etc/resolv.conf.

Advanced Diagnostics

Aggregates network-wide statistics and runs a battery of common-issue checks. Reports active connection counts, listening ports, total interfaces and flags two anti-patterns explicitly:

  • NetworkManager running on a Proxmox host. NetworkManager and Proxmox's ifupdown conflict — both try to manage interfaces, leading to bridges that flap or refuse to come up. If detected, the tool offers an interactive prompt to stop, disable and purge it (this is the only diagnostic that can modify the system, and only after explicit consent).
  • Duplicate IP addresses on different interfaces. Two interfaces holding the same IPv4 cause intermittent connectivity that's extremely hard to debug from inside the VM. Detected via ip -4 addr show | sort | uniq -d.

The NetworkManager prompt is the only modifying action here

Every other check in Advanced Diagnostics is read-only. NetworkManager removal is gated behind a yes/no dialog and runs systemctl stop / disable + apt-get purge -y network-manager. If you decline, nothing changes.

Troubleshooting

Test Connectivity reports the gateway as failed but the host has internet

Some routers do not respond to ICMP from their internal interface even when they happily forward traffic. Test with traceroute 8.8.8.8 from a console: if the second hop is your ISP and the public DNS test passed, the gateway is fine despite the failed ping.

Advanced Diagnostics keeps flagging duplicate IPs after I removed one

The check looks at the live kernel state, not the config file. Run ip -4 addr show to confirm; if the duplicate is still present, an interface is still holding it. Use ip addr del <IP>/<mask> dev <iface> to remove it from the live state, then edit /etc/network/interfaces if it is also persistent.

Related