Network issues on the server Device experiences unstable internet connection Possible causes and solutions
Network issues on the server Device experiences unstable internet connection Possible causes and solutions
Hey, I've been facing some issues with my server's internet connection lately. It's running Ubuntu Server 20.04, and what seems to be happening is that the internet works just fine for a while, but then suddenly it stops without any clear reason or my actions. This has started recently, especially when using Healthcheck.io to track my cron jobs. One of them updates my subdomain IP, which requires internet access to function properly. Without it, I don't receive notifications. The cron job runs every hour.
I'm managing the IP through NetworkManager since I use Cockpit for server management. My LAN connection is fine, and even the DNS server works correctly. All devices that rely on my server's IP can resolve DNS queries. I'm using PiHole as a DNS provider but not as a DHCP server. Plex also works outside the office—when I'm on LTE, I can connect and stream without issues. However, I can establish an SSH session over LTE using my WAN IP, but my server can't download packages, update software, or run speed tests.
When I ping 8.8.8.8, it works fine. This seems to be caused by the interface flags on my network card (enp9s0) being set to UP, BROADCAST, RUNNING, and MULTICAST. The MTU is 1500, and the IP is 192.168.2.10 with a 255.255.255.0 mask. Everything else looks normal.
Begin with simple connectivity checks when the connection drops. Verify DNS settings, ping the gateway and a public host (such as 1.1.1.1 or 8.8.8.8), use traceroutes, and examine ICMP and TCP/UDP traffic. See if other devices, servers, or clients are impacted. Determine when and how it recovers. Look for recurring patterns.
Certainly! Here’s the complete Netplan or interfaces file for you to share.
Thank you for your message. No network configuration files are being applied since I'm using NetworkManager instead. Feel free to share the details if it helps. The devices listed are: enp9s0-04111eca-7087-4cd5-8e77-352c6c21e726.nmconnection and enp9s0.nmconnection.
Hello, thank you for your message. When the connection drops, I’ve tried pinging 8.8.8.8 and 1.1.1.1 and always get a response. None of my other devices are affected, even those linked to my DNS server on my own machine. They still have full internet access. The issue seems to happen around every 12 hours, and I’ll share the heatcheck.io logs if you upload them. I’ll also use the tools you suggested once the problem resurfaces.
The files are properly configured, but as mentioned by @jj9987, consider running a trace route to check for any unusual activity.
When using traceroute6 to google.com, the output indicates a name or service is unknown. It seems you might be experiencing connectivity issues or lack of internet access, which could explain the results.
It's just the default package that was set up. Restarting helps restore my connection briefly, allowing me to run traceroute. The output I see when using traceroute now differs from what I previously observed. This is what appears after executing the same command with traceroute instead of *6 traceroute to google.com.