This article helps you identify and resolve unstable server connection issues. Follow the diagnostic steps for the category that best matches your symptoms.

1. Local / Client-Side Issues

Connection problems sometimes originate on your end. Check the following before investigating the server.

Local Network Issues

1. Restart your router and modem, then wait 60 seconds before reconnecting.
2. Test your internet connection by visiting a reliable website or running a speed test.
3. Try connecting from a different network (e.g. mobile hotspot) to rule out your local network as the cause.
4. If you are on Wi-Fi, switch to a wired connection to eliminate wireless interference.

VPN or Firewall Interference

1. Temporarily disable any active VPN and attempt to reconnect to the server.
2. Check your local firewall or security software to ensure it is not blocking the connection port.
3. If using a corporate firewall, contact your network administrator to verify outbound traffic is permitted.

DNS Resolution Issues

1. Try connecting to your server using its IP address directly instead of the hostname to bypass DNS.
2. Flush your local DNS cache:

• Windows: ipconfig /flushdns
• macOS: sudo dscacheutil -flushcache
• Linux: sudo systemd-resolve --flush-caches

3. Switch to a public DNS resolver such as 1.1.1.1 (Cloudflare) or 8.8.8.8 (Google) and test again.

2. Network Path / Routing Issues

Packet loss between your machine and the server can cause intermittent drops or high latency.

Packet Loss Along the Route

1. Run a traceroute or MTR to your server to identify where in the network path loss is occurring (see Section 6 – Diagnostic Tools).
2. Share the output with Contabo Support so we can investigate the affected network segment.

Issue on Contabo’s Side

1. Check the Contabo Status Page for any active incidents or maintenance affecting your region.
2. If an incident is listed, monitor the status page for updates. No further action is required on your end.
3. If no incident is listed but you suspect a routing problem on our infrastructure, contact Contabo Support with your traceroute/MTR results.

3. Server Network Configuration

Misconfigured settings on the server itself can cause connections to be dropped or refused.

Incorrect Network Interface Settings

1. Log in via the Contabo VNC/KVM console (bypasses the network) and verify the network interface is up: ip link show
2. Confirm the correct IP address, subnet mask, and gateway are configured: ip addr show and ip route show
3. Restart the networking service if settings appear correct but the interface is down: sudo systemctl restart networking (Debian/Ubuntu) or sudo systemctl restart NetworkManager (CentOS/RHEL).

Firewall Rules Dropping Packets

1. Review your firewall rules to ensure the required ports are not inadvertently blocked: sudo iptables -L -n -v or sudo ufw status verbose
2. Temporarily flush rules in a test environment to confirm the firewall is the cause: sudo iptables -F (use with caution – this removes all rules).
3. Re-add only the necessary rules and verify connectivity after each change.

NIC Issues or Driver Problems

1. Check kernel logs for NIC errors: sudo dmesg | grep -i eth or sudo journalctl -k | grep -i net
2. If driver errors appear, contact Contabo Support – NIC-level issues on virtual or dedicated servers require our intervention.

4. DDoS or Traffic Flooding

A DDoS attack or unexpected traffic surge can overwhelm your server and cause connection instability.

1. Check your server’s incoming traffic metrics in the Contabo Customer Control Panel to identify unusual spikes.
2. Inspect active connections: ss -nt | awk '{print $5}' | cut -d: -f1 | sort | uniq -c | sort -rn | head -20
3. If a DDoS attack is confirmed or suspected, contact Contabo Support immediately. We can apply network-level mitigation measures.
4. As a short-term measure, consider implementing rate limiting or blocking the offending IP ranges using your firewall.

5. Resource Exhaustion Affecting Network

High CPU or RAM usage can cause the server’s network stack to become unresponsive, resulting in connection timeouts.

CPU / RAM Overload

1. Check current resource usage: top or htop
2. Identify and stop any runaway processes consuming excessive CPU or memory.
3. If the server itself is consistently within normal limits but timeouts persist, the host node may be overloaded.

⚠️  Note: Host node overload cannot be resolved by the customer. Please contact Contabo Support so we can assess the situation and migrate your server to a less loaded host if necessary.

Network Bandwidth Overload

1. Monitor current bandwidth usage: vnstat -i eth0 -l or nload eth0
2. Identify processes generating the most traffic: nethogs or iftop
3. Throttle or stop high-bandwidth processes where possible.
4. If you are regularly hitting your bandwidth limit, consider upgrading your plan or reviewing application traffic patterns.

See How to reduce outgoing traffic for more information.

6. Diagnostic Tools

The following tools help pinpoint where a connection issue originates. For a full diagnostics guide, see How can I run diagnostics on my server?.

traceroute / tracert

Shows each hop between your machine and the server, and reports latency and packet loss per hop. Use this to narrow down which network segment is causing problems.

• Linux / macOS: traceroute <server-ip>
• Windows: tracert <server-ip>

MTR 

Combines the functions of traceroute and ping into a continuously updated report. Provides more reliable data on intermittent packet loss.

• Linux: mtr <server-ip> (install via sudo apt install mtr or sudo yum install mtr)
• Windows: WinMTR (GUI application, available at winmtr.net)
• Run for at least 60 seconds and save the output before contacting support.

Still Having Issues?

If you have worked through the steps above and the problem persists, please contact Contabo Support and include:

• Your server IP address and hostname
• A description of the symptoms and when they started
• Output from traceroute / MTR
• Any relevant log entries or error messages

This helps our team diagnose the issue faster and provide an accurate resolution.