How to Fix “Failed to Start Network Name Resolution” in Ubuntu

Booting up an Ubuntu desktop or server only to notice a red failure line in your systemd startup log stating “Failed to start Network Name Resolution” (or discovering that systemd-resolved.service has entered a failed state) triggers an immediate total internet outage. While your network card may possess a valid IP address and ping gateway IPs successfully, web browsers display “Server not found”, and executing sudo apt update terminates with “Temporary failure in name resolution”.

The systemd-resolved daemon is Ubuntu’s native network name resolution manager, listening on local loopback address 127.0.0.53:53 to handle DNS lookups for all system applications. When this service fails to start, the failure is almost universally caused by broken symbolic links in /etc/resolv.conf, port 53 socket conflicts with third-party DNS software (such as dnsmasq, Pi-hole, or BIND9), or syntax errors in /etc/systemd/resolved.conf. This tutorial walks you through practical terminal commands to repair systemd-resolved and restore DNS resolution.

Method 1: Inspect Service Failure Logs via journalctl

Before modifying network configuration files, query the systemd journal to discover the exact reason systemd-resolved refused to start.

1. Open your terminal by pressing Ctrl + Alt + T (or log in via local console).
2. Inspect the status of the name resolution service:

   systemctl status systemd-resolved.service

3. Query the systemd journal for specific failure logs:

   journalctl -u systemd-resolved.service -b --no-pager

4. Look for error indicators such as:
   • “Failed to bind UDP socket to 127.0.0.53:53: Address already in use” (Port 53 conflict).
   • “Failed to parse config file” (Syntax error in resolved.conf).

Method 2: Recreate the Canonical /etc/resolv.conf Symlink

In Ubuntu, /etc/resolv.conf must never be a static text file; it must be a symbolic link pointing to systemd-resolved’s dynamic stub resolver file. If third-party software overwrote this symlink with a static file, name resolution fails.

1. Check if /etc/resolv.conf is a symlink or a regular file:

   ls -la /etc/resolv.conf

2. If the output does not show /etc/resolv.conf -> ../run/systemd/resolve/stub-resolv.conf, remove the static or corrupted file:

   sudo rm -f /etc/resolv.conf

3. Recreate the proper symbolic link:

   sudo ln -s /run/systemd/resolve/stub-resolv.conf /etc/resolv.conf

4. Restart the service:

   sudo systemctl restart systemd-resolved

5. Test DNS resolution by pinging a domain:

   ping -c 3 ubuntu.com

Method 3: Resolve Port 53 Socket Conflicts (dnsmasq / Docker)

If another application—such as a local caching DNS forwarder, Pi-hole, dnsmasq, or a Docker container—is already listening on port 53, systemd-resolved cannot bind its stub listener.

1. Identify which process is holding port 53:

   sudo ss -tulpn | grep ':53 '

2. If a process such as dnsmasq or named is listed, determine whether that service is required. If unintended, stop and disable it:

   sudo systemctl stop dnsmasq && sudo systemctl disable dnsmasq

3. Alternatively, if you must run a custom DNS server locally, configure systemd-resolved to disable its local stub listener:

   sudo nano /etc/systemd/resolved.conf

4. Under the [Resolve] heading, set:

   DNSStubListener=no

5. Save the file (Ctrl + O, then Ctrl + X) and restart the service:

   sudo systemctl restart systemd-resolved

Method 4: Specify Static Fallback DNS Servers in resolved.conf

If your router DHCP fails to supply valid upstream DNS nameservers, configure systemd-resolved with trusted public DNS fallback servers.

1. Open the systemd-resolved configuration in nano:

   sudo nano /etc/systemd/resolved.conf

2. Uncomment and update the DNS and FallbackDNS directives:

   [Resolve]
   DNS=1.1.1.1 8.8.8.8
   FallbackDNS=1.0.0.1 8.8.4.4

3. Save and exit nano.
4. Enable and restart the service cleanly:

   sudo systemctl enable --now systemd-resolved

5. Verify that DNS queries resolve:

   resolvectl status

Ubuntu systemd-resolved Troubleshooting Reference Table

Use the matrix below to assess resolution failure modes and their proven solutions:

Observed Failure State	Underlying Root Cause	Standard Resolution Command
“Failed to bind UDP socket: Address in use”	Another service (dnsmasq/named) bound to port 53	sudo ss -tulpn | grep :53 & disable conflicting service
/etc/resolv.conf is a broken link	Missing /run/systemd/resolve/stub-resolv.conf	sudo ln -sf /run/systemd/resolve/stub-resolv.conf /etc/resolv.conf
resolvectl status shows zero DNS servers	DHCP client failed to advertise DNS from router	Define static DNS=1.1.1.1 in /etc/systemd/resolved.conf
Temporary failure in name resolution on boot	Network online target triggered before resolved	sudo systemctl enable systemd-resolved
Ping IP works, ping domain name fails	Stalled DNS query cache in systemd-resolved	sudo resolvectl flush-caches

Flushing the Systemd DNS Cache

Once systemd-resolved is active, cached negative DNS responses can cause applications to continue reporting “Temporary failure in name resolution” for several minutes. Flush the local cache immediately by executing: sudo resolvectl flush-caches. Verify cache statistics by running resolvectl statistics to ensure the cache has been reset to zero.