PYCURL ERROR 6 - "Couldn't resolve host 'mirrorlist.centos.org'"

Resolving PycURL Error 6: 'Couldn't resolve host' on CentOS/Hyper-V

This article provides a comprehensive guide to troubleshoot and resolve the 'Couldn't resolve host' error (PycURL Error 6) commonly encountered on CentOS virtual machines, especially when hosted on Hyper-V. Learn how to diagnose DNS issues, configure network settings, and ensure proper connectivity.

The 'Couldn't resolve host' error, specifically PycURL Error 6, is a common networking problem that indicates your system is unable to translate a hostname (like mirrorlist.centos.org) into an IP address. This is fundamentally a DNS (Domain Name System) resolution failure. While it can occur in various environments, it's frequently observed in virtualized settings such as CentOS running on Hyper-V, where network configurations might not be immediately obvious or correctly set up.

Understanding the Root Cause: DNS Resolution Failure

When you encounter 'Couldn't resolve host', it means your operating system's network stack cannot find the IP address associated with the requested hostname. This process typically involves querying DNS servers. Several factors can lead to this failure:

1. Incorrect DNS Server Configuration: The VM might not be configured to use valid and reachable DNS servers.
2. Network Connectivity Issues: The VM might not have a proper network connection to reach any DNS server or the internet.
3. Firewall Restrictions: A firewall (either on the VM, Hyper-V host, or external network) might be blocking DNS queries (port 53 UDP/TCP).
4. Hostname Typos: Although less common for well-known hosts like mirrorlist.centos.org, a typo in the hostname would also result in this error.
5. Hyper-V Virtual Switch Configuration: The virtual switch used by the VM might not be correctly configured to provide internet access or DNS forwarding.

flowchart TD
    A[Application Request 'mirrorlist.centos.org'] --> B{OS DNS Resolver}
    B --> C{Query Configured DNS Servers}
    C -- DNS Server Unreachable --> D[PycURL Error 6: 'Couldn't resolve host']
    C -- DNS Server Reachable but No Response --> D
    C -- DNS Server Responds with 'Host Not Found' --> D
    C -- DNS Server Responds with IP --> E[Connect to IP Address]
    E --> F[Success]

Flowchart illustrating the DNS resolution process and potential failure points leading to 'Couldn't resolve host'.

Diagnosing Network and DNS Issues on CentOS

Before attempting fixes, it's crucial to diagnose where the problem lies. Start by checking basic network connectivity and then move to DNS-specific tests.

ping -c 4 8.8.8.8

Test basic internet connectivity to Google's public DNS server.

If the ping command fails, your VM likely has no internet access. If it succeeds, the issue is more specific to DNS resolution.

cat /etc/resolv.conf

Check the configured DNS servers on your CentOS system.

This file lists the DNS servers your system uses. Ensure the IP addresses listed are correct and reachable. Common public DNS servers include 8.8.8.8 (Google) or 1.1.1.1 (Cloudflare).

dig mirrorlist.centos.org
nslookup mirrorlist.centos.org

Use dig or nslookup to explicitly test DNS resolution for the problematic hostname.

If these commands fail to resolve the hostname, it confirms a DNS resolution problem.

Resolving the 'Couldn't resolve host' Error

Based on your diagnosis, apply the following solutions.

1. Step 1: Configure DNS Servers Manually

If /etc/resolv.conf is empty or contains incorrect DNS servers, you can manually edit it. However, on modern CentOS systems using NetworkManager, it's better to configure DNS via NetworkManager or the network interface configuration files to ensure persistence across reboots.

For temporary testing, you can edit /etc/resolv.conf directly:

sudo vi /etc/resolv.conf

Add or modify the nameserver lines:

nameserver 8.8.8.8
nameserver 8.8.4.4

For a persistent change, edit the network interface configuration file (e.g., /etc/sysconfig/network-scripts/ifcfg-eth0 or ifcfg-ens192):

sudo vi /etc/sysconfig/network-scripts/ifcfg-eth0

Add or modify the DNS1 and DNS2 lines:

DNS1=8.8.8.8
DNS2=8.8.4.4
PEERDNS=no

After modifying, restart the network service:

sudo systemctl restart network

2. Step 2: Verify Hyper-V Virtual Switch Configuration

On the Hyper-V host, ensure the virtual machine is connected to a virtual switch that provides internet access. This typically means an 'External' virtual switch bound to a physical network adapter with internet connectivity. If using an 'Internal' or 'Private' switch, you'll need to configure NAT or routing on the host to provide internet access to the VM.

1. Open Hyper-V Manager.
2. Go to 'Virtual Switch Manager'.
3. Ensure the virtual switch connected to your CentOS VM is of type 'External' and is bound to the correct physical network adapter that has internet access.

3. Step 3: Check Firewall Settings

Both the CentOS VM's firewall (firewalld or iptables) and any external firewalls (e.g., on the Hyper-V host or network router) can block DNS traffic. Temporarily disable the firewall on the CentOS VM to rule it out:

sudo systemctl stop firewalld
sudo systemctl disable firewalld

If this resolves the issue, re-enable the firewall and configure it to allow outgoing DNS traffic (UDP port 53).

4. Step 4: Review Network Adapter Settings in Hyper-V

Sometimes, the virtual network adapter settings within Hyper-V for the specific VM can cause issues. Ensure the network adapter is enabled and configured correctly.

1. In Hyper-V Manager, right-click your CentOS VM and select 'Settings'.
2. Navigate to 'Network Adapter'.
3. Ensure it's connected to the correct virtual switch.
4. Consider removing and re-adding the network adapter if other steps fail, or try using a 'Legacy Network Adapter' for older CentOS versions if you suspect driver issues.

Advanced Troubleshooting: DHCP and NetworkManager

If your CentOS VM is configured to use DHCP, ensure the DHCP server is correctly providing DNS server addresses. If NetworkManager is managing your network, you can use nmcli to inspect and modify settings.

nmcli device show eth0 | grep DNS
nmcli connection show 'System eth0' | grep dns

Check DNS settings via NetworkManager for the 'eth0' interface (replace 'eth0' with your actual interface name).

To set DNS servers using nmcli for a specific connection (e.g., 'System eth0'):

sudo nmcli connection modify 'System eth0' ipv4.dns "8.8.8.8 8.8.4.4"
sudo nmcli connection modify 'System eth0' ipv4.dns-search "example.com"
sudo nmcli connection up 'System eth0'

Configure static DNS servers using nmcli and bring the connection up.