Jump to contentJump to page navigation: previous page [access key p]/next page [access key n]
documentation.suse.com / SUSE Linux Enterprise Server Documentation / Security and Hardening Guide / Network Security / Configuring a VPN Server
Applies to SUSE Linux Enterprise Server 15 SP2

24 Configuring a VPN Server

Today, Internet connections are cheap and available almost everywhere. However, not all connections are secure. Using a Virtual Private Network (VPN), you can create a secure network within an insecure network such as the Internet or Wi-Fi. It can be implemented in different ways and serves several purposes. In this chapter, we focus on the OpenVPN implementation to link branch offices via secure wide area networks (WANs).

24.1 Conceptual Overview

This section defines some terms regarding VPN and gives a brief overview of some scenarios.

24.1.1 Terminology


The two ends of a tunnel, the source or destination client.

Tap Device

A tap device simulates an Ethernet device (layer 2 packets in the OSI model, such as Ethernet frames). A tap device is used for creating a network bridge. It works with Ethernet frames.

Tun Device

A tun device simulates a point-to-point network (layer 3 packets in the OSI model, such as IP packets). A tun device is used with routing and works with IP frames.


Linking two locations through a primarily public network. From a more technical viewpoint, it is a connection between the client's device and the server's device. Usually a tunnel is encrypted, but it does need to be by definition.

24.1.2 VPN Scenarios

Whenever you set up a VPN connection, your IP packets are transferred over a secured tunnel. A tunnel can use either a tun or tap device. They are virtual network kernel drivers which implement the transmission of Ethernet frames or IP frames/packets.

Any user space program, such as OpenVPN, can attach itself to a tun or tap device to receive packets sent by your operating system. The program is also able to write packets to the device.

There are many solutions to set up and build a VPN connection. This section focuses on the OpenVPN package. Compared to other VPN software, OpenVPN can be operated in two modes:

Routed VPN

Routing is an easy solution to set up. It is more efficient and scales better than a bridged VPN. Furthermore, it allows the user to tune MTU (Maximum Transfer Unit) to raise efficiency. However, in a heterogeneous environment, if you do not have a Samba server on the gateway, NetBIOS broadcasts do not work. If you need IPv6, the drivers for the tun devices on both ends must support this protocol explicitly. This scenario is depicted in Figure 24.1, “Routed VPN”.

Routed VPN
Figure 24.1: Routed VPN
Bridged VPN

Bridging is a more complex solution. It is recommended when you need to browse Windows file shares across the VPN without setting up a Samba or WINS server. Bridged VPN is also needed to use non-IP protocols (such as IPX) or applications relying on network broadcasts. However, it is less efficient than routed VPN. Another disadvantage is that it does not scale well. This scenario is depicted in the following figures.

Bridged VPN - Scenario 1
Figure 24.2: Bridged VPN - Scenario 1
Bridged VPN - Scenario 2
Figure 24.3: Bridged VPN - Scenario 2
Bridged VPN - Scenario 3
Figure 24.4: Bridged VPN - Scenario 3

The major difference between bridging and routing is that a routed VPN cannot IP-broadcast while a bridged VPN can.

24.2 Setting Up a Simple Test Scenario

In the following example, we will create a point-to-point VPN tunnel. The example demonstrates how to create a VPN tunnel between one client and a server. It is assumed that your VPN server will use private IP addresses like IP_OF_SERVER and your client will use the IP address IP_OF_CLIENT. Make sure you select addresses which do not conflict with other IP addresses.

Warning: Use Only for Testing

This following scenario is provided as an example meant for familiarizing yourself with VPN technology. Do not use this as a real world scenario, as it can compromise the security and safety of your IT infrastructure!

Tip: Names for Configuration File

To simplify working with OpenVPN configuration files, we recommend the following:

  • Place your OpenVPN configuration files in the directory /etc/openvpn.

  • Name your configuration files MY_CONFIGURATION.conf.

  • If there are multiple files that belong to the same configuration, place them in a subdirectory like /etc/openvpn/MY_CONFIGURATION.

24.2.1 Configuring the VPN Server

To configure a VPN server, proceed as follows:

Procedure 24.1: VPN Server Configuration
  1. Install the package openvpn on the machine that will later become your VPN server.

  2. Open a shell, become root and create the VPN secret key:

    # openvpn --genkey --secret /etc/openvpn/secret.key
  3. Copy the secret key to your client:

    # scp /etc/openvpn/secret.key root@IP_OF_CLIENT:/etc/openvpn/
  4. Create the file /etc/openvpn/server.conf with the following content:

    dev tun
    secret secret.key
  5. Set up a tun device configuration by creating a file called /etc/sysconfig/network/ifcfg-tun0 with the following content:


    The notation openvpn@server points to the OpenVPN server configuration file located at /etc/openvpn/server.conf. For more information, see /usr/share/doc/packages/openvpn/README.SUSE.

  6. If you use a firewall, start YaST and open UDP port 1194 (Security and Users › Firewall › Allowed Services).

  7. Start the OpenVPN server service by setting the tun device to up:

    > sudo wicked ifup tun0

    You should see the confirmation:

    tun0            up

24.2.2 Configuring the VPN Clients

To configure the VPN client, do the following:

Procedure 24.2: VPN Client Configuration
  1. Install the package openvpn on your client VPN machine.

  2. Create /etc/openvpn/client.conf with the following content:

    dev tun
    secret secret.key

    Replace the placeholder IP_OF_CLIENT in the first line with either the domain name, or the public IP address of your server.

  3. Set up a tun device configuration by creating a file called /etc/sysconfig/network/ifcfg-tun0 with the following content:

  4. If you use a firewall, start YaST and open UDP port 1194 as described in Step 6 of Procedure 24.1, “VPN Server Configuration”.

  5. Start the OpenVPN server service by setting the tun device to up:

    > sudo wicked ifup tun0

    You should see the confirmation:

    tun0            up

24.2.3 Testing the VPN Example Scenario

After OpenVPN has successfully started, test the availability of the tun device with the following command:

ip addr show tun0

To verify the VPN connection, use ping on both client and server side to see if they can reach each other. Ping the server from the client:

ping -I tun0 IP_OF_SERVER

Ping the client from the server:

ping -I tun0 IP_OF_CLIENT

24.3 Setting Up Your VPN Server Using a Certificate Authority

The example in Section 24.2 is useful for testing, but not for daily work. This section explains how to build a VPN server that allows more than one connection at the same time. This is done with a public key infrastructure (PKI). A PKI consists of a pair of public and private keys for the server and each client, and a certificate authority (CA), which is used to sign every server and client certificate.

This setup involves the following basic steps:

24.3.1 Creating Certificates

Before a VPN connection can be established, the client must authenticate the server certificate. Conversely, the server must also authenticate the client certificate. This is called mutual authentication.

Creating certificates is not supported on SUSE Linux Enterprise Server. The following assumes you have created a CA certificate, a server certificate, and a client certificate on another system.

The server certificate is required in the PEM and unencrypted key in PEM formats. Copy the PEM version to /etc/openvpn/server_crt.pem on the VPN server. The unencrypted version needs to go to /etc/openvpn/server_key.pem.

Client certificates need to be of the format PKCS12 (preferred) or PEM. The certificate in PKCS12 format needs to contain the CA chain and needs to be copied to /etc/openvpn/CLIENT.p12. In case you have client certificates in PEM format containing the CA chain, copy them to /etc/openvpn/CLIENT.pem. In case you have split the PEM certificates into client certificate (*.ca), client key (*.key), and the CA certificate (*.ca), copy these files to /etc/openvpn/ on each client.

The CA certificate needs to be copied to /etc/openvpn/vpn_ca.pem on the server and each client.

Important: Splitting Client Certificates

If you split client certificates into client certificate, client key, and the CA certificate, you need to provide the respective file names in the OpenVPN configuration file on the respective clients (see Example 24.1, “VPN Server Configuration File”).

24.3.2 Configuring the VPN Server

As the basis of your configuration file, copy /usr/share/doc/packages/openvpn/sample-config-files/server.conf to /etc/openvpn/. Then customize it to your needs.

Example 24.1: VPN Server Configuration File
# /etc/openvpn/server.conf
port 1194 1
proto udp 2
dev tun0 3

# Security 4

ca    vpn_ca.pem
cert  server_crt.pem
key   server_key.pem

# ns-cert-type server 
remote-cert-tls client 5
dh   server/dh2048.pem 6

server 7
ifconfig-pool-persist /var/run/openvpn/ipp.txt 8

# Privileges 9
user nobody
group nobody

# Other configuration 10
keepalive 10 120
# status      /var/log/openvpn-status.tun0.log 11
# log-append  /var/log/openvpn-server.log 12
verb 4


The TCP/UDP port on which OpenVPN listens. You need to open the port in the firewall, see Chapter 23, Masquerading and Firewalls. The standard port for VPN is 1194, so you can usually leave that as it is.


The protocol, either UDP or TCP.


The tun or tap device. For the difference between these, see Section 24.1.1, “Terminology”.


The following lines contain the relative or absolute path to the root server CA certificate (ca), the root CA key (cert), and the private server key (key). These were generated in Section 24.3.1, “Creating Certificates”.


Require that peer certificates have been signed with an explicit key usage and extended key usage based on RFC3280 TLS rules.


The Diffie-Hellman parameters. Create the required file with the following command:

openssl dhparam -out /etc/openvpn/dh2048.pem 2048


Supplies a VPN subnet. The server can be reached by


Records a mapping of clients and its virtual IP address in the given file. Useful when the server goes down and (after the restart) the clients get their previously assigned IP address.


For security reasons, run the OpenVPN daemon with reduced privileges. To do so, specify that it should use the group and user nobody.


Several configuration options—see the comment in the example configuration file: /usr/share/doc/packages/openvpn/sample-config-files.


Enable this option to write short status updates with statistical data (operational status dump) to the named file. By default, this is not enabled.

All output is written to the system journal, which can be displayed with journalctl. If you have more than one configuration file (for example, one for home and another for work), it is recommended to include the device name into the file name. This avoids overwriting output files accidentally. In this case, it is tun0, taken from the dev directive—see 3.


By default, log messages go to syslog. Overwrite this behavior by removing the hash character. In that case, all messages go to /var/log/openvpn-server.log. Do not forget to configure a logrotate service. See man 8 logrotate for further details.

After having completed this configuration, you can see log messages of your OpenVPN server under /var/log/openvpn.log. After having started it for the first time, it should finish with:

... Initialization Sequence Completed

If you do not see this message, check the log carefully for any hints of what is wrong in your configuration file.

24.3.3 Configuring the VPN Clients

As the basis of your configuration file, copy /usr/share/doc/packages/openvpn/sample-config-files/client.conf to /etc/openvpn/. Then customize it to your needs.

Example 24.2: VPN Client Configuration File
# /etc/openvpn/client.conf
client 1
dev tun 2
proto udp 3
remote IP_OR_HOST_NAME 1194 4
resolv-retry infinite

remote-cert-tls server 5

# Privileges 6
user nobody
group nobody

# Try to preserve some state across restarts.

# Security 7
pkcs12 client1.p12

comp-lzo 8


Specifies that this machine is a client.


The network device. Both clients and server must use the same device.


The protocol. Use the same settings as on the server.


This is security option for clients which ensures that the host they connect to is a designated server.


Replace the placeholder IP_OR_HOST_NAME with the respective host name or IP address of your VPN server. After the host name, the port of the server is given. You can have multiple lines of remote entries pointing to different VPN servers. This is useful for load balancing between different VPN servers.


For security reasons, run the OpenVPN daemon with reduced privileges. To do so, specify that it should use the group and user nobody.


Contains the client files. For security reasons, use a separate pair of files for each client.


Turn on compression. Only use this parameter if compression is enabled on the server as well.

24.4 Setting Up a VPN Server or Client Using YaST

You can also use YaST to set up a VPN server. However, the YaST module does not support OpenVPN. Instead, it provides support for the IPsec protocol (as implemented in the software StrongSwan). Like OpenVPN, IPsec is a widely supported VPN scheme.

Procedure 24.3: Setting Up an IPsec Server
  1. To start the YaST VPN module, select Applications › VPN Gateways and Clients.

  2. Under Global Configuration, activate Enable VPN Daemon.

  3. To create a new VPN, click New VPN, then enter a name for the connection.

  4. Under Type, select Gateway (Server).

  5. Then choose the scenario:

    • The scenarios Secure communication with a pre-shared key and Secure communication with a certificate are best suited to Linux client setups.

    • The scenario Provide access to Android, iOS, Mac OS X clients sets up a configuration that is natively supported by modern versions of Android, iOS, and macOS. It is based on a pre-shared key setup with an additional user name and password authentication.

    • The scenario Provide access to Windows 7, Windows 8 clients is a configuration that is natively supported by Windows and BlackBerry devices. It is based on a certificate setup with an additional user name and password authentication.

    For this example, choose Secure communication with a pre-shared key.

  6. To specify the key, click Edit Credentials. Activate Show key, then type the secret key. Confirm with OK.

  7. Choose whether and how to limit access within your VPN under Provide VPN clients access to. To enable only certain IP ranges, specify these in CIDR format, separated by commas in Limited CIDRs. For more information about the CIDR format, see https://en.wikipedia.org/wiki/Classless_Inter-Domain_Routing.

  8. Under Clients' address pool, specify the format of IP addresses your VPN should provide to its clients.

  9. To finish, click OK. The YaST VPN module will now automatically add and enable firewall rules to allow clients to connect to the new VPN.

    To view the connection status, in the following confirmation window, click Yes. You will then see the output of systemctl status for your VPN, which allows you to check if the VPN is running and configured correctly.

24.5 For More Information

For more information about VPN in general, see:

  • https://openvpn.net: the OpenVPN home page

  • man openvpn

  • /usr/share/doc/packages/openvpn/sample-config-files/: example configuration files for different scenarios.

  • /usr/src/linux/Documentation/networking/tuntap.txt, to install the kernel-source package.