Project

General

Profile

Introduction to strongSwan

This document is intended to give an introduction to strongSwan for new users (or existing users with catching-up to do).

Prerequisites

It is assumed that the reader of this document...
  • understands the principles of networking, from setting IP addresses and DNS servers to basic firewalling
  • grasps the basic concepts of authentication based on public-key cryptography and a public-key infrastructure (PKI)
  • knows how to install binary software packages or how to compile source code following instructions
  • is adept on the console

If you don't have such knowledge, there exist many ready-to-use appliances that provide remote IPsec access.

Securing a Network

strongSwan is a complete IPsec solution providing encryption and authentication to servers and clients.
It can be used to secure communications with remote networks, so that connecting remotely is the same as
connecting locally.

  • Gateway: The gateway is usually your firewall, but this can be any host within your network.
    Often the gateway is also able to serve a small network with DHCP and DNS.
    In the image above the hosts moon and sun serve as gateways for the internal hosts alice and bob,
    respectively.
  • Remote access / Roadwarrior clients: Usually, roadwarriors are laptops and other mobile devices
    connecting from remote to your network using the gateway. In the image above carol represents a
    roadwarrior who wants to access either of the two networks behind the two gateways.
  • Remote hosts / Host-to-Host: This can be a remote web server or a backup system. This is illustrated
    in the image by host winnetou and either of the gateways. The connection between the two hosts can
    usually be initiated by either one of them.
  • Remote sites / Site-to-Site: Hosts in two or more subnets at different locations should be able to access
    each other. Again referring to the image above, the two subnets 10.1.0.0/24 and 10.2.0.0/24 behind
    gateways moon and sun, respectively, might be connected, so that the hosts alice and bob may securely
    communicate with one another.

On our website you'll find dozens of complete configuration examples covering these and similar situations.

IKE and IPsec Basics

strongSwan is basically a keying daemon, which uses the Internet Key Exchange protocols (IKEv1 and IKEv2)
to establish security associations (SA) between two peers. IKE provides strong authentication of both peers and derives
unique cryptographic session keys. Such an IKE session is often denoted IKE_SA in our documentation.
Besides authentication and key material IKE also provides the means to exchange configuration information (e.g. virtual
IP addresses
) and to negotiate IPsec SAs, which are often called CHILD_SAs. IPsec SAs define which network traffic is
to be secured and how it has to be encrypted and authenticated.

A CHILD_SA consists of two components:
  1. The actual IPsec SAs (there are two, one in each direction) describing the algorithms and keys used to encrypt and authenticate the traffic
  2. The policies (there are at least two) that define which network traffic shall use such an SA

The policies work both ways, that is, only traffic matching an inbound policy will be allowed after decryption. Policies are derived from the
traffic selectors (TS) negotiated via IKE when establishing a CHILD_SA.
Unprotected traffic that the kernel receives and for which there is a matching inbound IPsec policy will be dropped. This is a security feature.

The actual IPsec traffic is not handled by strongSwan but instead by the network and IPsec stack of the operating
system kernel.

strongSwan installs the negotiated IPsec SAs and SPs into the kernel by using a platform dependent kernel API.

The mentioned distinction between policies and SAs often leads to misconceptions. For instance, referring to the image
above, if host moon has a site-to-site tunnel to host sun (connecting the two networks 10.1.0.0/24 and 10.2.0.0/24),
and host carol has a roadwarrior connection to host sun (from which carol received a virtual IP address of 10.3.0.10),
then carol wont be able to automatically communicate with alice, even if forwarding is enabled on sun. This is because
there is no IPsec policy allowing traffic between carol (10.3.0.10) and alice (10.1.0.10). An additional SA between moon
and sun, connecting the virtual subnet 10.3.0.0/24 with 10.1.0.0/24 would be a possible solution to this issue.

Generally, IPsec processing and routing are two different topics. IPsec is often just bumped into the stack (policy based)
and the original routing decision for the unprotected packet also applies to the protected packet.
This is the reason charon, by default, installs specific routes to the remote part of the TS (in newer versions of charon (>5.5.0),
routes aren't installed for transport mode CHILD_SAs).
An exception to this is route based IPsec, which uses interfaces to control what packets are going to be processed
by each tunnel to a unique participant. Route based IPsec is less flexible than policy based IPsec.

More information about IPsec and IKE can be found on our wiki.

Authentication Basics

To ensure that the peer with which an IKE_SA is established is really who it claims to be it has to be authenticated.

strongSwan provides several methods to do this:

  • Public Key Authentication: This uses RSA, ECDSA or EdDSA X.509 certificates to verify the authenticity of the peer.
    • Certificates can be self-signed (in which case they have to be installed on all peers), or signed by a common
      Certificate Authority (CA). The latter simplifies deployment and configuration a lot as the gateway only
      needs the CA certificate to authenticate all peers that provide a valid certificate signed by that CA.
    • Certificate Revocation Lists (CRLs) or the Online Certificate Status Protocol (OCSP) may be
      used to verify the validity of certificates.
    • To securely store private keys smart cards may be used via the PKCS#11 plugin.
    • In order to prevent man-in-the-middle attacks the identity claimed by the peer has to be confirmed by
      the certificate, either by the subject or a subjectAltName extension.
  • Pre-Shared-Key (PSK): A pre-shared-key is an easy to deploy option but it requires strong secrets to be secure.
    • If the PSK is known to many users (which is often the case with IKEv1 XAuth with PSK) any user who knows
      the secret could impersonate the gateway. Therefore this method is not recommended for large scale
      deployments.
  • Extensible Authentication Protocol (EAP): This covers several possible authentication methods, some are
    based on username/password authentication (EAP-MD5, EAP-MSCHAPv2, EAP-GTC) or on certificates (EAP-TLS),
    some can even tunnel other EAP methods (EAP-TTLS, EAP-PEAP).
    • The actual authentication of users may be delegated to a RADIUS server with the eap-radius plugin.
    • EAP authentication can only be used with IKEv2 and for some methods with IKEv1 using the xauth-eap plugin.
  • eXtended Authentication (XAuth): XAuth provides a flexible authentication framework within IKEv1. It is mainly
    used for username/password based authentication. Also, it is generally used as a second authentication method
    after a mutual authentication with either certificates or PSK. With IKEv1 hybrid authentication is is, however,
    possible to authenticate the gateway with a certificate and use only XAuth to authenticate the client.

With IKEv2 it is possible to use multiple authentication rounds, for instance, to first authenticate the "machines" with
certificates and then the "user" with an username/password-based authentication scheme (e.g EAP-MSCHAPv2). It is also
possible to use asymmetric authentication, for instance, by authenticating the gateway with a certificate and the client
with a username/password-based EAP method (in the first authentication round). Please be aware that not all IKEv2
implementations support this extension (RFC 4739).

Again, our website provides dozens of configuration examples covering these and other authentication options.

Configuration Files

The recommended way of configuring strongSwan is via the powerful vici interface and the swanctl command line tool. The swanctl.conf configuration file used by swanctl is stored together with certificates and corresponding private keys in the swanctl directory. Global strongSwan settings as well as plugin-specific configurations are defined in strongswan.conf.

Alternatively, the legacy ipsec stroke interface and its ipsec.conf and ipsec.secrets configuration files may be used.

Other Configuration Sources

The configuration may also be loaded from an SQL database or provided by custom plugins like the one used with the NetworkManager plugin.

Installation

The installation of strongSwan is covered in a separate part of the wiki.

Using binary packages provided by your distribution is generally recommended as it makes maintenance easier.
Unfortunately, this means that you are often not able to use the most recent version.

Invocation and Maintenance

strongSwan is usually managed with the swanctl command, while the IKE daemon charon is controlled by systemd on modern distros.
With legacy installations, strongSwan is controlled by the ipsec command, where ipsec start will start the starter daemon which in turn
starts and configures the keying daemon charon.

Connections and CHILD_SAs defined in swanctl.conf (or in conn sections in ipsec.conf) can be started on three different occasions:

  • On traffic: If start_action=trap/auto=route is used, IPsec trap policies for the configured traffic (local|remote_ts/left|rightsubnet)
    will be installed and traffic matching these policies will trigger acquire events that cause the daemon to establish the required IKE/IPsec SAs.
    This is also used for passthrough/drop IPsec policies, to let specific traffic bypass other policies/SAs, or drop it completely.
  • On startup: CHILD_SAs configured with start_action=start (or auto=start) will automatically be established when the daemon is started.
    They are not automatically restarted when they go down for some reason. You need to specify other configuration settings
    (dpd_action/dpdaction and/or close_action/closeaction) to restart them automatically, but even then, the setup is not bullet-proof
    and will potentially leak packets. You are encouraged to use trap policies and read the SecurityRecommendations to take care of any problems.
  • Manually: A connection that uses no start_action (or auto=add in ipsec.conf) has to be established manually with swanctl --initiate (or ipsec up) or by a peer/roadwarrior.
    Depending on the configuration, it is also possible to use swanctl --install (or ipsec route) to install policies manually for such connections, like start_action=trap/auto=route would do it on startup.

After an SA has been established swanctl --terminate (or ipsec down) may be used to tear down the IKE_SA or individual CHILD_SAs.

Whenever the swanctl.conf file or credentials the swanctl directory are changed they may be reloaded with the different --load* commands (with ipsec.conf the ipsec update or ipsec reload commands may be used to reload the config). Already established
connections are not affected by these commands (unless start_action=start is used), if that is required the SAs or even the daemon must be restarted.

For legacy configs, if ipsec.secrets or the files in ipsec.d have been changed the ipsec reread... commands may be used to reload these files.
End-entity certificates placed in ipsec.d/certs are not loaded automatically, instead they are loaded whenever referenced
with left|rightcert in a conn section. Using the ipsec purge... commands may be required in order for the new files to be used (not required when using swanctl to load credentials).

Using the different swanctl --list* (ipsec list...) commands will provide information about loaded or cached certificates, supported algorithms and loaded plugins.

Logging and Monitoring

If you run into problems, increasing the log level might help you understand what exactly went wrong. The different
logging options are described on our wiki and the strongswan.conf man page (recommended log settings for
debugging problems may be found on HelpRequests).

Whenever you encounter a log message similar to "received ... error notify", where ... is, for instance,
NO_PROPOSAL_CHOSEN or TS_UNACCEPTABLE, you should consult the logs of the remote peer so as to find out why
it generated that error notify in the first place.

Starting the daemon with ipsec start --nofork) prevents it from forking and will log directly to the console (in case
loggers are configured in strongswan.conf make sure one of them logs to stderr or stdout).

The swanctl --list* commands (or ipsec status and statusall) will provide information about the established and configured connections.

On Linux, the iproute2 package provides the ip xfrm state and ip xfrm policy commands to request detailed
information about the IPsec SAs and policies installed in the kernel. Adding the -s option will display extensive
statistical information like the number of transmitted or invalid packages. On other platforms the setkey command
from the ipsec-tools package provides similar information.

tcpdump and wireshark are also often useful to debug problems.

When testing a connection with ping make sure to select a source IP address (with the -I option) that is included in the
local traffic selector (also see Site-to-Site Configurations below).

PKI

To use certificate based authentication you'll need to create either self-signed certificates or setup a whole public-key
infrastructure (PKI), consisting of a Certificate Authority (CA), optional intermediate CAs and end-entity certificates plus
certificate revocation lists (CRL) or other methods like OCSP to verify the validity of certificates.

One of the easiest ways to generate certificates is to use the ipsec pki utility. Since setting up a whole PKI can be quite
complex, we only provide instructions to get you started.

OpenSSL is also a widespread alternative to generate certificates, as are several GUI based CA management utilities.
Commercial CA management tools like Microsoft's are also often used for large scale CAs.

Certificate requirements

  1. The generated leaf (end) certificates need to authenticate the corresponding peer's IKE ID for authentication to succeed.
  2. To authenticate against another instance of strongSwan using a or several certificates (if attribute certificates are used), the certificate must authenticate the ID that the host sends (so Alice tries to authenticate against Bob as Alice (herself), so Alice's certificate has to have at least a SAN field with the correct type (FQDN) and value "Alice" or the DN (not CN!) has to be "Alice". In other words, you can use the complete DN or any of the SAN fields' values (assuming the type is correct. See Notes regarding certificates for details) as an ID.
  3. Additionally, the certificate has to be trusted by Bob by, for example, it being known beforehand as valid to Bob or the certificate being authenticated/issued by a certificate authority that Bob trusts.
  4. For authentication to succeed, the other peer has to possess the complete chain of the certificate (from root certificate (the root CA) to the leaf certificate (the host's certificate) with all intermediate certificates (all intermediate CAs). This is done by sending any intermediate certificates to the remote host.

The requirements for certificate authentication with other IKE implementations (implementations of other vendors, like Apple or Microsoft/Windows) are mentioned in the respective articles in the interoperability section of the main page.

Notes regarding certificates

Because a foreign root certificate is not automatically trusted, a certificate's root certificate is not sent to another host. That other host already has to have that root certificate available in the first place for authentication to succeed.

A local certificate is only sent to another host if at least one of the following is true:
  1. You set leftsendcert=always in the conn that is used
  2. The remote peer requested a certificate issued by a trusted CA by sending a CERTREQ payload to the local peer that indicates one of the CAs in the path from the local host's certificate to its root certificate.

Routing

On Linux, strongSwan installs routes into routing table 220 by default and hence requires the kernel to support policy
based routing.

You can make the daemon install the routes into any table you like, or you can disable it completely.
For those purposes, the charon.install_routes, charon.routing_table and charon.routing_table_prio settings
in strongswan.conf may be used. When a tunnel is established between two subnets, charon tries to find local IPs in the tunneled
local subnets. Such an IP must be configurd with scope global to be viable for the lookup. If a valid IP is found, charon will install
a route pointing to the remote subnet where the source IP is set to the found IP. This results in routes like the following:

10.1.0.0/24 via 10.2.0.1 src 10.2.0.2

In that example, the local IP would be 10.2.0.2. The remote subnet would be 10.1.0.0/24.
This is done so packets to the remote subnet are sent with the correct source IP, so the IPsec policies match and traffic
from the local machine to the remote subnet will be secured with IPsec.

To avoid conflicts with these routes (especially if virtual IPs are used), the kernel-netlink plugin manually parses the
host's routing tables to determine a suitable source address when sending IKE packets. On hosts with a (very) high number
of routes this is quite inefficient. In that case, setting charon.plugins.kernel-netlink.fwmark in strongswan.conf is
recommended as it will allow using a more efficient source address lookup.

In order to detect connectivity changes strongSwan parses the events that the kernel sends when a route is installed or
deleted and hence could cause high CPU load when you run it on a system that receives a lot of routes via dynamic
routing, for example. You can disable it using the charon.process_route setting in strongswan.conf.

If IPv6 is used then make sure to bypass NDP traffic if necessary.

It is possible that you encounter MSS/MTU problems when tunneling traffic. Please refer to Forwarding and Split-Tunneling
for details.

Remote Access Configurations

In this section we present example configurations for common remote access use cases. In these so called roadwarrior
scenarios mobile clients will be able to connect to a remote network.

Because these clients most likely connect from unknown IP addresses the gateway will use right=%any to literally accept
connections from anywhere. To simplify routing traffic back to the clients and because roadwarriors are often located
behind one or more NAT devices, the use of virtual IP addresses is necessary.

The virtual IPs can either be from a distinct subnet or actually from the subnet behind the gateway (by use of the
farp and optionally the dhcp plugins).

Whether roadwarriors will send all traffic to the gateway or use split-tunneling, that is, only send traffic for specific
destinations through the tunnel, is also something to consider. It is explained more detailed in Forwarding and
Split-Tunneling.

The above page also explains how traffic is forwarded to hosts behind the gateway.

IKEv2 (Windows 7/8, Linux, Android 4+, Mac OS X, iOS 8+)

The gateway configurations shown in the Windows 7 section of the wiki may be used for all IKEv2 clients. In both use
cases presented there the gateway is authenticated with a certificate, while the clients will either authenticate themselves
with certificates, or use username and password. Both configurations may be implemented on a gateway to leave it to the
clients to select an authentication method.

With the eap-radius plugin the user authentication may be delegated to a RADIUS server (e.g. an existing Active Directory DC).

Both the strongSwan VPN Client for Android 4 and newer and the strongSwan NetworkManager plugin may be
used with either of these configs.

For Linux roadwarriors that don't want or can't use the NetworkManager plugin this client config may be used.
Alternatively, the charon-cmd command line IKE client provides a simple means to establish roadwarrior connections since 5.1.0.

Our app for Mac OS X supports IKEv2 and simple EAP authentication.

With iOS 8 and Mac OS X 10.10 (Yosemite) Apple introduced support for IKEv2 in their clients. A GUI to configure such connections
is currently not provided, so it's necessary to write (or generate) custom configuration profiles.

IKEv1 (iOS, Mac OS X, Android, Windows)

The configuration presented on the iOS and Mac OS X page should work for all IKEv1 clients that support XAuth.

For Windows hosts before Windows 7 it is recommended to use a third-party IPsec client like Shrew instead of the
built-in IKEv1/L2TP client.

Instead of generating a private key and certificate pair for each client you may also use the same key/certificate pair for
all clients. The actual client authentication will then be based on XAuth (this is similar to hybrid authentication, but also works
for clients that don't support it, or implement it incorrectly, like some iOS versions did). Even though the private key/certificate
pair is "public", this still ensures proper authentication of the gateway, but might simplify deployment to clients.

XAuth with PSK may also be used (see #218) but is not recommended for larger deployments.

The XAuth credentials provided by the clients may be verified against the same RADIUS server used for IKEv2 clients with
the help of the xauth-eap plugin.

Site-to-Site Configurations

For site-to-site connections you may refer to the configuration examples. Out test suites for all features can be found here.

The most important difference compared to the remote access case is that the initiator will not request a virtual IP address
but instead use leftsubnet to tunnel traffic from one or more local subnets. For IKEv2 multiple subnets (in CIDR notation) can
be added to left|rightsubnet, separated by commas. If IKEv1 is used a separate conn section has to be added for each
combination of left and right subnet as only the first subnet in left|rightsubnet will be used (using either conn %default or the
also keyword can reduce each of these configs to a few lines).

One thing that often confuses users new to IPsec is that testing a net-to-net scenario from either of the gateways often requires
one to select the source address used specifically (e.g. with ping -I) because the external IP of either gateway might not be
included in the tunneled subnets. If that is something you require either add the external IPs to the list of subnets in left|rightsubnet
or add a specific host-to-host config.

Host-to-Host Configurations

Host-to-host connections are very easy to setup. You basically have to configure right to the hostname or IP address of the peer
and configure the desired authentication, neither leftsubnet nor rightsubnet have to be set explicitly.

Again, our test suite provides several examples.