TURN Server Auto Discovery

TURN is a protocol that is often used to improve the connectivity of Peer-to-Peer (P2P) applications (as defined in section 2.7 of ). TURN allows a connection to be established when one or both sides are incapable of a direct P2P connection. It is an important building block for interactive, real-time communication using audio, video, collaboration etc. While TURN services are extensively used today, the means to auto discover TURN servers do not exist. TURN clients are usually explicitly configured with a well known TURN server. To allow TURN applications to operate seamlessly across different types of networks and encourage the use of TURN without the need for manual configuration, it is important that there exists an auto discovery mechanism for TURN services. Web Real-Time Communication (WebRTC) usages and related extensions, which are mostly based on web applications, need TURN server discovery mechanisms. This document describes three discovery mechanisms, so as to maximize opportunity for discovery, based on the network in which the TURN client finds itself. The three discovery mechanisms are: A resolution mechanism based on straightforward Naming Authority Pointer (S-NAPTR) resource records in the Domain Name System (DNS). describes details on retrieving a list of server transport addresses from DNS that can be used to create a TURN allocation. DNS Service Discovery A mechanism based on anycast address for TURN. In general, if a client wishes to communicate using one of its interfaces using a specific IP address family, it SHOULD query the TURN server(s) that has been discovered for that specific interface and address family. How to select an interface and IP address family is out of the scope of this document.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in .

TURN clients, by default, discover TURN server(s) by means of local or manual TURN configuration (i.e., TURN servers configured at the system level). Configuration discovered from an application, e.g., a Java Script specified TURN server for Web Real-Time Communication (WebRTC) usages and related extensions, is considered as local configuration. An implementation may give the user an opportunity (e.g., by means of configuration file options or menu items) to specify a TURN server for each address family. A client can choose auto-discovery in the absence of local configuration, if local configuration doesn’t work or in addition to local configuration. This document does not offer a recommendation on server selection. A TURN client that implements the auto discovery algorithm, to discover TURN servers in the attached network, uses the following mechanisms for discovery: Service Resolution : The TURN client attempts to perform TURN service resolution using the host's DNS domain. DNS SD: DNS Service Discovery. Anycast : Send TURN allocate request to the assigned TURN anycast request for each combination of interface and address family. Not all TURN servers may be discovered using NAPTR records or DNS SD; Similarly, not all TURN servers may support anycast. For best results, a client SHOULD implement all discovery mechanisms described above. The document does not prescribe a strict order that a client must follow for discovery. An implementation may choose to perform all the above steps in parallel for discovery OR choose to follow any desired order and stop the discovery procedure if a mechanism succeeds. On hosts with more than one interface or address family (IPv4/v6), the TURN server discovery procedure has to be performed for each combination of interface and address family. A client MAY choose to perform the discovery procedure only for a desired interface/address combination if the client does not wish to discover a TURN server for all combinations of interface and address family.

This mechanism is performed in two steps: 1. A DNS domain name is retrieved for each combination of interface and address family. 2. Retrieved DNS domain names are then used for S-NAPTR lookups as per . Further DNS lookups may be necessary to determine TURN server IP address(es).

A client has to determine the domain in which it is located. The following sections provide two possible mechanisms to learn the domain name, but other means of retrieving domain names may be used, which are outside the scope of this document e.g. local configuration. Implementations may allow the user to specify a default name that is used if no specific name has been configured.

DHCP can be used to determine the domain name related to an interface's point of network attachment. Network operators may provide the domain name to be used for service discovery within an access network using DHCP. Sections 3.2 and 3.3 of define DHCP IPv4 and IPv6 access network domain name options, OPTION_V4_ACCESS_DOMAIN and OPTION_V6_ACCESS_DOMAIN respectively, to identify a domain name that is suitable for service discovery within the access network. For IPv4, the discovery procedure MUST request the access network domain name option in a Parameter Request List option, as described in . defines the DHCP IPv4 domain name option; while this option is less suitable, a client MAY request for it if the access network domain name defined in is not available. For IPv6, the discovery procedure MUST request for the access network domain name option in an Options Request Option (ORO) within an Information-request message, as described in . If neither option can be retrieved the procedure fails for this interface. If a result can be retrieved it will be used as an input for S-NAPTR resolution.

For a TURN client with an understanding of the protocol mechanics of calling applications, the client may wish to extract the domain name from its own identity i.e canonical identifier used to reach the user.

A client may support multiple users, potentially with different domains, or for a single user to use different domains for different services. The means to choose and extract the domain name may be different based on the type of identifier, service being used etc., which are outside the scope of this document.

Once the TURN discovery procedure has retrieved domain names, the resolution mechanism described in is followed. An S-NAPTR lookup with 'RELAY' application service and the desired protocol tag is made to obtain information necessary to connect to the authoritative TURN server within the given domain. If no TURN-specific S-NAPTR records can be retrieved, the discovery procedure fails for this domain name (and the corresponding interface and IP protocol version). If more domain names are known, the discovery procedure may perform the corresponding S-NAPTR lookups immediately. However, before retrying a lookup that has failed, a client must wait a time period that is appropriate for the encountered error (NXDOMAIN, timeout, etc.).

DNS-based Service Discovery (DNS-SD) and Multicast DNS (mDNS) provide generic solutions for discovering services available in a local network. DNS-SD/mDNS define a set of naming rules for certain DNS record types that they use for advertising and discovering services. Section 4.1 of specifies that a service instance name in DNS-SD has the following structure: <Instance> . <Service> . <Domain> The <Domain> portion specifies the DNS sub-domain where the service instance is registered. It may be "local.", indicating the mDNS local domain, or it may be a conventional domain name such as "example.com.". The <Service> portion of the TURN service instance name MUST be "_turn._udp" or "_turn._tcp" or "_turns._udp" or "_turns._tcp", as introduced in .

A TURN client can proactively discover TURN servers being advertised in the site by multicasting a PTR query to one or all of the following: "_turn._udp.local." "_turn._tcp.local" "_turns._udp.local." "_turns._tcp.local" A TURN server can send out gratuitous multicast DNS answer packets whenever it starts up, wakes from sleep, or detects a change in network configuration. TURN clients receive these gratuitous packets and cache information contained in it.

IP anycast can also be used for TURN service discovery. A packet sent to an anycast address is delivered to the "topologically nearest" network interface with the anycast address. Using the TURN anycast address, the only two things that need to be deployed in the network for discovery are the two things that actually use TURN. When a client requires TURN services, it sends a TURN allocate request to the assigned anycast address. The TURN anycast server responds with a 300 (Try Alternate) error as described in ; The response contains the TURN unicast address in the ALTERNATE-SERVER attribute. For subsequent communication with the TURN server, the client uses the responding server's unicast address. This has to be done because two packets addressed to an anycast address may reach two different anycast servers. The client, thus, also needs to ensure that the initial request fits in a single packet. An implementation may choose to send out every new TURN Allocation request to the anycast address to discover the closest and the most optimal unicast address for the TURN server.

A change of IP address on an interface may invalidate the result of the TURN server discovery procedure. For instance, if the IP address assigned to a mobile host changes due to host mobility, it may be required to re-run the TURN server discovery procedure without relying on earlier gained information. New requests should be made to the newly learned TURN servers learned after TURN discovery re-run. However, if an earlier learned TURN server is still accessible using the new IP address, procedures described for mobility using TURN defined in can be used for ongoing streams.

WebRTC endpoints SHOULD treat any TURN server discovered through the mechanisms described in this specification as an enterprise/gateway or access network server, in accordance with Recursively Encapsulated TURN .

IANA has assigned a single IPv4 address from the 192.0.0.0/24 prefix and registered it in the "IANA IPv4 Special-Purpose Address Registry" .

IANA has assigned a single IPv6 address from the 2001:0000::/23 prefix and registered it in the "IANA IPv6 Special-Purpose Address Registry" .

Use of Session Traversal Utilities for NAT (STUN) authentication is OPTIONAL for TURN servers provided by the local network or by the access network. A network provided TURN server MAY be configured to accept Allocation requests without STUN authentication, and a TURN client MAY be configured to accept Allocation success responses without STUN authentication from a network provided TURN server. Making STUN authentication OPTIONAL is a downgrade of a MUST level requirement defined in . The downgrade allows TURN servers provided by local or access network to accept Allocation requests from new and/or guest users in the network who do not necessarily possess long term credentials for STUN authentication. The intention, in such deployments, being to provide TURN services to all users in the local or access network. However, this opens up a TURN server to a variety of attacks described in Section 17 of . A TURN server in such cases must be configured to only process STUN requests from the trusted local network or subscribers of the access network. Operational measures must be taken in order protect the TURN server; some of these measures include, but not limited to, access control by means of access-lists, firewalls, subscriber quota limits, ingress filtering etc. A TURN client MUST use (D)TLS in the absence of STUN long-term credential mechanism or STUN Extension for Third-Party Authorization . It is RECOMMENDED that the TURN client use one of the following techniques with (D)TLS to validate the TURN server: For certificate-based authentication, a pre-populated trust anchor store allows a TURN client to perform path validation for the server certificate obtained during the (D)TLS handshake. If the client used a domain name to discover the TURN server, that domain name also provides a mechanism for validation of the TURN server. The client MUST use the rules and guidelines given in section 6 of to validate the TURN server identity. Certification authorities that issue TURN server certificates SHOULD support the CN-ID, DNS-ID, SRV-ID and URI-ID identifier types. TURN service providers SHOULD prefer the use of DNS-ID, SRV-ID and URI-ID over CN-ID identifier types in certificate requests (as described in Section 2.3 from ) and the wildcard character '*' SHOULD NOT be included in presented identifier. For TURN servers that don't have a certificate trust chain (e.g., because they are on a home network or a corporate network), a configured list of TURN servers can contain the Subject Public Key Info (SPKI) fingerprint of the TURN servers. The public key is used for the same reasons HTTP pinning uses the public key. Raw public key-based authentication, as defined in , could also be used to authenticate a TURN server. An auto-discovered TURN server is considered to be only as trusted as the path between the client and the TURN server. In order to safely use auto-discovered TURN servers for sessions with 'strict privacy' requirements, the user needs to be able to define privacy criteria (e.g. a trusted list of servers, networks, or domains) that are considered acceptable for such traffic. Any discovered TURN server outside the criteria is considered untrusted and therefore MUST NOT be used for privacy sensitive communication. In some auto-discovery scenarios, it might not be possible for the TURN client to use (D)TLS authentication to validate the TURN server. However, fall-back to clear text in such cases could leave the TURN client open to on-path injection of spoofed TURN messages. Instead, with an explicit administrator choice, a TURN client SHOULD fall-back to encryption-only (D)TLS when (D)TLS authentication is not available. Another reason to fall-back to encryption-only is for privacy, which is analogous to SMTP opportunistic encryption where one does not require privacy but one desires privacy when possible. It is suggested that a TURN client attempt (D)TLS with authentication and encryption, falling back to encryption-only if the TURN server cannot be authenticated via (D)TLS. A TURN client could fall back to clear text, with an explicit administrator choice, if it does not support unauthenticated (D)TLS, but fallback to clear text is NOT RECOMMENDED because it makes the client more susceptible to man-in-the-middle attacks and on-path packet injection.

The primary attack against the methods described in this document is one that would lead to impersonation of a TURN server. An attacker could attempt to compromise the S-NAPTR resolution. Security considerations described in are applicable here as well. In addition to considerations related to S-NAPTR, it is important to recognize that the output of this is entirely dependent on its input. An attacker who can control the domain name can also control the final result. Because more than one method can be used to determine the domain name, a host implementation needs to consider attacks against each of the methods that are used. If DHCP is used, the integrity of DHCP options is limited by the security of the channel over which they are provided. Physical security and separation of DHCP messages from other packets are commonplace methods that can reduce the possibility of attack within an access network; alternatively, DHCP authentication can provide a degree of protection against modification. When using DHCP discovery, clients are encouraged to use unicast DHCP INFORM queries instead of broadcast queries which are more easily spoofed in insecure networks.

Since DNS-SD is just a specification for how to name and use records in the existing DNS system, it has no specific additional security requirements over and above those that already apply to DNS queries and DNS updates. For DNS queries, DNS Security Extensions (DNSSEC) should be used where the authenticity of information is important. For DNS updates, secure updates should generally be used to control which clients have permission to update DNS records. For mDNS, in addition to what has been described above, a principal security threat is a security threat inherent to IP multicast routing and any application that runs on it. A rogue system can advertise that it is a TURN server. Discovery of such rogue systems as TURN servers, in itself, is not a security threat if there is a means for the TURN client to authenticate and authorize the discovered TURN servers.

In a network without any TURN server that is aware of the TURN anycast address, outgoing TURN requests could leak out onto the external Internet, possibly revealing information. Using an IANA-assigned well-known TURN anycast address enables border gateways to block such outgoing packets. In the default-free zone, routers should be configured to drop such packets. Such configuration can occur naturally via BGP messages advertising that no route exists to said address. Sensitive clients that do not wish to leak information about their presence can set an IP TTL on their TURN requests that limits how far they can travel into the public Internet.

The authors would like to thank Simon Perrault, Paul Kyzivat, Troy Shields, Eduardo Gueiros, Ted Hardie, Bernard Aboba, Karl Stahl, Brian Weis, Ralph Dromes, Ben Campbell, Suresh Krishnan and Brandon Williams for their review and valuable comments. Thanks to Adam Roach for his detailed review and suggesting DNS Service Discovery as an additional discovery mechanism.