Distributed Denial of Service (DDoS) Open Threat Signaling Requirements

Distributed Denial of Service (DDoS) attacks continue to plague networks around the globe, from Tier-1 service providers on down to enterprises and small businesses. Attack scale and frequency similarly have continued to increase, in part as a result of software vulnerabilities leading to reflection and amplification attacks. Once staggering attack traffic volume is now the norm, and the impact of larger-scale attacks attract the attention of international press agencies. The greater impact of contemporary DDoS attacks has led to increased focus on coordinated attack response. Many institutions and enterprises lack the resources or expertise to operate on-premise attack mitigation solutions themselves, or simply find themselves constrained by local bandwidth limitations. To address such gaps, security service providers have begun to offer on-demand traffic scrubbing services, which aim to separate the DDoS traffic from legitimate traffic and forward only the latter. Today each such service offers its own interface for subscribers to request attack mitigation, tying subscribers to proprietary implementations while also limiting the subset of network elements capable of participating in the attack response. As a result of incompatibility across services, attack responses may be fragmentary or otherwise incomplete, leaving key players in the attack path unable to assist in the defense. The lack of a common method to coordinate a real-time response among involved actors and network domains inhibits the speed and effectiveness of DDoS attack mitigation. This document describes the required characteristics of a DOTS protocol enabling requests for DDoS attack mitigation, reducing attack impact and leading to more efficient defensive strategies. DOTS communicates the need for defensive action in anticipation of or in response to an attack, but does not dictate the form any defensive action takes. DOTS supplements calls for help with pertinent details about the detected attack, allowing entities participating in DOTS to form ad hoc, adaptive alliances against DDoS attacks as described in the DOTS use cases . The requirements in this document are derived from those use cases and .

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in . This document adopts the following terms: A distributed denial-of-service attack, in which traffic originating from multiple sources are directed at a target on a network. DDoS attacks are intended to cause a negative impact on the availability of servers, services, applications, and/or other functionality of an attack target. Denial-of-service considerations are discussed in detail in . A network connected entity with a finite set of resources, such as network bandwidth, memory or CPU, that is the focus of a DDoS attack. Potential targets include servers, services and applications. Collected behavioral characteristics defining the nature of a DDoS attack. This document makes no assumptions regarding telemetry collection methodology. An action or set of actions taken to recognize and filter out DDoS attack traffic while passing legitimate traffic to the attack target. A set of countermeasures enforced against traffic destined for the target or targets of a detected or reported DDoS attack, where countermeasure enforcement is managed by an entity in the network path between attack sources and the attack target. Mitigation methodology is out of scope for this document. An entity, typically a network element, capable of performing mitigation of a detected or reported DDoS attack. For the purposes of this document, this entity is a black box capable of mitigation, making no assumptions about availability or design of countermeasures, nor about the programmable interface between this entity and other network elements. The mitigator and DOTS server are assumed to belong to the same administrative entity. A DOTS-aware software module responsible for requesting attack response coordination with other DOTS-aware elements. A DOTS-aware software module handling and responding to messages from DOTS clients. The DOTS server SHOULD enable mitigation on behalf of the DOTS client, if requested, by communicating the DOTS client’s request to the mitigator and returning selected mitigator feedback to the requesting DOTS client. A DOTS server MAY also be a mitigator. Any DOTS-aware software module capable of participating in a DOTS siganling session. A logical DOTS agent resulting from the logical concatenation of a DOTS server and a DOTS client, analogous to a SIP Back-to-Back User Agent (B2BUA) . DOTS gateways are discussed in detail in . A bidirectional, mutually authenticated communication channel between DOTS agents characterized by resilience even in conditions leading to severe packet loss, such as a volumetric DDoS attack causing network congestion. A concise authenticated status/control message transmitted between DOTS agents, used to indicate client’s need for mitigation, as well as to convey the status of any requested mitigation. A message transmitted between DOTS agents over the signal channel, used as a keep-alive and to measure peer health. A message sent from a DOTS client to a DOTS server over the signal channel, indicating the DOTS client’s need for mitigation, as well as the scope of any requested mitigation, optionally including additional attack details to supplement server-initiated mitigation. A message sent from a DOTS server to a DOTS client over the signal channel. Note that a server signal is not a response to client signal, but a DOTS server-initiated status message sent to DOTS clients with which the server has established signaling sessions. A secure communication layer between DOTS clients and DOTS servers used for infrequent bulk exchange of data not easily or appropriately communicated through the signal channel under attack conditions. A policy matching a network traffic flow or set of flows and rate-limiting or discarding matching traffic. A filter list of addresses, prefixes and/or other identifiers indicating sources from which traffic should be blocked, regardless of traffic content. A list of addresses, prefixes and/or other identifiers from indicating sources from which traffic should always be allowed, regardless of contradictory data gleaned in a detected attack. A DOTS client exchanging messages with multiple DOTS servers, each in a separate administrative domain.

This section describes the required features and characteristics of the DOTS protocol. DOTS is an advisory protocol. An active DDoS attack against the entity controlling the DOTS client need not be present before establishing DOTS communication between DOTS agents. Indeed, establishing a relationship with peer DOTS agents during normal network conditions provides the foundation for more rapid attack response against future attacks, as all interactions setting up DOTS, including any business or service level agreements, are already complete. DOTS must at a minimum make it possible for a DOTS client to request a DOTS server’s aid in mounting a coordinated defense against a suspected attack, signaling within or between domains as requested by local operators. DOTS clients should similarly be able to withdraw aid requests. DOTS requires no justification from DOTS clients for requests for help, nor do DOTS clients need to justify withdrawing help requests: the decision is local to the DOTS clients’ domain. Regular feedback between DOTS clients and DOTS server supplement the defensive alliance by maintaining a common understanding of DOTS peer health and activity. Bidirectional communication between DOTS clients and DOTS servers is therefore critical. Yet DOTS must also work with a set of competing operational goals. On the one hand, the protocol must be resilient under extremely hostile network conditions, providing continued contact between DOTS agents even as attack traffic saturates the link. Such resiliency may be developed several ways, but characteristics such as small message size, asynchronous, redundant message delivery and minimal connection overhead (when possible given local network policy) will tend to contribute to the robustness demanded by a viable DOTS protocol. Operators of peer DOTS-enabled domains may enable quality- or class-of-service traffic tagging to increase the probability of successful DOTS signal delivery, but DOTS requires no such policies be in place. The DOTS solution indeed must be viable especially in their absence. On the other hand, DOTS must include protections ensuring message confidentiality, integrity and authenticity to keep the protocol from becoming another vector for the very attacks it’s meant to help fight off. DOTS clients must be able to authenticate DOTS servers, and vice versa, for DOTS to operate safely, meaning the DOTS agents must have a way to negotiate and agree upon the terms of protocol security. Attacks against the transport protocol should not offer a means of attack against the message confidentiality, integrity and authenticity. The DOTS server and client must also have some common method of defining the scope of any mitigation performed by the mitigator, as well as making adjustments to other commonly configurable features, such as listen ports, exchanging black- and white-lists, and so on. Finally, DOTS should provide sufficient extensibility to meet local, vendor or future needs in coordinated attack defense, although this consideration is necessarily superseded by the other operational requirements.

Extensibility: Protocols and data models developed as part of DOTS MUST be extensible in order to keep DOTS adaptable to operational and proprietary DDoS defenses. Future extensions MUST be backward compatible. Resilience and Robustness: The signaling protocol MUST be designed to maximize the probability of signal delivery even under the severely constrained network conditions imposed by particular attack traffic. The protocol MUST be resilient, that is, continue operating despite message loss and out-of-order or redundant message delivery. Bidirectionality: To support peer health detection, to maintain an open signal channel, and to increase the probability of signal delivery during attack, the signal channel MUST be bidirectional, with client and server transmitting signals to each other at regular intervals, regardless of any client request for mitigation. Unidirectional messages MUST be supported within the bidirectional signal channel to allow for unsolicited message delivery, enabling asynchronous notifications between agents. Sub-MTU Message Size: To avoid message fragmentation and the consequently decreased probability of message delivery, signaling protocol message size MUST be kept under signaling path Maximum Transmission Unit (MTU), including the byte overhead of any encapsulation, transport headers, and transport- or message-level security. Bulk Data Exchange: Infrequent bulk data exchange between DOTS agents can also significantly augment attack response coordination, permitting such tasks as population of black- or white-listed source addresses; address or prefix group aliasing; exchange of incident reports; and other hinting or configuration supplementing attack response. As the resilience requirements for the DOTS signal channel mandate small signal message size, a separate, secure data channel utilizing an established reliable transport protocol MUST be used for bulk data exchange.

Use of Common Transport Protocols: DOTS MUST operate over common widely deployed and standardized transport protocols. While the User Datagram Protocol (UDP) SHOULD be used for the signal channel, the Transmission Control Protocol (TCP) MAY be used if necessary due to network policy or middlebox capabilities or configurations. The data channel MUST use TCP; see below. Session Health Monitoring: Peer DOTS agents MUST regularly send heartbeats to each other after mutual authentication in order to keep the DOTS session open. A session MUST be considered active until a DOTS agent explicitly ends the session, or either DOTS agent fails to receive heartbeats from the other after a mutually agreed upon timeout period has elapsed. Session Redirection: In order to increase DOTS operational flexibility and scalability, DOTS servers SHOULD be able to redirect DOTS clients to another DOTS server at any time. Due to the decreased probability of DOTS server signal delivery due to link congestion, it is RECOMMENDED DOTS servers avoid redirecting while mitigation is enabled during an active attack against a target in the DOTS client’s domain. Either the DOTS servers have to fate-share the security state, the client MUST have separate security state with each potential redirectable server, or be able to negotiate new state as part of redirection. Mitigation Status: DOTS MUST provide a means to report the status of an action requested by a DOTS client. In particular, DOTS clients MUST be able to request or withdraw a request for mitigation from the DOTS server. The DOTS server MUST acknowledge a DOTS client’s request to withdraw from coordinated attack response in subsequent signals, and MUST cease mitigation activity as quickly as possible. However, a DOTS client rapidly toggling active mitigation may result in undesirable side-effects for the network path, such as route or DNS flapping. A DOTS server therefore MAY continue mitigating for a mutually negotiated period after receiving the DOTS client’s request to stop. A server MAY refuse to engage in coordinated attack response with a client. To make the status of a client’s request clear, the server MUST indicate in server signals whether client-initiated mitigation is active. When a client-initiated mitigation is active, and threat handling details such as mitigation scope and statistics are available to the server, the server SHOULD include those details in server signals sent to the client. DOTS clients SHOULD take mitigation statistics into account when deciding whether to request the DOTS server cease mitigation. Mitigation Lifetime: A DOTS client SHOULD indicate the desired lifetime of any mitigation requested from the DOTS server. As DDoS attack duration is unpredictable, the DOTS client SHOULD be able to extend mitigation lifetime with periodic renewed requests for help. When the mitigation lifetime comes to an end, the DOTS server SHOULD delay session termination for a protocol-defined grace period to allow for delivery of delayed mitigation renewals over the signal channel. After the grace period elapses, the DOTS server MAY terminate the session at any time. If a DOTS client does not include a mitigation lifetime in requests for help sent to the DOTS server, the DOTS server will use a reasonable default as defined by the protocol. As above, the DOTS client MAY extend a current mitigation request’s lifetime trivially with renewed requests for help. A DOTS client MAY also request an indefinite mitigation lifetime, enabling architectures in which the mitigator is always in the traffic path to the resources for which the DOTS client is requesting protection. DOTS servers MAY refuse such requests for any reason. The reasons themselves are not in scope. Mitigation Scope: DOTS clients MUST indicate the desired scope of any mitigation, for example by using Classless Internet Domain Routing (CIDR) , prefixes, for IPv6 prefixes, the length/prefix convention established in the Border Gateway Protocol (BGP) , SIP URIs , E.164 numbers, DNS names, or by a resource group alias agreed upon with the server through the data channel. If there is additional information available narrowing the scope of any requested attack response, such as targeted port range, protocol, or service, DOTS clients SHOULD include that information in client signals. DOTS clients MAY also include additional attack details. Such supplemental information is OPTIONAL, and DOTS servers MAY ignore it when enabling countermeasures on the mitigator. As an active attack evolves, clients MUST be able to adjust as necessary the scope of requested mitigation by refining the scope of resources requiring mitigation. Mitigation Efficacy: When a mitigation request by a DOTS client is active, DOTS clients SHOULD transmit a metric of perceived mitigation efficacy to the DOTS server, per “Automatic or Operator-Assisted CPE or PE Mitigators Request Upstream DDoS Mitigation Services” in . DOTS servers MAY use the efficacy metric to adjust countermeasures activated on a mitigator on behalf of a DOTS client. Conflict Detection and Notification: Multiple DOTS clients controlled by a single administrative entity may send conflicting mitigation requests for pool of protected resources , as a result of misconfiguration, operator error, or compromised DOTS clients. DOTS servers attempting to honor conflicting requests may flap network route or DNS information, degrading the networks attempting to participate in attack response with the DOTS clients. DOTS servers SHALL detect such conflicting requests, and SHALL notify the DOTS clients in conflict. The notification SHOULD indicate the nature and scope of the conflict, for example, the overlapping prefix range in a conflicting mitigation request. Network Address Translator Traversal: The DOTS protocol MUST operate over networks in which Network Address Translation (NAT) is deployed. As UDP is the recommended transport for DOTS, all considerations in “Middlebox Traversal Guidelines” in apply to DOTS. Regardless of transport, DOTS protocols MUST follow established best common practices (BCPs) for NAT traversal.

The data channel is intended to be used for bulk data exchanges between DOTS agents. Unlike the signal channel, which must operate nominally even when confronted with despite signal degradation due to packet loss, the data channel is not expected to be constructed to deal with attack conditions. As the primary function of the data channel is data exchange, a reliable transport is required in order for DOTS agents to detect data delivery success or failure. The data channel must be extensible. We anticipate the data channel will be used for such purposes as configuration or resource discovery. For example, a DOTS client may submit to the DOTS server a collection of prefixes it wants to refer to by alias when requesting mitigation, to which the server would respond with a success status and the new prefix group alias, or an error status and message in the event the DOTS client’s data channel request failed. The transactional nature of such data exchanges suggests a separate set of requirements for the data channel, while the potentially sensitive content sent between DOTS agents requires extra precautions to ensure data privacy and authenticity. Reliable transport: Messages sent over the data channel MUST be delivered reliably, in send order. Data privacy and integrity: Transmissions over the data channel is likely to contain operationally or privacy-sensitive information or instructions from the remote DOTS agent. Theft or modification of data channel transmissions could lead to information leaks or malicious transactions on behalf of the sending agent (see below). Consequently data sent over the data channel MUST be encrypted and authenticated using current industry best practices. DOTS servers MUST enable means to prevent leaking operationally or privacy-sensitive data. Although administrative entities participating in DOTS may detail what data may be revealed to third-party DOTS agents, such considerations are not in scope for this document. Session configuration: To help meet the general and operational requirements in this document, DOTS servers MUST provide methods for DOTS client operators to configure DOTS session behavior using the data channel. DOTS server implementations MUST have mechanisms to configure the following: Acceptable signal lossiness, as described in GEN-002. Heartbeat intervals, as described in OP-002. Maximum mitigation lifetime, as described in OP-005. Resource identifiers, as described in OP-006. DOTS server implementations MAY expose additional configurability. Additional configurability is implementation-specific. Black- and whitelist management: DOTS servers SHOULD provide methods for DOTS clients to manage black- and white-lists of traffic destined for resources belonging to a client. For example, a DOTS client should be able to create a black- or whitelist entry; retrieve a list of current entries from either list; update the content of either list; and delete entries as necessary. How the DOTS server determines client ownership of address space is not in scope.

DOTS must operate within a particularly strict security context, as an insufficiently protected signal or data channel may be subject to abuse, enabling or supplementing the very attacks DOTS purports to mitigate. Peer Mutual Authentication: DOTS agents MUST authenticate each other before a DOTS session is considered valid. The method of authentication is not specified, but should follow current industry best practices with respect to any cryptographic mechanisms to authenticate the remote peer. Message Confidentiality, Integrity and Authenticity: DOTS protocols MUST take steps to protect the confidentiality, integrity and authenticity of messages sent between client and server. While specific transport- and message-level security options are not specified, the protocols MUST follow current industry best practices for encryption and message authentication. In order for DOTS protocols to remain secure despite advancements in cryptanalysis and traffic analysis, DOTS agents MUST be able to negotiate the terms and mechanisms of protocol security, subject to the interoperability and signal message size requirements above. Message Replay Protection: In order to prevent a passive attacker from capturing and replaying old messages, DOTS protocols MUST provide a method for replay detection.

The value of DOTS is in standardizing a mechanism to permit elements, networks or domains under or under threat of DDoS attack to request aid mitigating the effects of any such attack. A well-structured DOTS data model is therefore critical to the development of a successful DOTS protocol. Structure: The data model structure for the DOTS protocol may be described by a single module, or be divided into related collections of hierarchical modules and sub-modules. If the data model structure is split across modules, those distinct modules MUST allow references to describe the overall data model’s structural dependencies. Versioning: To ensure interoperability between DOTS protocol implementations, data models MUST be versioned. The version number of the initial data model SHALL be 1. Each published change to the initial published DOTS data model SHALL increment the data model version by 1. How the protocol represents data model versions is not defined in this document. Mitigation Status Representation: The data model MUST provide the ability to represent a request for mitigation and the withdrawal of such a request. The data model MUST also support a representation of currently requested mitigation status, including failures and their causes. Mitigation Scope Representation: The data model MUST support representation of a requested mitigation’s scope. As mitigation scope may be represented in several different ways, per OP-006 above, the data model MUST be capable of flexible representation of mitigation scope. Mitigation Lifetime Representation: The data model MUST support representation of a mitigation request’s lifetime, including mitigations with no specified end time. Mitigation Efficacy Representation: The data model MUST support representation of a DOTS client’s understanding of the efficacy of a mitigation enabled through a mitigation request. TBD: how do we represent the efficacy? Relationship to Transport: The DOTS data model MUST NOT depend on the specifics of any transport to represent fields in the model.