]> Tsinghua University

Beijing 100084 China cuiyong@tsinghua.edu.cn http://www.cuiyong.net/

Zhongguancun Laboratory

Beijing 100094 China +86-185-1028-7458 gaoyj@zgclab.edu.cn

Hickory Hill Consulting

7453 Hickory Hill Saline Michigan 48176 United States of America +1-734-604-0332 shares@ndzh.com

Security IETF DDoS Mitigation, BGP flow specification The BGP Flow Specification enables the distribution of traffic filter policies (traffic filters and actions) via BGP, facilitating DDoS traffic filtering. However, the traffic filter in FSv1 and FSv2 predominantly focuses on IP header fields, which may not adequately address volumetric DDoS attack traffic characterized by fixed patterns within the packet content. This document introduces a new flow specification filter type designed for packet content filtering. The match field includes ptype, otype, offset, content-length, content, and mask encoded in the Flowspec NLRI. This new filter aims to leverage network devices such as routers and switches to defend against simple volumetric DDoS attacks, reducing the overall defence cost of carrier network.

Introduction BGP flow specification describes the distribution of traffic filter policies through BGP, allowing for efficient traffic management and DDoS attack mitigation. Existing versions, FSv1 and FSv2, primarily offer n-tuple matching conditions for policy enforcement, enabling actions such as packet dropping, re-directing, or limitation, etc. These filter rules can be propagated to all BGP peers simultaneously without necessitating router configuration changes. Despite their utility, existing filters reliance on IP header fields for traffic filtering is increasingly inadequate for new kinds of DDoS attack. DDoS attacks such as ACK flood, UDP flood, ICMP flood and HTTP flood attacks, etc, have been found to potentially characterise a fixed content in the packet. This document delineates a new flow specification filter type that facilitates packet content filtering, leveraging ptype, otype, offset, content-length, content, and mask fields within the FlowSpec NLRI to enhance DDoS defense mechanisms.

Terminology 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 BCP 14 when, and only when, they appear in all capitals, as shown here.

Definitions and Acronyms

• DDoS: Distributed Denial of Service
• NLRI: Network Layer Reachability Information
• FSv1: Flow Specification Version 1, defined in and
• FSv2: Flow Specification Version 2 define in [draft-hares-idr-flowspec-v2-ddos-00]

The Packet Content Filter for FSv1 This document specifies a new flow specification filter type that is encoded in the BGP FS NLRI and we follow the FSv1 definition format. The packet content filter is defined as follows: Type TBD – Packet-Content Encoding:< type (1 octet), value> The value field is encoded using ptype, otype, offset, content-length, content and mask. Encoding: < ptype (4 bits), otype (4 bits), offset (2 octets), content-length (1 octets), content (Variable), mask (Variable)>

Ptype Field The ptype is defined as a 4-bit unsigned integer that defines the packet type via AFI, because some filters are added to hardware that are IPv4 or IPv6 specific. Figure 1： Ptype field

Otype and Offset Fields The otype and offset fields define the starting position of the packet content used for matching. To avoid the effect of variable header length on the offset, we use the hierarchical way like [draft-khare-idr-bgp-flowspec-payload-match-08]. The Otype is defined as a 4-bit unsigned integer. The detail are as follows: Figure 2： Otype field Otype 0 is defined as the start of the IP header. Otype 1 is defined as the start of the data portion of the IP header after the IP options. Otype 2 is defined as the start of the TCP or UDP data. Type 2 will only be used if it is the Layer 4 transport protocol is TCP (6) or UDP (17). For other IP protocols, type 1 or type 2 can be used. The offset is defined as a 2-octet unsigned integer that specifies the count of octets to be bypassed from the otype's starting position to match the packet content. It is worth noting that packet fragmentation will cause the offset value to change, so it is not enough to filter the fragmented packets through the packet content filter. One possible way is to filter the first packet through the payload filter, and then use its header information along with the fragment filter to filter the subsequent packets. Example:

• By setting otype 0 and an offset of 0, the match is configured to start precisely at the beginning of the IP header.
• By setting otype 1 and an offset of 2, the match will start two octets past the initial data portion of the IP header, skipping over any IP options. This configuration, for example, in UDP packets, specifically targets the start of the destination port information.
• By setting otype 2 and an offset of 10, the match will start ten octets into the content of the TCP/UDP packet.

Content-length, Content and Mask Fields The content-length is a one octet unsigned integer field that contains the length of the value field in octets. Content and mask fields have a same length which defined by the content-length. The content provides a string to be matched. Based on the information provided by equipment vendors and operators, 8 octets is usually sufficient for the identification of DDoS attacks. Mask is a string containing 0 and 1, where 1 represents what will be matched and 0 represents characters that can be ignored. The content and mask are operated AND by bit to get the final content of the packet that needs to be matched.

Example of Encoding An example of a FlowSpec NLRI encoding for:"all packets to 192.0.2.0/24 and have fixed content 5858 in the tcp payload with offset of 0".

length	destination	packet content
0x0b	01 18 c0 00 02	TBD 40 12 00 00 02 58 58 ff ff

Description of each field of the FlowSpec NLRI.

Value	Description
0x0f	length	15 octets (if len<240, 1 octet)
0x01	type	Type 1 - Destination Prefix
0x18	length	24 bits
0xc0	prefix	192
0x00	prefix	0
0x02	prefix	2
TBD	type	Type TBD - Packet Content
0x40	length	64 bits
0x12	ptype, otype	IPv4, TCP payload
0x0000	offset	0 octets
0x02	content-length	2 octets
0x5858	content	0x5858
0xffff	mask	0xffff

The Packet Content Filter for FSv2

Filter Encoding To adapt to the updates of flowSpec, this document also gives the definition of the packet content filters in FSv2. The formate is as same as the NLRI formate for extented IP filters, as shown in Figure 3: Figure 3： NLRI formate for extented IP filters The formate of Sub-TLVs is showns in Figure 4: Figure 4: formate for Sub-TLV The defination of packet content filter in sub-TLV formate are as follows: Figure 5: Defination of the packet content filter where fileds have the same length as FSv1. Encoding: < ptype (4 bits), otype (4 bits), offset (2 octets), content-length (1 octets), content (Variable), mask (Variable)>

Filter Ordering Rule Compared to FSv1, FSv2 adds filter ordering function. According to the defination of ordering rules in FSv2, the transmission of SubTLVs within a flow specification rule MUST be sent ascending order by SubTLV type. If the SubTLV types are the same, then the value fields are compared using mechanisms defined in and and MUST be in ascending order. However, due to multiple fields in the value of the packet content filter, the mechanisms defined in and do not apply. To give the default ordering rules of packet content filters, this document gives the defination as follows:

1. Filters with a larger content-length are implemented first.
2. If they have the same content-length, compare otype and the larger type is implemented first.
3. If they have the same content-lengte and otype, compare offset and the larger value will be implemented first.
4. If they have the same content-length, otype, and offset, compare the content as a binary string using the memcmp() function as defined by [ISO_IEC_9899]. The common prefix is compared. If the common prefix is not equal, the string with the lower prefix has higher precedence.

When there are multiple component exist in multiple NLRI, determine the order of the packet content filters according to the above ordering rules.

Use Cases Here is a use case for ordering rules with multiple NLRI and multiple components. There are five components, with the same destination IP and user order, each of which contains a packet content filter with different values: The rules will be installed as:

IANA Considerations In accordance with the procedures outlined for managing the "Flow Spec Component Types" registry, IANA is hereby requested to assign a new Type Value. This assignment is sought from the First Come First Served range, as detailed below: The introduction of the "Packet Content filter" Type Value is purposed to expand the capability of BGP FSv1 by enabling more granular control over traffic filtering. This is especially pertinent for addressing complex patterns within packet content that are characteristic of Distributed Denial of Service (DDoS) attacks and other security challenges. The proposed Packet Content filter facilitate the specification of detailed criteria for traffic matching, including but not limited to, content inspection at specific packet offsets. In the following update we will add the definition of FSv2.

Security Considerations No new security issues are introduced to the BGP protocol by this specification

Normative References This document defines a Border Gateway Protocol Network Layer Reachability Information (BGP NLRI) encoding format that can be used to distribute (intra-domain and inter-domain) traffic Flow Specifications for IPv4 unicast and IPv4 BGP/MPLS VPN services. This allows the routing system to propagate information regarding more specific components of the traffic aggregate defined by an IP destination prefix. It also specifies BGP Extended Community encoding formats, which can be used to propagate Traffic Filtering Actions along with the Flow Specification NLRI. Those Traffic Filtering Actions encode actions a routing system can take if the packet matches the Flow Specification. This document obsoletes both RFC 5575 and RFC 7674. "Dissemination of Flow Specification Rules" (RFC 8955) provides a Border Gateway Protocol (BGP) extension for the propagation of traffic flow information for the purpose of rate limiting or filtering IPv4 protocol data packets. This document extends RFC 8955 with IPv6 functionality. It also updates RFC 8955 by changing the IANA Flow Spec Component Types registry. In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings.

Acknowledgements We wish to thank Jeffery Hass and Li Yang for their valuable comments and suggestions on this document. We also wish to thank Rui Xu and Yannan Hu for their contribution in the implementation and validation of the packet content filter software.