The Generalized TTL Security Mechanism (GTSM) is designed to protect
a router's TCP/IP based control plane from CPU-utilization based
attacks. In particular, while cryptographic techniques can protect
the router-based infrastructure (e.g., BGP [RFC1771], [RFC1772]) from
a wide variety of attacks, many attacks based on CPU overload can be
prevented by the simple mechanism described in this document. Note
that the same technique protects against other scarce-resource
attacks involving a router's CPU, such as attacks against
processor-line card bandwidth.
GTSM is based on the fact that the vast majority of protocol peerings
are established between routers that are adjacent [PEERING]. Thus
most protocol peerings are either directly between connected
interfaces or at the worst case, are between loopback and loopback,
with static routes to loopbacks. Since TTL spoofing is considered
nearly impossible, a mechanism based on an expected TTL value can
provide a simple and reasonably robust defense from infrastructure
attacks based on forged protocol packets.
Finally, the GTSM mechanism is equally applicable to both TTL (IPv4)
and Hop Limit (IPv6), and from the perspective of GTSM, TTL and Hop
Limit have identical semantics. As a result, in the remainder of
this document the term "TTL" is used to refer to both TTL or Hop
Limit (as appropriate).
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 BCP 14, RFC 2119
[RFC2119].
GTSM is predicated upon the following assumptions:
(i) The vast majority of protocol peerings are between adjacent
routers [PEERING].
Gill, et al. Experimental [Page 2]
RFC 3682 Generalized TTL Security Mechanism February 2004
(ii) It is common practice for many service providers to ingress
filter (deny) packets that have the provider's loopback
addresses as the source IP address.
(iii) Use of GTSM is OPTIONAL, and can be configured on a per-peer
(group) basis.
(iv) The router supports a method of classifying traffic destined
for the route processor into interesting/control and
not-control queues.
(iv) The peer routers both implement GTSM.
This document assumes that GTSM will be manually configured between
protocol peers. That is, no automatic GTSM capability negotiation,
such as is envisioned by RFC 2842 [RFC2842] is assumed or defined.
Throughout this document, we assume that potential attackers have
evolved in both sophistication and access to the point that they can
send control traffic to a protocol session, and that this traffic
appears to be valid control traffic (i.e., has the source/destination
of configured peer routers).
We also assume that each router in the path between the attacker and
the victim protocol speaker decrements TTL properly (clearly, if
either the path or the adjacent peer is compromised, then there are
worse problems to worry about).
Since the vast majority of our peerings are between adjacent routers,
we can set the TTL on the protocol packets to 255 (the maximum
possible for IP) and then reject any protocol packets that come in
from configured peers which do NOT have an inbound TTL of 255.
GTSM can be disabled for applications such as route-servers and other
large diameter multi-hop peerings. In the event that an the attack
comes in from a compromised multi-hop peering, that peering can be
shut down (a method to reduce exposure to multi-hop attacks is
outlined below).
GTSM SHOULD NOT be enabled by default. The following process
describes the per-peer behavior:
Gill, et al. Experimental [Page 3]
RFC 3682 Generalized TTL Security Mechanism February 2004
(i) If GTSM is enabled, an implementation performs the following
procedure:
(a) For directly connected routers,
o Set the outbound TTL for the protocol connection to 255.
o For each configured protocol peer:
Update the receive path Access Control List (ACL) or
firewall to only allow protocol packets to pass onto the
Route Processor (RP) that have the correct <source,
destination, TTL> tuple. The TTL must either be 255
(for a directly connected peer), or 255-(configured-
range-of-acceptable-hops) for a multi-hop peer. We
specify a range here to achieve some robustness to
changes in topology. Any directly connected check MUST
be disabled for such peerings.
It is assumed that a receive path ACL is an ACL that is
designed to control which packets are allowed to go to
the RP. This procedure will only allow protocol packets
from adjacent router to pass onto the RP.
(b) If the inbound TTL is 255 (for a directly connected
peer), or 255-(configured-range-of-acceptable-hops) (for
multi-hop peers), the packet is NOT processed. Rather,
the packet is placed into a low priority queue, and
subsequently logged and/or silently discarded. In this
case, an ICMP message MUST NOT be generated.
(ii) If GTSM is not enabled, normal protocol behavior is followed.
When a multi-hop protocol session is required, we set the expected
TTL value to be 255-(configured-range-of-acceptable-hops). This
approach provides a qualitatively lower degree of security for the
protocol implementing GTSM (i.e., a DoS attack could theoretically be
launched by compromising some box in the path). However, GTSM will
still catch the vast majority of observed DDoS attacks against a
given protocol. Note that since the number of hops can change
rapidly in real network situations, it is considered that GTSM may
not be able to handle this scenario adequately and an implementation
MAY provide OPTIONAL support.
Gill, et al. Experimental [Page 4]
RFC 3682 Generalized TTL Security Mechanism February 2004
In general, GTSM is not used for intra-domain protocol peers or
adjacencies. The special case of iBGP peers can be protected by
filtering at the network edge for any packet that has a source
address of one of the loopback addresses used for the intra-domain
peering. In addition, the current best practice is to further
protect such peers or adjacencies with an MD5 signature [RFC2385].
The use of the TTL field to protect BGP originated with many
different people, including Paul Traina and Jon Stewart. Ryan
McDowell also suggested a similar idea. Steve Bellovin, Jay
Borkenhagen, Randy Bush, Vern Paxon, Pekka Savola, and Robert Raszuk
also provided useful feedback on earlier versions of this document.
David Ward provided insight on the generalization of the original
BGP-specific idea.
The approach described here is based on the observation that a TTL
(or Hop Limit) value of 255 is non-trivial to spoof, since as the
packet passes through routers towards the destination, the TTL is
decremented by one. As a result, when a router receives a packet, it
may not be able to determine if the packet's IP address is valid, but
it can determine how many router hops away it is (again, assuming
none of the routers in the path are compromised in such a way that
they would reset the packet's TTL).
Note, however, that while engineering a packet's TTL such that it has
a particular value when sourced from an arbitrary location is
difficult (but not impossible), engineering a TTL value of 255 from
non-directly connected locations is not possible (again, assuming
none of the directly connected neighbors are compromised, the packet
hasn't been tunneled to the decapsulator, and the intervening routers
are operating in accordance with RFC 791 [RFC791]).
Gill, et al. Experimental [Page 5]
RFC 3682 Generalized TTL Security Mechanism February 2004
An exception to the observation that a packet with TTL of 255 is
difficult to spoof occurs when a protocol packet is tunneled to a
decapsulator who then forwards the packet to a directly connected
protocol peer. In this case the decapsulator (tunnel endpoint) can
either be the penultimate hop, or the last hop itself. A related
case arises when the protocol packet is tunneled directly to the
protocol peer (the protocol peer is the decapsulator).
When the protocol packet is encapsulated in IP, it is possible to
spoof the TTL. It may also be impossible to legitimately get the
packet to the protocol peer with a TTL of 255, as in the IP in MPLS
cases described below.
Finally, note that the security of any tunneling technique depends
heavily on authentication at the tunnel endpoints, as well as how the
tunneled packets are protected in flight. Such mechanisms are,
however, beyond the scope of this memo.
Protocol packets may be tunneled over IP directly to a protocol peer,
or to a decapsulator (tunnel endpoint) that then forwards the packet
to a directly connected protocol peer (e.g., in IP-in-IP [RFC2003],
GRE [RFC2784], or various forms of IPv6-in-IPv4 [RFC2893]). These
cases are depicted below.
Peer router ---------- Tunnel endpoint router and peer
TTL=255 [tunnel] [TTL=255 at ingress]
[TTL=255 at egress]
Peer router ---------- Tunnel endpoint router ----- On-link peer
TTL=255 [tunnel] [TTL=255 at ingress] [TTL=254 at ingress]
[TTL=254 at egress]
In the first case, in which the encapsulated packet is tunneled
directly to the protocol peer, the encapsulated packet's TTL can be
set arbitrary value. In the second case, in which the encapsulated
packet is tunneled to a decapsulator (tunnel endpoint) which then
forwards it to a directly connected protocol peer, RFC 2003 specifies
the following behavior:
When encapsulating a datagram, the TTL in the inner IP header is
decremented by one if the tunneling is being done as part of
forwarding the datagram; otherwise, the inner header TTL is not
changed during encapsulation. If the resulting TTL in the inner
IP header is 0, the datagram is discarded and an ICMP Time
Gill, et al. Experimental [Page 6]
RFC 3682 Generalized TTL Security Mechanism February 2004
Exceeded message SHOULD be returned to the sender. An
encapsulator MUST NOT encapsulate a datagram with TTL = 0.
Hence the inner IP packet header's TTL, as seen by the decapsulator,
can be set to an arbitrary value (in particular, 255). As a result,
it may not be possible to deliver the protocol packet to the peer
with a TTL of 255.
Protocol packets may also be tunneled over MPLS to a protocol peer
which either the penultimate hop (when the penultimate hop popping
(PHP) is employed [RFC3032]), or one hop beyond the penultimate hop.
These cases are depicted below.
Peer router ---------- Penultimate Hop (PH) and peer
TTL=255 [tunnel] [TTL=255 at ingress]
[TTL<=254 at egress]
Peer router ---------- Penultimate Hop -------- On-link peer
TTL=255 [tunnel] [TTL=255 at ingress] [TTL <=254 at ingress]
[TTL<=254 at egress]
TTL handling for these cases is described in RFC 3032. RFC 3032
states that when the IP packet is first labeled:
... the TTL field of the label stack entry MUST BE set to the
value of the IP TTL field. (If the IP TTL field needs to be
decremented, as part of the IP processing, it is assumed that
this has already been done.)
When the label is popped:
When a label is popped, and the resulting label stack is empty,
then the value of the IP TTL field SHOULD BE replaced with the
outgoing TTL value, as defined above. In IPv4 this also requires
modification of the IP header checksum.
where the definition of "outgoing TTL" is:
The "incoming TTL" of a labeled packet is defined to be the value
of the TTL field of the top label stack entry when the packet is
received.
Gill, et al. Experimental [Page 7]
RFC 3682 Generalized TTL Security Mechanism February 2004
The "outgoing TTL" of a labeled packet is defined to be the larger
of:
a) one less than the incoming TTL,
b) zero.
In either of these cases, the minimum value by which the TTL could be
decremented would be one (the network operator prefers to hide its
infrastructure by decrementing the TTL by the minimum number of LSP
hops, one, rather than decrementing the TTL as it traverses its MPLS
domain). As a result, the maximum TTL value at egress from the MPLS
cloud is 254 (255-1), and as a result the check described in section
3 will fail.
While the GTSM method is less effective for multi-hop protocol
sessions, it does close the window on several forms of attack.
However, in the multi-hop scenario GTSM is an OPTIONAL extension.
Protection of the protocol infrastructure beyond what is provided by
the GTSM method will likely require cryptographic machinery such as
is envisioned by Secure BGP (S-BGP) [SBGP1,SBGP2], and/or other
extensions. Finally, note that in the multi-hop case described
above, we specify a range of acceptable TTLs in order to achieve some
robustness to topology changes. This robustness to topological
change comes at the cost of the loss of some robustness to different
forms of attack.
[RFC791] Postel, J., "Internet Protocol Specification", STD 5, RFC
791, September 1981.
[RFC1771] Rekhter, Y. and T. Li (Editors), "A Border Gateway
Protocol (BGP-4)", RFC 1771, March 1995.
[RFC1772] Rekhter, Y. and P. Gross, "Application of the Border
Gateway Protocol in the Internet", RFC 1772, March 1995.
[RFC2003] Perkins, C., "IP Encapsulation with IP", RFC 2003, October
1996.
Gill, et al. Experimental [Page 8]
RFC 3682 Generalized TTL Security Mechanism February 2004
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC2385] Heffernan, A., "Protection of BGP Sessions via the TCP MD5
Signature Option", RFC 2385, August 1998.
[RFC2461] Narten, T., Nordmark, E. and W. Simpson, "Neighbor
Discover for IP Version 6 (IPv6)", RFC 2461, December
1998.
[RFC2784] Farinacci, D., "Generic Routing Encapsulation (GRE)", RFC
2784, March 2000.
[RFC2842] Chandra, R. and J. Scudder, "Capabilities Advertisement
with BGP-4", RFC 2842, May 2000.
[RFC2893] Gilligan, R. and E. Nordmark, "Transition Mechanisms for
IPv6 Hosts and Routers", RFC 2893, August 2000.
[RFC3032] Rosen, E. Tappan, D., Fedorkow, G., Rekhter, Y.,
Farinacci, D., Li, T. and A. Conta, "MPLS Label Stack
Encoding", RFC 3032, January 2001.
[RFC3036] Andersson, L., Doolan, P., Feldman, N., Fredette, A. and
B. Thomas, "LDP Specification", RFC 3036, January 2001.
[RFC3392] Chandra, R. and J. Scudder, "Capabilities Advertisement
with BGP-4", RFC 3392, November 2002.
[SBGP1] Kent, S., C. Lynn, and K. Seo, "Secure Border Gateway
Protocol (Secure-BGP)", IEEE Journal on Selected Areas in
Communications, volume 18, number 4, April, 2000.
[SBGP2] Kent, S., C. Lynn, J. Mikkelson, and K. Seo, "Secure
Border Gateway Protocol (S-BGP) -- Real World Performance
and Deployment Issues", Proceedings of the IEEE Network
and Distributed System Security Symposium, February, 2000.
[BFD] Katz, D. and D. Ward, "Bidirectional Forwarding
Detection", Work in Progress, June 2003.
[PEERING] Empirical data gathered from the Sprint and AOL backbones,
October, 2002.
Gill, et al. Experimental [Page 9]
RFC 3682 Generalized TTL Security Mechanism February 2004
[RFC2028] Hovey, R. and S. Bradner, "The Organizations Involved in
the IETF Standards Process", BCP 11, RFC 2028, October
1996.
[RFC2434] Narten, T., and H. Alvestrand, "Guidelines for Writing an
IANA Considerations Section in RFCs", BCP 26, RFC 2434,
October 1998.
[RFC3618] Meyer, D. and W. Fenner, Eds., "The Multicast Source
Discovery Protocol (MSDP)", RFC 3618, October 2003.
Copyright (C) The Internet Society (2004). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assignees.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.
Gill, et al. Experimental [Page 11]