One document matched: draft-ietf-mpls-ldp-hello-crypto-auth-02.xml
<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd">
<?rfc toc="yes"?>
<?rfc tocompact="no"?>
<?rfc tocdepth="6"?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes"?>
<rfc category="std" docName="draft-ietf-mpls-ldp-hello-crypto-auth-02.txt"
ipr="trust200902">
<front>
<title abbrev="LDP Hello Cryptographic Authentication">LDP Hello
Cryptographic Authentication</title>
<author fullname="Lianshu Zheng" initials="L." surname="Zheng">
<organization>Huawei Technologies</organization>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country>China</country>
</postal>
<email>vero.zheng@huawei.com</email>
</address>
</author>
<author fullname="Mach(Guoyi) Chen" initials="M." surname="Chen">
<organization>Huawei Technologies</organization>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country>China</country>
</postal>
<email>mach.chen@huawei.com</email>
</address>
</author>
<author fullname="Manav Bhatia" initials="M." surname="Bhatia">
<organization>Alcatel-Lucent</organization>
<address>
<postal>
<street/>
<city/>
<region/>
<code/>
<country>India</country>
</postal>
<email>manav.bhatia@alcatel-lucent.com</email>
</address>
</author>
<date day="29" month="August" year="2013"/>
<abstract>
<t>This document introduces a new optional Cryptographic Authentication
TLV that LDP can use to secure its Hello messages. It secures the Hello
messages against spoofing attacks and some well known attacks against
the IP header. This document describes a mechanism to secure the LDP
Hello messages using National Institute of Standards and Technology
(NIST) Secure Hash Standard family of algorithms.</t>
</abstract>
<note title="Requirements Language">
<t>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 <xref
target="RFC2119">RFC 2119</xref>.</t>
</note>
</front>
<middle>
<section title="Introduction">
<t>The Label Distribution Protocol (LDP) <xref target="RFC5036"/> sets
up LDP sessions that runs between LDP peers. The peers could either be
directly connected at the link level or could be multiple hops away. An
LDP Label Switching Router (LSR) could either be configured with the
identity of its peers or could discover them using LDP Hello messages.
These messages are sent encapsulated in UDP addressed to "all routers on
this subnet" or to a specific IP address. Periodic Hello messages are
also used to maintain the relationship between LDP peers necessary to
keep the LDP session active. </t>
<t>Unlike other LDP messages, the Hello messages are sent using UDP and
not TCP. This implies that these messages cannot use the security
mechanisms defined for TCP <xref target="RFC5926"> </xref>. Besides a
note that some configuration may help protect against bogus discovery
messages, <xref target="RFC5036"/> does not really provide any security
mechanism to protect the Hello messages.</t>
<t>Spoofing a Hello packet for an existing adjacency can cause the valid
adjacency to time out and in turn can result in termination of the
associated session. This can occur when the spoofed Hello specifies a
smaller Hold Time, causing the receiver to expect Hellos within this
smaller interval, while the true neighbor continues sending Hellos at
the previously agreed lower frequency. Spoofing a Hello packet can also
cause the LDP session to be terminated directly, which can occur when
the spoofed Hello specifies a different Transport Address, other than
the previously agreed one between neighbors. Spoofed Hello messages have
been observed and reported as a real problem in production networks
<xref target="I-D.ietf-karp-routing-tcp-analysis"/>.</t>
<t><xref target="RFC5036"/> describes that the threat of spoofed Basic
Hellos can be reduced by accepting Basic Hellos only on interfaces to
which LSRs that can be trusted are directly connected, and ignoring
Basic Hellos not addressed to the "all routers on this subnet" multicast
group. Spoofing attacks via Extended Hellos are a potentially more
serious threat. An LSR can reduce the threat of spoofed Extended Hellos
by filtering them and accepting only those originating at sources
permitted by an access list. However, filtering using access lists
requires LSR resource, and does not prevent IP-address spoofing.</t>
<t>This document introduces a new Cryptographic Authentication TLV which
is used in LDP Hello message as an optional parameter. It enhances the
authentication mechanism for LDP by securing the Hello message against
spoofing attack. It also introduces a cryptographic sequence number
carried in the Hello messages that can be used to protect against replay
attacks. The LSRs could be configured to only accept Hello messages from
specific peers when authentication is in use.</t>
<t>Using this Cryptographic Authentication TLV, one or more secret keys
(with corresponding key IDs) are configured in each system. For each LDP
Hello packet, the key is used to generate and verify a HMAC Hash that is
stored in the LDP Hello packet. For cryptographic hash function, this
document proposes to use SHA-1, SHA-256, SHA-384, and SHA-512 defined in
US NIST Secure Hash Standard (SHS) <xref target="FIPS-180-3"/>. The HMAC
authentication mode defined in NIST FIPS 198 is used <xref
target="FIPS-198"/>. Of the above, implementations MUST include support
for at least HMAC-SHA-256 and SHOULD include support for HMAC-SHA-1 and
MAY include support for either of HMAC-SHA-384 or HMAC-SHA-512.</t>
</section>
<section title="Cryptographic Authentication TLV">
<section title="Optional Parameter for Hello Message">
<t><xref target="RFC5036"/> defines the encoding for the Hello
message. Each Hello message contains zero or more Optional Parameters,
each encoded as a TLV. Three Optional Parameters are defined by <xref
target="RFC5036"/>. This document defines a new Optional Parameter:
the Cryptographic Authentication parameter.</t>
<figure align="left">
<artwork><![CDATA[
Optional Parameter Type
------------------------------- --------
IPv4 Transport Address 0x0401 (RFC5036)
Configuration Sequence Number 0x0402 (RFC5036)
IPv6 Transport Address 0x0403 (RFC5036)
Cryptographic Authentication 0x0404 (this document, TBD by IANA)
]]></artwork>
</figure>
<t>The Cryptographic Authentication TLV Encoding is described in
section 2.2.</t>
</section>
<section title="Cryptographic Authentication TLV Encoding">
<figure align="left">
<artwork><![CDATA[
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0|0| Auth (TBD) | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Authentication Key ID |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cryptographic Sequence Number (High Order 32 Bits) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cryptographic Sequence Number (Low Order 32 Bits) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
| Authentication Data (Variable) |
~ ~
| |
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
]]></artwork>
</figure>
<t>- Type: TBD, Cryptographic Authentication</t>
<t>- Length: Specifying the length in octets of the value field.</t>
<t>- Auth Key ID: 32 bit field that identifies the algorithm and the
secret key used to create the message digest carried in LDP
payload.</t>
<t>- Cryptographic Sequence Number: 64-bit strictly increasing
sequence number that is used to guard against replay attacks. The
64-bit sequence number MUST be incremented for every LDP Hello packet
sent by the LDP router. Upon reception, the sequence number MUST be
greater than the sequence number in the last LDP Hello packet accepted
from the sending LDP neighbor. Otherwise, the LDP packet is considered
a replayed packet and dropped.</t>
<t>LDP routers implementing this specification MUST use available
mechanisms to preserve the sequence number's strictly increasing
property for the deployed life of the LDP router (including cold
restarts). One mechanism for accomplishing this could be to use the
high-order 32 bits of the sequence number as a wrap/boot count that is
incremented anytime the LDP router loses its sequence number state.
Techniques such as sequence number space partitioning described above
or non-volatile storage preservation can be used but are really beyond
the scope of this specification. Sequence number wrap is described in
<xref target="Sequence-Wrap"></xref>.</t>
<t>- Authentication Data:</t>
<t>This field carries the digest computed by the Cryptographic
Authentication algorithm in use. The length of the Authentication Data
varies based on the cryptographic algorithm in use, which is shown as
below:</t>
<figure align="left">
<artwork><![CDATA[
Auth type Length
--------------- ----------
HMAC-SHA1 20 bytes
HMAC-SHA-256 32 bytes
HMAC-SHA-384 48 bytes
HMAC-SHA-512 64 bytes
]]></artwork>
</figure>
</section>
<section anchor="Sequence-Wrap" title="Sequence Number Wrap">
<t>When incrementing the sequence number for each transmitted LDP packet,
the sequence number should be treated as an unsigned 64-bit value. If the lower
order 32-bit value wraps, the higher order 32-bit value should be incremented and saved in
non-volatile storage. If by some chance the LDP router is deployed long enough
that there is a possibility that the 64-bit sequence number may wrap, all keys,
independent of key distribution mechanism, MUST be reset to avoid the possibility
of replay attacks. Once the keys have been changed, the higher order sequence number
can be reset to 0 and saved to non-volatile storage.</t>
</section>
</section>
<section anchor="Crypto_Auth_Procedure" title="Cryptographic Authentication Procedure">
<t> As noted earlier, the Auth Key ID maps to the authentication algorithm and
the secret key used to generate and verify the message digest. This
specification discusses the computation of LDP Cryptographic Authentication
data when any of the NIST SHS family of algorithms is used in the Hashed
Message Authentication Code (HMAC) mode. <vspace blankLines="1" />
The currently valid algorithms (including mode) for LDP Cryptographic
Authentication include: <vspace blankLines="1" />
HMAC-SHA-1, HMAC-SHA-256, HMAC-SHA-384 and HMAC-SHA-512
<vspace blankLines="1" />
Of the above, implementations of this specification MUST include support for at least
HMAC-SHA-256 and SHOULD include support for HMAC-SHA-1 and MAY also
include support for HMAC-SHA-384 and HMAC-SHA-512.<vspace blankLines="1" />
Implementations of this standard MUST use HMAC-SHA-256 as the default authentication algorithm.
</t>
</section>
<section anchor="Cross-Protocol" title="Cross Protocol Attack Mitigation">
<t>In order to prevent cross protocol replay attacks for protocols sharing common
keys, the two octet LDP Cryptographic Protocol ID is appended to the authentication key
prior to use. Other protocols using cryptographic authentication as specified herein
MUST similarly append their respective Cryptographic Protocol IDs to their keys
in this step. Refer to <xref target="IANA">IANA Considerations</xref>.
</t></section>
<section title="Cryptographic Aspects">
<t>In the algorithm description below, the following nomenclature, which
is consistent with <xref target="FIPS-198"/>, is used:</t>
<t> H is the specific hashing algorithm (e.g. SHA-256). <vspace blankLines="1" />
K is the Authentication Key from the LDP security association.<vspace blankLines="1" />
Ks is a Protocol Specific Authentication Key obtained by appending Authentication Key (K)
with the two-octet LDP Cryptographic Protocol ID appended.<vspace blankLines="1" />
Ko is the cryptographic key used with the hash algorithm.<vspace blankLines="1" />
B is the block size of H, measured in octets rather than bits.<vspace blankLines="1" />
Note that B is the internal block size, not the hash size. </t>
<t> <list style="empty"> <t> For SHA-1 and SHA-256: B == 64 </t>
<t>For SHA-384 and SHA-512: B == 128 </t>
</list></t>
<t> L is the length of the hash, measured in octets rather than bits.<vspace blankLines="1" />
XOR is the exclusive-or operation. <vspace blankLines="1" />
Opad is the hexadecimal value 0x5c repeated B times. <vspace blankLines="1" />
Ipad is the hexadecimal value 0x36 repeated B times. <vspace blankLines="1" />
Apad is a value which is the same length as the hash output or message digest.
In case of IPv4, the first 4 octets contain the IPv4 source address
followed by the hexadecimal value 0x878FE1F3 repeated (L-4)/4 times.
In case of IPv6, the first 16 octets contain the IPv6 source address followed
by the hexadecimal value 0x878FE1F3 repeated (L-16)/4 times.
This implies that hash output is always
a length of at least 16 octets. <vspace blankLines="1" /> </t>
<section title="Preparing the Cryptographic Key">
<t>The LDP Cryptographic Protocol ID is appended to the Authentication Key (K)
yielding a Protocol Specific Authentication Key (Ks).
In this application, Ko is always L octets long. While
<xref target="RFC2104"> </xref> supports a key that is up to B
octets long, this application uses L as the Ks length consistent
with <xref target="RFC4822"> </xref>, <xref target="RFC5310"> </xref>,
<xref target="RFC5709"> </xref> and <xref target="RFC6506"> </xref>. According to
<xref target="FIPS-180-3"/>, Section 3, keys greater than L octets do not
significantly increase the function strength. Ks is computed as
follows: </t>
<t>
If the Protocol Specific Authentication Key (Ks) is L octets long, then Ko is
equal to Ks.
If the Protocol Specific Authentication Key (Ks) is more than L octets long, then
Ko is set to H(Ks). If the Protocol Specific Authentication Key (Ks) is less than
L octets long, then Ko is set to the Protocol Specific Authentication Key (Ks)
with zeros appended to the end of the Protocol Specific
Authentication Key (Ks) such that Ko is L octets long.</t>
</section>
<section title="Computing the Hash">
<t>First, the Authentication Data field in the Cryptographic
Authentication TLV is filled with the value Apad. Then, to compute
HMAC over the Hello packet it performs:</t>
<t>H(Ko XOR Opad || H(Ko XOR Ipad || (Hello Packet)))</t>
<t>Hello Packet refers to the LDP Hello packet excluding the IP
header.</t>
<t> When XORing Ko and Ipad, or XORing Ko and Opad, Ko must be padded with zeros to the
length of Ipad and the Opad. </t>
</section>
<section title="Result">
<t>The resultant Hash becomes the Authentication Data that is sent in
the Authentication Data field of the Cryptographic Authentication TLV.
The length of the Authentication Data field is always identical to the
message digest size of the specific hash function H that is being
used.</t>
<t>This also means that the use of hash functions with larger
output sizes will also increase the size of the LDP packet
as transmitted on the wire.</t>
</section>
</section>
<section title="Processing Hello Message Using Cryptographic Authentication">
<section title="Transmission Using Cryptographic Authentication">
<t>Prior to transmitting Hello message, the Length in the
Cryptographic Authentication TLV header is set as per the
authentication algorithm that is being used. It is set to 24 for
HMAC-SHA-1, 36 for HMAC-SHA-256, 52 for HMAC-SHA-384 and 68 for
HMAC-SHA-512.</t>
<t>The Auth Key ID field is set to the ID of the current
authentication key. The HMAC Hash is computed as explained in Section
3. The resulting Hash is stored in the Authentication Data field prior
to transmission. The authentication key MUST NOT be carried in the
packet.</t>
</section>
<section title="Receipt Using Cryptographic Authentication">
<t>The receiving LSR applies acceptability criteria for received
Hellos using cryptographic authentication. If the Cryptographic
Authentication TLV is unknown to the receiving LSR, the received
packet MUST be discarded according to Section 3.5.1.2.2 of <xref
target="RFC5036"/>.</t>
<t>If the Auth Key ID field does not match the ID of a configured
authentication key, the received packet MUST be discarded.</t>
<t>If the cryptographic sequence number in the LDP packet is less than
or equal to the last sequence number received from the same neighbor,
the LDP packet MUST be discarded.</t>
<t>Before the receiving LSR performs any processing, it needs to save
the values of the Authentication Data field. The receiving LSR then
replaces the contents of the Authentication Data field with Apad,
computes the Hash, using the authentication key specified by the
received Auth Key ID field, as explained in Section 3. If the locally
computed Hash is equal to the received value of the Authentication
Data field, the received packet is accepted for other normal checks
and processing as described in <xref target="RFC5036"/>. Otherwise, if
the locally computed Hash is not equal to the received value of the
Authentication Data field, the received packet MUST be discarded.</t>
</section>
</section>
<section anchor="Security" title="Security Considerations">
<t>Section 1 of this document describes the security issues arising from
the use of unauthenticated LDP Hello messages. In order to address those
issues, it is RECOMMENDED that all deployments use the Cryptographic
Authentication TLV to authenticate the Hello messages.</t>
<t>The quality of the security provided by the Cryptographic
Authentication TLV depends completely on the strength of the
cryptographic algorithm in use, the strength of the key being used, and
the correct implementation of the security mechanism in communicating
LDP implementations. Also, the level of security provided by the
Cryptographic Authentication TLV varies based on the authentication type
used.</t>
<t>It should be noted that the authentication method described in this
document is not being used to authenticate the specific originator of a
packet but is rather being used to confirm that the packet has indeed
been issued by a router that has access to the Authentication Key.</t>
<t>Deployments SHOULD use sufficiently long and random values for the
Authentication Key so that guessing and other cryptographic attacks on
the key are not feasible in their environments. Furthermore, it is
RECOMMENDED that Authentication Keys incorporate at least 128
pseudo-random bits to minimize the risk of such attacks. In support of
these recommendations, management systems SHOULD support hexadecimal
input of Authentication Keys.</t>
<t>The mechanism described herein is not perfect and does not need to be
perfect. Instead, this mechanism represents a significant increase in
the effort required for an adversary to successfully attack the LDP
Hello protocol while not causing undue implementation, deployment, or
operational complexity.</t>
</section>
<section anchor="IANA" title="IANA Considerations">
<t>The IANA is requested to as assign a new TLV from the "Multiprotocol
Label Switching Architecture (MPLS) Label Switched Paths (LSPs)
Parameters - TLVs" registry, "TLVs and sub-TLVs" sub- registry.</t>
<t><figure>
<artwork><![CDATA[Value Meaning Reference
----- -------------------------------- ---------
TBD Cryptographic Authentication TLV this document (sect 3.2)
]]></artwork>
</figure></t>
<t>The IANA is also requested to as assign value from the
"Authentication Cryptographic Protocol ID", registry under
the "Keying and Authentication for Routing Protocols (KARP) Parameters"
category.</t>
<t><figure>
<artwork><![CDATA[
Value Meaning Reference
----- -------------------------------- ---------
TBD LDP Cryptographic Protocol ID this document (sect 4)
]]></artwork>
</figure></t>
</section>
<section anchor="Acknowledgements" title="Acknowledgements">
<t>The authors would like to thank Liu Xuehu for his work on background
and motivation for LDP Hello authentication. The authors also would like
to thank Adrian Farrel, Eric Rosen, Sam Hartman, Eric Gray, Kamran Raza
and Acee Lindem for their valuable comments.</t>
<t> We would also like to thank the authors of RFC 5709 and RFC 6506 from
where we have taken most of the cryptographic computation procedures from. </t>
</section>
</middle>
<back>
<references title="Normative References">
<?rfc include="reference.RFC.2119"?>
<?rfc include='reference.RFC.2104'?>
<?rfc include='reference.RFC.5036'?>
<?rfc include='reference.RFC.5709'?>
<?rfc include='reference.RFC.5310'?>
<?rfc include='reference.RFC.4822'?>
<?rfc include='reference.RFC.6506'?>
<reference anchor="FIPS-180-3">
<front>
<title>Secure Hash Standard (SHS), FIPS PUB 180-3</title>
<author fullname="National Institute of Standards and Technology">
<organization/>
</author>
<date month="October" year="2008"/>
</front>
</reference>
<reference anchor="FIPS-198">
<front>
<title>The Keyed-Hash Message Authentication Code (HMAC), FIPS PUB
198</title>
<author fullname="US National Institute of Standards & Technology">
<organization/>
</author>
<date month="March" year="2002"/>
</front>
</reference>
</references>
<references title="Informative References">
<?rfc include='reference.RFC.5926'?>
<?rfc include='reference.I-D.ietf-karp-routing-tcp-analysis'?>
</references>
</back>
</rfc>
| PAFTECH AB 2003-2026 | 2026-04-22 08:38:21 |