Network Working Group Luyuan Fang (Ed)
Internet Draft Michael Behringer
Category: Informational Cisco Systems, Inc.
Expires: August 2007 Ross Callon
Juniper Networks
J. L. Le Roux
France Telecom
Raymond Zhang
British Telecom
Paul Knight
Nortel
Yaakov Stein
RAD Data Communications
February 2007
Security Framework for MPLS and GMPLS Networks
draft-fang-mpls-gmpls-security-framework-00.txt
Status of this Memo
This memo provides information for the Internet community. It does
not specify an Internet standard of any kind. Distribution of this
memo is unlimited.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.
Internet-Drafts are draft documents valid for a maximum of six
months and may be updated, replaced, or obsoleted by other
documents at any time. It is inappropriate to use Internet-Drafts
as reference material or to cite them other than as "work in
progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.
IPR Disclosure Acknowledgement
By submitting this Internet-Draft, each author represents that any
applicable patent or other IPR claims of which he or she is aware
have been or will be disclosed, and any of which he or she becomes
aware will be disclosed, in accordance with Section 6 of BCP 79.
Copyright Notice
Copyright (C) The IETF Trust (2007).
Fang, et al. Informational 1
MPLS/GMPLS Security framework
February 2007
Abstract
This document provides a security framework for Multiprotocol Label
Switching (MPLS) and Generalized Multiprotocol Label Switching
(GMPLS) Networks (MPLS and GMPLS are described in [RFC3031] and
[RFC3945]). This document addresses the security aspects that are
relevant in the context of MPLS and GMPLS. It describes the
security threats, the related defensive techniques, and the
mechanisms for detection and reporting. This document gives
emphasis to RSVP-TE and LDP security considerations, as well as
Inter-AS and Inter-provider security considerations for building
and maintaining MPLS and GMPLS networks across different domains or
different Service Providers.
Table of Contents
1. Introduction..................................................3
1.1. Structure of This Document.................................4
1.2. Contributors...............................................5
2. Terminology...................................................5
2.1. Terminology................................................5
2.2. Acronyms and Abbreviations.................................7
3. Security Reference Models.....................................7
4. Security Threats..............................................9
4.1. Attacks on the Control Plane..............................10
4.2. Attacks on the Data Plane.................................13
5. Defensive Techniques for MPLS/GMPLS Networks.................15
5.1. Cryptographic techniques..................................16
5.2. Authentication............................................24
5.3. Access Control techniques.................................25
5.4. Use of Isolated Infrastructure............................29
5.5. Use of Aggregated Infrastructure..........................30
5.6. Service Provider Quality Control Processes................30
5.7. Deployment of Testable MPLS/GMPLS Service.................31
6. Monitoring, Detection, and Reporting of Security Attacks.....31
7. Service Provider General Security Requirements...............32
7.1. Protection within the Core Network........................32
7.2. Protection on the User Access Link........................36
7.3. General Requirements for MPLS/GMPLS Providers.............38
8. Inter-provider Security Requirements.........................38
8.1. Control Plane Protection..................................39
Fang, et al. Informational 2
MPLS/GMPLS Security framework
February 2007
8.2. Data Plane Protection.....................................43
9. Security Considerations......................................44
10. IANA Considerations........................................45
11. Normative References.......................................45
12. Informational References...................................46
13. Author's Addresses.........................................47
14. Acknowledgement............................................49
Conventions used in this document
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 RFC2119 [RFC
2119].
1. Introduction
Security is an important aspect of all networks, MPLS and GMPLS
networks being no exception.
MPLS and GMPLS are described in [RFC3031] [RFC3945]. Various
security considerations have been addressed in each of the many
RFCs that address on MPLS and GMPLS technologies, but there has not
been a single document which provides general security
considerations. The motivation for creating this document is to
provide a comprehensive and consistent security framework for MPLS
and GMPLS networks. Each individual document may point to this
document for general security considerations in addition to
providing the security considerations which are specific to the
particular technologies the document is describing.
In this document, we first describe the security threats that are
relevant in the context of MPLS and GMPLS, and the defensive
techniques that can be used to combat those threats. We consider
security issues deriving both from malicious or incorrect behavior
of users and other parties and from negligent or incorrect behavior
of the providers. An important part of security defense is the
detection and report of a security attack, which is also addressed
in this document.
We then discuss the possible service provider security requirements
in a MPLS or GMPLS environment. The users have expectations that
need to be met on the security characteristics of MPLS or GMPLS
networks. These will include the security requirements for MPLS and
GMPLS supporting equipments, and the provider operation security
Fang, et al. Informational 3
MPLS/GMPLS Security framework
February 2007
requirements. The service providers must protect their network
infrastructure, and make it secure to the level required to provide
services over their MPLS or GMPLS networks.
Inter-As and Inter-provider security are discussed with special
emphasis, since the security risk factors are higher with inter-
provider connections. Depending on different MPLS or GMPLS
techniques used, the degree of risk and the mitigation
methodologies vary. This document discusses the security aspects
and requirements for certain basic MPLS and GMPLS techniques and
inter-connection models. This document does not attempt to cover
all current and future MPLS and GMPLS technologies, since it is not
within the scope of this document to analyze the security
properties of specific technologies.
It is important to clarify that, in this document; we limit
ourselves to describing the providers' security requirements that
pertain to MPLS and GMPLS networks. Readers may refer to the
"Security Best Practices Efforts and Documents" [opsec effort] and
"Security Mechanisms for the Internet" [RFC3631] for general
network operation security considerations. It is not our intention,
however, to formulate precise "requirements" on each specific
technology in terms of defining the mechanisms and techniques that
must be implemented to satisfy such security requirements.
1.1. Structure of This Document
This document is organized as follows. In Section 2, we define the
terminology used in the document. In section 3, we define the
security reference models for security in MPLS/GMPLS networks,
which we use in the rest of the document. In Section 4, we describe
the security threats that are specific of MPLS and GMPLS. In
Section 5, we review defense techniques that may be used against
those threats. In Section 6, we describe how attacks may be
detected and reported. In Section 7, we describe security
requirements that the provider may have in order to guarantee the
security of the network infrastructure to provide MPLS/GMPLS
services. In section 8, we discuss Inter-provider security
requirements. Finally, in Section 9, we discuss security
considerations of this document.
This document has used relevant content from RFC 4111 "Security
Framework of Provider Provisioned VPN" [RFC4111], and "MPLS
InterCarrier Interconnect Technical Specification" [MFA MPLS ICI]
in the Inter-provider security discussion. We acknowledge the
authors of these documents for the valuable information and text.
Fang, et al. Informational 4
MPLS/GMPLS Security framework
February 2007
1.2. Contributors
As the design team members of MPLS security Framework, the
following made significant contributions to this document.
Nabil Bitar, Verizon
Monique Morrow, Cisco systems, Inc.
Jerry Ash, AT&T
2. Terminology
2.1. Terminology
This document uses MPLS and GMPLS specific terminology. Definitions
and details about MPLS and GMPLS terminology can be found in
[RFC3031] and [RFC3945]. The most important definitions are
repeated in this section, for other definitions the reader is
referred to [RFC3031] and [RFC3945].
CE: Customer Edge device. A Customer Edge device is a router or a
switch in the customer network interfacing with the Service
Provider's network.
Forwarding equivalence class (FEC): A group of IP packets which are
forwarded in the same manner (e.g., over the same path, with the
same forwarding treatment)
Label: A short fixed length physically contiguous identifier which
is used to identify a FEC, usually of local significance.
Label switched hop: the hop between two MPLS nodes, on which
forwarding is done using labels.
Label switched path (LSP): The path through one or more LSRs at one
level of the hierarchy followed by a packets in a particular FEC.
Label switching router (LSR): an MPLS node which is capable of
forwarding native L3 packets
Layer 2: the protocol layer under layer 3 (which therefore offers
the services used by layer 3). Forwarding, when done by the
swapping of short fixed length labels, occurs at layer 2 regardless
of whether the label being examined is an ATM VPI/VCI, a frame
relay DLCI, or an MPLS label.
Fang, et al. Informational 5
MPLS/GMPLS Security framework
February 2007
Layer 3: the protocol layer at which IP and its associated routing
protocols operate link layer synonymous with layer 2.
Loop detection: a method of dealing with loops in which loops are
allowed to be set up, and data may be transmitted over the loop,
but the loop is later detected.
Loop prevention: a method of dealing with loops in which data is
never transmitted over a loop.
Label stack: an ordered set of labels.
Merge point: a node at which label merging is done
MPLS domain: a contiguous set of nodes which operate MPLS routing
and forwarding and which are also in one Routing or Administrative
Domain.
MPLS edge node: an MPLS node that connects an MPLS domain with a
node which is outside of the domain, either because it does not run
MPLS, and/or because it is in a different domain. Note that if an
LSR has a neighboring host which is not running MPLS, that that LSR
is an MPLS edge node.
P: Provider Router. The Provider Router is a router in the Service
Provider's core network that does not have interfaces directly
towards the customer. A P router is used to interconnect the PE
routers.
MPLS egress node: an MPLS edge node in its role in handling traffic
as it leaves an MPLS domain
MPLS ingress node: an MPLS edge node in its role in handling
traffic as it enters an MPLS domain
MPLS label: a label which is carried in a packet header, and which
represents the packet's FEC
MPLS node: a node which is running MPLS. An MPLS node will be
aware of MPLS control protocols, will operate one or more L3
routing protocols, and will be capable of forwarding packets based
on labels. An MPLS node may optionally be also capable of
forwarding native L3 packets.
MultiProtocol Label Switching (MPLS): an IETF working group and the
effort associated with the working group
Fang, et al. Informational 6
MPLS/GMPLS Security framework
February 2007
PE: Provider Edge device. The Provider Edge device is the equipment
in the Service Provider's network that interfaces with the
equipment in the customer's network.
SP: Service Provider.
VPN: Virtual Private Network. Restricted communication between a
set of sites, making use of an IP backbone which is shared by
traffic that is not going to or coming from those sites. [RFC4110].
2.2. Acronyms and Abbreviations
AS Autonomous System
ASBR Autonomous System Border Router
ATM Asynchronous Transfer Mode
BGP Border Gateway Protocol
FEC Forwarding Equivalence Class
GMPLS Generalized Multi-Protocol Label Switching
IGP Interior Gateway Protocol
IP Internet Protocol
LDP Label Distribution Protocol
L2 Layer 2
L3 Layer 3
LSP Label Switched Path
LSR Label Switching Router
MPLS MultiProtocol Label Switching
MP-BGP Multi-Protocol BGP
PCE Path Calculation Element
PSN Packet-Switched Network
RSVP-TE Resource Reservation Protocol with Traffic Engineering
Extensions
TTL Time-To-Live
VPN Virtual Private Network
3. Security Reference Models
This section defines a reference model for security in MPLS/GMPLS
networks.
A MPLS/GMPLS core network is defined here as the central network
infrastructure (P and PE routers). A MPLS/GMPLS core network
consists of one or more SP networks. All network elements in the
core are under the operational control of one or more MPLS/GMPLS
service providers. Even if the MPLS/GMPLS core is provided by
several service providers, towards the end users it appears as a
single zone of trust. However, when several service providers
Fang, et al. Informational 7
MPLS/GMPLS Security framework
February 2007
provide together an MPLS/GMPLS core, each SP still needs to secure
itself against the other SPs.
A MPLS/GMPLS end user is a company, institution or residential
client of the SP.
This document defines each MPLS in a single domain a trusted zone.
A primary concern is about security aspects that relate to breaches
of security from the "outside" of a trusted zone to the "inside" of
this zone. Figure 1 depicts the concept of trusted zones within the
MPLS/GMPLS framework.
/-------------\
+------------+ / \ +------------+
| MPLS/GMPLS +---/ \--------+ MPLS |
| user | MPLS/GMPLS Core | user |
| site +---\ /XXX-----+ site |
+------------+ \ / XXX +------------+
\-------------/ | |
| |
| +------\
+--------/ "Internet"
MPLS/GMPLS Core with user connections and Internet connection
Figure 1: The MPLS/GMPLS trusted zone model
The trusted zone defined is the MPLS/GMPLS core/network in a single
AS within a single Service Provider.
In principle the trusted zones should be separate; however,
typically MPLS core networks also offer Internet access, in which
case a transit point (marked with "XXX" in the figure 1) is
defined. In the case of MPLS/GMPLS inter-provider connection, the
trusted zone ends at the ASBR (marked with "B" in the figure 2) of
the considered AS/provider.
A key requirement of MPLS and GMPLS networks is that the security
of the trusted zone not be compromised by interconnecting the
MPLS/GMPLS core infrastructure with another provider core
(MPLS/GMPLS or non-MPLS/GMPLS), Internet, or end user access.
In addition, neighbors may be trusted or untrusted. Neighbors may
be authorized or unauthorized. Even though a neighbor may be
authorized for communication, it may not be trusted. For example,
Fang, et al. Informational 8
MPLS/GMPLS Security framework
February 2007
when connecting with another provider ASBRs to set up inter-AS
LSPs, the other provider is considered as an untrusted but
authorized neighbor.
+---------------+ +----------------+
| | | |
| MPLS/GMPLS ASBR1----ASBR3 MPLS/GMPLS |
CE1--PE1 Network | | Network PE2--CE2
| Provider A ASBR2----ASBR4 Provider B |
| | | |
+---------------+ +----------------+
For Provider A:
Trusted Zone: Provider A MPSL/GMPLS network
Trusted neighbor: PE1, ASBR1, ASBR2
Authorized but untrusted neighbor: provider B
Unauthorized neighbor: CE1, CE2
Figure 2. MPLS/GMPLS trusted zone and authorized neighbor
Security against threats that originate within the same trusted
zone as their targets (for example, attacks from within the core
network) is outside the scope of this document.
Also outside the scope are all aspects of network security which
are independent of whether a network is a MPLS/GMPLS network (for
example, attacks from the Internet to a user web-server which is
connected through the MPLS/GMPLS network will not be considered
here, unless the way the MPLS/GMPLS network is provisioned could
make a difference to the security of this user server).
4. Security Threats
This section discusses the various network security threats that
may endanger MPLS/GMPLS networks. The discussion is limited to
those threats that are unique to MPLS/GMPLS networks, or that
affect MPLS/GMPLS network in unique ways.
A successful attack on a particular MPLS/GMPLS network or on a
service provider's MPLS/GMPLS infrastructure may cause one or more
of the following ill effects:
- Observation, modification, or deletion of provider/user data.
- Replay of provider/user data.
Fang, et al. Informational 9
MPLS/GMPLS Security framework
February 2007
- Injection of non-authentic data into a provider/user traffic
stream.
- Traffic pattern analysis on provider/user traffic.
- Disruption of provider/user connectivity.
- Degradation of provider service quality.
It is useful to consider that threats, whether malicious or
accidental, may come from different categories of sources. For
example they may come from:
- Other users whose services are provided by the same MPLS/GMPLS
core.
- The MPLS/GMPLS service provider or persons working for it.
- Other persons who obtain physical access to a MPLS/GMPLS service
provider site.
- Other persons who use social engineering methods to influence
behavior of service provider personnel.
- Users of the MPLS/GMPLS network itself, i.e. intra-VPN threats.
(Such threats are beyond the scope of this document.)
- Others i.e. attackers from the Internet at large.
- Other service provider in the case of MPLS/GMPLS Inter-provider
connection. The core of the other provider may or may not be using
MPLS/GMPLS core.
Given that security is generally a compromise between expense and
risk, it is also useful to consider the likelihood of different
attacks occurring. There is at least a perceived difference in the
likelihood of most types of attacks being successfully mounted in
different environments, such as:
- A MPLS/GMPLS inter-connecting with another provider's core
- A MPLS/GMPLS transiting the public Internet
Most types of attacks become easier to mount and hence more likely
as the shared infrastructure via which service is provided expands
from a single service provider to multiple cooperating providers to
the global Internet. Attacks that may not be of sufficient
likeliness to warrant concern in a closely controlled environment
often merit defensive measures in broader, more open environments.
The following sections discuss specific types of exploits that
threaten MPLS/GMPLS networks.
4.1. Attacks on the Control Plane
This category encompasses attacks on the control structures
operated by the service provider with MPLS/GMPLS cores.
Fang, et al. Informational 10
MPLS/GMPLS Security framework
February 2007
4.1.1. LSP creation by an unauthorized element
The unauthorized element can be a local CE or a router in another
domain. An unauthorized element can generate MPLS signaling
messages. At the least, this can result in extra control plane and
forwarding state, and if successful, network bandwidth could be
reserved unnecessarily.
4.1.2. LSP message interception
This threat might be accomplished by monitoring network traffic,
although it would require physical intrusion. If successful, it
could provide information leading to label spoofing attacks. It
also raises confidentiality issues.
4.1.3. Attacks against RSVP-TE
RSVP-TE, described in [RFC3209], is the control protocol used to
set up GMPLS and traffic engineered MPLS tunnels.
There are two major types of attacks against an MPLS domain based
on RSVP-TE. The attacker may set up numerous unauthorized LSPs, or
may send a storm of RSVP messages in a DoS attack. It has been
demonstrated that unprotected routers running RSVP can be
effectively disabled by both types of DoS attacks.
These attacks may even be combined, by using the unauthorized LSPs
to transport additional RSVP (or other) messages across routers
where they might otherwise be filtered out. RSVP attacks can be
launched against adjacent routers at the border with the attacker,
or against non-adjacent routers within the MPLS domain, if there is
no effective mechanism to filter them out.
4.1.4. Attacks against LDP
LDP, described in [RFC3036], is the control protocol used to set up
non-TE MPLS tunnels.
There are two significant types of attack against LDP. An
unauthorized network element can establish an LDP session by
sending LDP Hello and LDP Init messages, leading to the potential
setup of an LSP, as well as accompanying LDP state table
consumption. Even without successfully established LSPs, an
attacker can launch a DoS attack in the form of a storm of LDP
Hello messages and/or LDP TCP Syn messages, leading to high CPU
utilization on the target router.
Fang, et al. Informational 11
MPLS/GMPLS Security framework
February 2007
4.1.5. Denial of Service Attacks on the Network Infrastructure
DoS attacks could be accomplished through an MPLS signaling storm,
resulting in high CPU utilization and possibly leading to control
plane resource starvation.
Control plane DOS attacks can be mounted specifically against the
mechanisms the service provider uses to provide various services,
or against the general infrastructure of the service provider e.g.
P routers or shared aspects of PE routers. (Attacks against the
general infrastructure are within the scope of this document only
if the attack happens in relation with the MPLS/GMPLS
infrastructure, otherwise is not MPLS/GMPLS-specific issue.)
The attacks described in the following sections may each have
denial of service as one of their effects. Other DOS attacks are
also possible.
4.1.6. Attacks on the Service Provider MPLS/GMPLS Equipment Via
Management Interfaces
This includes unauthorized access to service provider
infrastructure equipment, for example to reconfigure the equipment
or to extract information (statistics, topology, etc.) pertaining
to the network.
4.1.7. Social Engineering Attacks on the Service Provider
Infrastructure
Attacks in which the service provider network is reconfigured or
damaged, or in which confidential information is improperly
disclosed, may be mounted through manipulation of service provider
personnel. These types of attacks are MPLS/GMPLS-specific if they
affect MPLS/GMPLS-serving mechanisms.
4.1.8. Cross-connection of Traffic Between Users
This refers to the event where expected isolation between separate
users (who may be VPN users) is breached. This includes cases such
as:
- A site being connected into the "wrong" VPN.
- Traffic being replicated and sent to an unauthorized
user.
- Two or more VPNs being improperly merged together.
- A point-to-point VPN connecting the wrong two points.
Fang, et al. Informational 12
MPLS/GMPLS Security framework
February 2007
- Any packet or frame being improperly delivered outside the VPN
to which it belongs.
Mis-connection or cross-connection of VPNs may be caused by service
provider or equipment vendor error, or by the malicious action of
an attacker. The breach may be physical (e.g. PE-CE links mis-
connected) or logical (improper device configuration).
Anecdotal evidence suggests that the cross-connection threat is one
of the largest security concerns of users (or would-be users).
4.1.9. Attacks Against User Routing Protocols
This encompasses attacks against underlying routing protocols that
are run by the service provider and that directly support the
MPLS/GMPLS core. (Attacks against the use of routing protocols for
the distribution of backbone (non-VPN) routes are beyond the scope
of this document.) Specific attacks against popular routing
protocols have been widely studied and described in [Beard].
4.1.10. Other Attacks on Control Traffic
Besides routing and management protocols (covered separately in the
previous sections) a number of other control protocols may be
directly involved in delivering the services by the MPLS/GMPLS
core. These include but may not be limited to:
- MPLS signaling (LDP, RSVP-TE) discussed above in subsections
4.1.4 and 4.1.3
- PCE signaling
- IPsec signaling (IKE)
- L2TP
- BGP-based membership discovery
- Database-based membership discovery (e.g. RADIUS-based)
Attacks might subvert or disrupt the activities of these protocols,
for example via impersonation or DOS attacks.
4.2. Attacks on the Data Plane
This category encompasses attacks on the provider or end user's
data. Note that from the MPLS/GMPLS network end user's point of
view, some of this might be control plane traffic, e.g. routing
protocols running from the user site A to the user site B via an L2
or L3 connection which may be some type of VPN.
Fang, et al. Informational 13
MPLS/GMPLS Security framework
February 2007
4.2.1. Unauthorized Observation of Data Traffic
This refers to "sniffing" provider/end user packets and examining
their contents. This can result in exposure of confidential
information. It can also be a first step in other attacks
(described below) in which the recorded data is modified and re-
inserted, or re-inserted as-is.
4.2.2. Modification of Data Traffic
This refers to modifying the contents of packets as they traverse
the MPLS/GMPLS core.
4.2.3. Insertion of Non-Authentic Data Traffic: Spoofing and
Replay
This refers to the insertion (or "spoofing") into the user packets
that do not belong there, with the objective of having them
accepted by the recipient as legitimate. Also included in this
category is the insertion of copies of once-legitimate packets that
have been recorded and replayed.
4.2.4. Unauthorized Deletion of Data Traffic
This refers to causing packets to be discarded as they traverse the
MPLS/GMPLS networks. This is a specific type of Denial of Service
attack.
4.2.5. Unauthorized Traffic Pattern Analysis
This refers to "sniffing" provider/user packets and examining
aspects or meta-aspects of them that may be visible even when the
packets themselves are encrypted. An attacker might gain useful
information based on the amount and timing of traffic, packet
sizes, source and destination addresses, etc. For most users, this
type of attack is generally considered to be significantly less of
a concern than the other types discussed in this section.
4.2.6. Denial of Service Attacks
Denial of Service (DOS) attacks are those in which an attacker
attempts to disrupt or prevent the use of a service by its
legitimate users. Taking network devices out of service, modifying
their configuration, or overwhelming them with requests for service
are several of the possible avenues for DOS attack.
Overwhelming the network with requests for service, otherwise known
as a "resource exhaustion" DOS attack, may target any resource in
Fang, et al. Informational 14
MPLS/GMPLS Security framework
February 2007
the network e.g. link bandwidth, packet forwarding capacity,
session capacity for various protocols, CPU power, and so on.
DOS attacks of the resource exhaustion type can be mounted against
the data plane of a particular provider or end-user by attempting
to insert (spoofing) an overwhelming quantity of non-authentic data
into the provider/end user network from the outside of the trusted
zone. Potential results might be to exhaust the bandwidth available
to that provider/end user or to overwhelm the cryptographic
authentication mechanisms of the provider or end user.
Data plane resource exhaustion attacks can also be mounted by
overwhelming the service provider's general (MPLS/GMPLS-
independent) infrastructure with traffic. These attacks on the
general infrastructure are not usually a MPLS/GMPLS-specific issue,
unless the attack is mounted by another MPLS/GMPLS network user
from a privileged position. (E.g. a MPLS/GMPLS network user might
be able to monopolize network data plane resources and thus disrupt
other users.)
5. Defensive Techniques for MPLS/GMPLS Networks
The defensive techniques discussed in this document are intended to
describe methods by which some security threats can be addressed.
They are not intended as requirements for all MPLS/GMPLS
implementations. The MPLS/GMPLS provider should determine the
applicability of these techniques to the provider's specific
service offerings, and the end user may wish to assess the value of
these techniques to the user's service requirements.
The techniques discussed here include encryption, authentication,
filtering, firewalls, access control, isolation, aggregation, and
other techniques.
Nothing is ever 100% secure. Defense therefore involves protecting
against those attacks that are most likely to occur and/or that
have the most dire consequences if successful. For those attacks
that are protected against, absolute protection is seldom
achievable; more often it is sufficient just to make the cost of a
successful attack greater than what the adversary will be willing
to expend.
Successfully defending against an attack does not necessarily mean
the attack must be prevented from happening or from reaching its
target. In many cases the network can instead be designed to
withstand the attack. For example, the introduction of non-
authentic packets could be defended against by preventing their
Fang, et al. Informational 15
MPLS/GMPLS Security framework
February 2007
introduction in the first place, or by making it possible to
identify and eliminate them before delivery to the MPLS/GMPLS
user's system. The latter is frequently a much easier task.
5.1. Cryptographic techniques
MPLS/GMPLS defenses against a wide variety of attacks can be
enhanced by the proper application of cryptographic techniques.
These are the same cryptographic techniques which are applicable to
general network communications. In general, these techniques can
provide confidentiality (encryption) of communication between
devices, authentication of the identities of the devices, and can
ensure that it will be detected if the data being communicated is
changed during transit.
Several aspects of authentication are addressed in some detail in a
separate "Authentication" section.
Encryption adds complexity to a service, and thus it may not be a
standard offering within every user service. There are a few
reasons why encryption may not be a standard offering within every
user service. Encryption adds an additional computational burden to
the devices performing encryption and decryption. This may reduce
the number of user connections which can be handled on a device or
otherwise reduce the capacity of the device, potentially driving up
the provider's costs. Typically, configuring encryption services
on devices adds to the complexity of the device configuration and
adds incremental labor cost. Packet sizes are typically increased
when the packets are secured, increasing the network traffic load
and adding to the likelihood of packet fragmentation with its
increased overhead. (This packet length increase can often be
mitigated to some extent by data compression techniques, but at the
expense of additional computational burden.) Finally, some
providers may employ enough other defensive techniques, such as
physical isolation or filtering/firewall techniques, that they may
not perceive additional benefit from encryption techniques.
The trust model among the MPLS/GMPLS user, the MPLS/GMPLS provider,
and other parts of the network is a key element in determining the
applicability of encryption for any specific MPLS/GMPLS
implementation. In particular, it determines where encryption
should be applied:
- If the data path between the user's site and the
provider's PE is not trusted, then encryption may be used
on the PE-CE link.
- If some part of the backbone network is not trusted,
particularly in implementations where traffic may travel
Fang, et al. Informational 16
MPLS/GMPLS Security framework
February 2007
across the Internet or multiple provider networks, then
the PE-PE traffic may be encrypted.
- If the user does not trust any zone outside of its
premises, it may require end-to-end or CE-CE encryption
service. This service fits within the scope of this
MPLS/GMPLS security framework when the CE is provisioned
by the MPLS/GMPLS provider.
- If the user requires remote access to a its site from a
system at a location which is not a customer location (for
example, access by a traveler) there may be a requirement
for encrypting the traffic between that system and an
access point or at a customer site. If the MPLS/GMPLS
provider provides the access point, then the customer must
cooperate with the provider to handle the access control
services for the remote users. These access control
services are usually implemented using encryption, as
well.
Although CE-CE encryption provides confidentiality against third-
party interception, if the MPLS/GMPLS provider has complete
management control over the CE (encryption) devices, then it may be
possible for the provider to gain access to the user's traffic or
internal network. Encryption devices can potentially be configured
to use null encryption, bypass encryption processing altogether, or
provide some means of sniffing or diverting unencrypted traffic.
Thus an implementation using CE-CE encryption needs to consider the
trust relationship between the MPLS/GMPLS user and provider.
MPLS/GMPLS users and providers may wish to negotiate a service
level agreement (SLA) for CE-CE encryption that will provide an
acceptable demarcation of responsibilities for management of
encryption on the CE devices. The demarcation may also be affected
by the capabilities of the CE devices. For example, the CE might
support some partitioning of management, a configuration lock-down
ability, or allow both parties to verify the configuration. In
general, the MPLS/GMPLS user needs to have a fairly high level of
trust that the MPLS/GMPLS provider will properly provision and
manage the CE devices, if the managed CE-CE model is used.
5.1.1. IPsec in MPLS/GMPLS
IPsec [RFC4301] [RFC4302] [RFC4305] [RFC4306] [RFC2411] is the
security protocol of choice for encryption at the IP layer (Layer
3). IPsec provides robust security for IP traffic between pairs of
devices. Non-IP traffic must be converted to IP (e.g. by
encapsulation) in order to exploit IPsec.
Fang, et al. Informational 17
MPLS/GMPLS Security framework
February 2007
In the MPLS/GMPLS model, IPsec can be employed to protect IP
traffic between PEs, between a PE and a CE, or from CE to CE. CE-
to-CE IPsec may be employed in either a provider-provisioned or a
user-provisioned model. Likewise, encryption of data which is
performed within the user's site is outside the scope of this
document, since it is simply handled as user data by the MPLS/GMPLS
core.
IPsec does not itself specify an encryption algorithm. It can use
a variety of encryption algorithms, with various key lengths, such
as AES encryption. There are trade-offs between key length,
computational burden, and the level of security of the encryption.
A full discussion of these trade-offs is beyond the scope of this
document. In practice, any currently recommended IPsec encryption
offers enough security to substantially reduce the likelihood of
being directly targeted by an attacker; other weaker links in the
chain of security are likely to be attacked first. MPLS/GMPLS
users may wish to use a Service Level Agreement (SLA) specifying
the Service Provider's responsibility for ensuring data
confidentiality, rather than analyzing the specific encryption
techniques used in the MPLS/GMPLS service.
For many of the MPLS/GMPLS provider's network control messages and
some user requirements, cryptographic authentication of messages
without encryption of the contents of the message may provide
acceptable security. Using IPsec, authentication of messages is
provided by the Authentication Header (AH) or through the use of
the Encapsulating Security Protocol (ESP) with authentication only.
Where control messages require authentication but do not use IPsec,
then other cryptographic authentication methods are available.
Message authentication methods currently considered to be secure
are based on hashed message authentication codes (HMAC) [RFC2104]
implemented with a secure hash algorithm such as Secure Hash
Algorithm 1 (SHA-1) [RFC3174].
The currently recommended mechanism to provide a combination of
confidentiality, data origin authentication, and connectionless
integrity is the use of AES in CCM (Counter with CBC-MAC) mode
(AES-CCM) [AES-CCM], with an explicit initialization vector (IV),
as the IPsec ESP.
MPLS/GMPLS which provide differentiated services based on traffic
type may encounter some conflicts with IPsec encryption of traffic.
Since encryption hides the content of the packets, it may not be
possible to differentiate the encrypted traffic in the same manner
as unencrypted traffic. Although DiffServ markings are copied to
Fang, et al. Informational 18
MPLS/GMPLS Security framework
February 2007
the IPsec header and can provide some differentiation, not all
traffic types can be accommodated by this mechanism.
5.1.2. Encryption for device configuration and management
For configuration and management of MPLS/GMPLS devices, encryption
and authentication of the management connection at a level
comparable to that provided by IPsec is desirable.
Several methods of transporting MPLS/GMPLS device management
traffic offer security and confidentiality.
- Secure Shell (SSH) offers protection for TELNET [STD-8] or
terminal-like connections to allow device configuration.
- SNMP v3 [STD62] provides encrypted and authenticated protection
for SNMP-managed devices.
- Transport Layer Security (TLS) [RFC4346] and the closely-related
Secure Sockets Layer (SSL) are widely used for securing HTTP-
based communication, and thus can provide support for most XML-
and SOAP-based device management approaches.
- As of 2004, there is extensive work proceeding in several
organizations (OASIS, W3C, WS-I, and others) on securing device
management traffic within a "Web Services" framework, using a
wide variety of security models, and providing support for
multiple security token formats, multiple trust domains,
multiple signature formats, and multiple encryption
technologies.
- IPsec provides the services with security and confidentiality at
the network layer. With regards to device management, its
current use is primarily focused on in-band management of user-
managed IPsec gateway devices.
5.1.3. Cryptographic techniques for MPLS Pseudowires
5.1.4. 5.1.3 Security Considerations for MPLS Pseudowires
In addition to IP traffic, MPLS networks may be used to transport
other services such as Ethernet, ATM, frame relay, and TDM. This is
done by setting up pseudowires (PWs) that tunnel the native service
through the MPLS core by encapsulating at the edges. The PWE
architecture is defined in [RFC3985].
Fang, et al. Informational 19
MPLS/GMPLS Security framework
February 2007
PW tunnels may be set up using the PWE control protocol based on
LDP [RFC4447], and thus security considerations for LDP will most
likely be applicable to the PWE3 control protocol as well.
PW user packets contain at least one MPLS label (the PW label) and
may contain one or more MPLS tunnel labels. After the label stack
there is a four-byte control word (which is optional for some PW
types), followed by the native service payload. It must be
stressed that encapsulation of MPLS PW packets in IP for the
purpose of enabling use of IPsec mechanisms is not a valid option.
The PW client traffic may be secured by use of mechanisms beyond
the scope of this document.
5.1.5. End-to-end vs. hop-by-hop encryption tradeoffs in
MPLS/GMPLS
In MPLS/GMPLS, encryption could potentially be applied to the
MPLS/GMPLS traffic at several different places. This section
discusses some of the tradeoffs in implementing encryption in
several different connection topologies among different devices
within a MPLS/GMPLS network.
Encryption typically involves a pair of devices which encrypt the
traffic passing between them. The devices may be directly
connected (over a single "hop"), or there may be intervening
devices which transport the encrypted traffic between the pair of
devices. The extreme cases involve using encryption between every
adjacent pair of devices along a given path (hop-by-hop), or using
encryption only between the end devices along a given path (end-to-
end). To keep this discussion within the scope of this document,
the latter ("end-to-end") case considered here is CE-to-CE rather
than fully end-to-end.
Figure 3 depicts a simplified topology showing the Customer Edge
(CE) devices, the Provider Edge (PE) devices, and a variable number
(three are shown) of Provider core (P) devices which might be
present along the path between two sites in a single VPN, operated
by a single service provider (SP).
Site_1---CE---PE---P---P---P---PE---CE---Site_2
Figure 3: Simplified topology traversing through MPLS/GMPLS core
Fang, et al. Informational 20
MPLS/GMPLS Security framework
February 2007
Within this simplified topology, and assuming that P devices are
not to be involved with encryption, there are four basic feasible
configurations for implementing encryption on connections among the
devices:
1) Site-to-site (CE-to-CE) - Encryption can be configured between
the two CE devices, so that traffic will be encrypted throughout
the SP's network.
2) Provider edge-to-edge (PE-to-PE) - Encryption can be configured
between the two PE devices. Unencrypted traffic is received at one
PE from the customer's CE, then it is encrypted for transmission
through the SP's network to the other PE, where it is decrypted and
sent to the other CE.
3) Access link (CE-to-PE) - Encryption can be configured between
the CE and PE, on each side (or on only one side).
4) Configurations 2 and 3 above can also be combined, with
encryption running from CE to PE, then PE to PE, then PE to CE.
Among the four feasible configurations, key tradeoffs in
considering encryption include:
- Vulnerability to link eavesdropping - assuming an attacker can
observe the data in transit on the links, would it be protected
by encryption?
- Vulnerability to device compromise - assuming an attacker can get
access to a device (or freely alter its configuration), would the
data be protected?
- Complexity of device configuration and management - given the
number of sites per VPN customer as Nce and the number of PEs
participating in a given VPN as Npe, how many device configurations
need to be created or maintained, and how do those configurations
scale?
- Processing load on devices - how many encryption or decryption
operations must be done given P packets? - This influences
considerations of device capacity and perhaps end-to-end delay.
- Ability of SP to provide enhanced services (QoS, firewall,
intrusion detection, etc.) - Can the SP inspect the data in order
to provide these services?
Fang, et al. Informational 21
MPLS/GMPLS Security framework
February 2007
These tradeoffs are discussed for each configuration, below:
1) Site-to-site (CE-to-CE)
Link eavesdropping - protected on all links
Device compromise - vulnerable to CE compromise
Complexity - single administration, responsible for one device per
site (Nce devices), but overall configuration per VPN scales as
Nce**2
Processing load - on each of two CEs, each packet is either
encrypted or decrypted (2P)
Enhanced services - severely limited; typically only Diffserv
markings are visible to SP, allowing some QoS services
2) Provider edge-to-edge (PE-to-PE)
Link eavesdropping - vulnerable on CE-PE links; protected on SP's
network links
Device compromise - vulnerable to CE or PE compromise
Complexity - single administration, Npe devices to configure.
(Multiple sites may share a PE device so Npe is typically much
less than Nce.) Scalability of the overall configuration
depends on the PPVPN type: If the encryption is separate per
VPN context, it scales as Npe**2 per customer VPN. If the
encryption is per-PE, it scales as Npe**2 for all customer VPNs
combined.
Processing load - on each of two PEs, each packet is either
encrypted or decrypted (2P)
Enhanced services - full; SP can apply any enhancements based on
detailed view of traffic
3) Access link (CE-to-PE)
Link eavesdropping - protected on CE-PE link; vulnerable on SP's
network links
Device compromise - vulnerable to CE or PE compromise
Complexity - two administrations (customer and SP) with device
configuration on each side (Nce + Npe devices to configure) but
since there is no mesh the overall configuration scales as Nce.
Processing load - on each of two CEs, each packet is either
encrypted or decrypted, plus on each of two PEs, each packet is
either encrypted or decrypted (4P)
Enhanced services - full; SP can apply any enhancements based on
detailed view of traffic
4) Combined Access link and PE-to-PE (essentially hop-by-hop)
Link eavesdropping - protected on all links
Fang, et al. Informational 22
MPLS/GMPLS Security framework
February 2007
Device compromise - vulnerable to CE or PE compromise
Complexity - two administrations (customer and SP) with device
configuration on each side (Nce + Npe devices to configure).
Scalability of the overall configuration depends on the PPVPN
type: If the encryption is separate per VPN context, it scales
as Npe**2 per customer VPN. If the encryption is per-PE, it
scales as Npe**2 for all customer VPNs combined.
Processing load - on each of two CEs, each packet is either
encrypted or decrypted, plus on each of two PEs, each packet is
both encrypted and decrypted (6P)
Enhanced services - full; SP can apply any enhancements based on
detailed view of traffic
Given the tradeoffs discussed above, a few conclusions can be made:
- Configurations 2 and 3 are subsets of 4 that may be appropriate
alternatives to 4 under certain threat models; the remainder of
these conclusions compare 1 (CE-to-CE) vs. 4 (combined access links
and PE-to-PE).
- If protection from link eavesdropping is most important, then
configurations 1 and 4 are equivalent.
- If protection from device compromise is most important and the
threat is to the CE devices, both cases are equivalent; if the
threat is to the PE devices, configuration 1 is best.
- If reducing complexity is most important, and the size of the
network is very small, configuration 1 is the best. Otherwise
configuration 4 is the best because rather than a mesh of CE
devices it requires a smaller mesh of PE devices. Also under some
PPVPN approaches the scaling of 4 is further improved by sharing
the same PE-PE mesh across all VPN contexts. The scaling advantage
of 4 may be increased or decreased in any given situation if the CE
devices are simpler to configure than the PE devices, or vice-
versa.
- If the overall processing load is a key factor, then 1 is best.
- If the availability of enhanced services support from the SP is
most important, then 4 is best.
As a quick overall conclusion, CE-to-CE encryption provides greater
protection against device compromise but this comes at the cost of
enhanced services and at the cost of operational complexity due to
the Order(n**2) scaling of a larger mesh.
Fang, et al. Informational 23
MPLS/GMPLS Security framework
February 2007
This analysis of site-to-site vs. hop-by-hop encryption tradeoffs
does not explicitly include cases of multiple providers cooperating
to provide a PPVPN service, public Internet VPN connectivity, or
remote access VPN service, but many of the tradeoffs will be
similar.
5.2. Authentication
In order to prevent security issues from some Denial-of-Service
attacks or from malicious misconfiguration, it is critical that
devices in the MPLS/GMPLS should only accept connections or control
messages from valid sources. Authentication refers to methods to
ensure that message sources are properly identified by the
MPLS/GMPLS devices with which they communicate. This section
focuses on identifying the scenarios in which sender authentication
is required, and recommends authentication mechanisms for these
scenarios.
Cryptographic techniques (authentication and encryption) do not
protect against some types of denial of service attacks,
specifically resource exhaustion attacks based on CPU or bandwidth
exhaustion. In fact, the processing required to decrypt and/or
check authentication may in some cases increase the effect of these
resource exhaustion attacks. Cryptographic techniques may however,
be useful against resource exhaustion attacks based on exhaustion
of state information (e.g., TCP SYN attacks).
The MPLS user plane, as presently defined, is not amenable to
source authentication as there are no source identifiers in the
MPLS packet to authenticate. The MPLS label is only locally
meaningful, and identifies a downstream semantic rather than an
upstream source.
When the MPLS payload carries identifiers that may be authenticated
(e.g., IP packets), authentication may be carried out at the client
level, but this does not help the MPLS service provider as these
client identifiers belong to an external non-trusted network.
5.2.1. Management System Authentication
Management system authentication includes the authentication of a
PE to a centrally-managed directory server, when directory-based
Fang, et al. Informational 24
MPLS/GMPLS Security framework
February 2007
"auto-discovery" is used. It also includes authentication of a CE
to the configuration server, when a configuration server system is
used.
5.2.2. Peer-to-peer Authentication
Peer-to-peer authentication includes peer authentication for
network control protocols (e.g. LDP, BGP, etc.), and other peer
authentication (i.e. authentication of one IPsec security gateway
by another).
5.2.3. Cryptographic techniques for authenticating identity
Cryptographic techniques offer several mechanisms for
authenticating the identity of devices or individuals. These
include the use of shared secret keys, one-time keys generated by
accessory devices or software, user-ID and password pairs, and a
range of public-private key systems. Another approach is to use a
hierarchical Certificate Authority system to provide digital
certificates.
This section describes or provides references to the specific
cryptographic approaches for authenticating identity. These
approaches provide secure mechanisms for most of the authentication
scenarios required in securing a MPLS/GMPLS network.
5.3. Access Control techniques
Access control techniques include packet-by-packet or packet-flow-
by-packet-flow access control by means of filters and firewalls, as
well as by means of admitting a "session" for a
control/signaling/management protocol. Enforcement of access
control by isolated infrastructure addresses is discussed in
another section of this document.
In this document, we distinguish between filtering and firewalls
based primarily on the direction of traffic flow. We define
filtering as being applicable to unidirectional traffic, while a
firewall can analyze and control both sides of a conversation.
There are two significant corollaries of this definition:
- Routing or traffic flow symmetry: A firewall typically requires
routing symmetry, which is usually enforced by locating a firewall
where the network topology assures that both sides of a
Fang, et al. Informational 25
MPLS/GMPLS Security framework
February 2007
conversation will pass through the firewall. A filter can operate
upon traffic flowing in one direction, without considering traffic
in the reverse direction.
- Statefulness: Since it receives both sides of a conversation, a
firewall may be able to interpret a significant amount of
information concerning the state of that conversation, and use this
information to control access. A filter can maintain some limited
state information on a unidirectional flow of packets, but cannot
determine the state of the bi-directional conversation as precisely
as a firewall.
5.3.1. Filtering
It is relatively common for routers to filter data packets. That
is, routers can look for particular values in certain fields of the
IP or higher level (e.g., TCP or UDP) headers. Packets which match
the criteria associated with a particular filter may either be
discarded or given special treatment.
In discussing filters, it is useful to separate the Filter
Characteristics which may be used to determine whether a packet
matches a filter from the Packet Actions which are applied to those
packets which match a particular filter.
o Filter Characteristics
Filter characteristics are used to determine whether a particular
packet or set of packets matches a particular filter.
In many cases filter characteristics may be stateless. A stateless
filter is one which determines whether a particular packet matches
a filter based solely on the filter definition, normal forwarding
information (such as the next hop for a packet), and the
characteristics of that individual packet. Typically stateless
filters may consider the incoming and outgoing logical or physical
interface, information in the IP header, and information in higher
layer headers such as the TCP or UDP header. Information in the IP
header to be considered may for example include source and
destination IP address, Protocol field, Fragment Offset, and TOS
field. Filters also may consider fields in the TCP or UDP header
such as the Port fields as well as the SYN field in the TCP header.
Stateful filtering maintains packet-specific state information, to
aid in determining whether a filter has been met. For example, a
device might apply stateless filters to the first fragment of a
fragmented IP packet. If the filter matches, then the data unit ID
may be remembered and other fragments of the same packet may then
be considered to match the same filter. Stateful filtering is more
Fang, et al. Informational 26
MPLS/GMPLS Security framework
February 2007
commonly done in firewalls, although firewall technology may be
added to routers.
o Actions based on Filter Results
If a packet, or a series of packets, matches a specific filter,
then there are a variety of actions which may be taken based on
that filter match. Examples of such actions include:
- Discard
In many cases filters may be set to catch certain undesirable
packets. Examples may include packets with forged or invalid source
addresses, packets which are part of a DOS or DDOS attack, or
packets which are trying to access resources which are not
permitted (such as network management packets from an unauthorized
source). Where such filters are activated, it is common to silently
discard the packet or set of packets matching the filter. The
discarded packets may of course also be counted and/or logged.
- Set CoS
A filter may be used to set the Class of Service associated with
the packet.
- Count packets and/or bytes
- Rate Limit
In some cases the set of packets which match a particular filter
may be limited to a specified bandwidth. In this case packets
and/or bytes would be counted, and would be forwarded normally up
to the specified limit. Excess packets may be discarded, or may be
marked (for example by setting a "discard eligible" bit in the IP
ToS field or the MPLS EXP field).
- Forward and Copy
It is useful in some cases to forward some set of packets normally,
but to also send a copy to a specified other address or interface.
For example, this may be used to implement a lawful intercept
capability, or to feed selected packets to an Intrusion Detection
System.
o Other Issues related to Use of Packet Filters
There may be a very wide variation in the performance impact of
filtering. This may occur both due to differences between
Fang, et al. Informational 27
MPLS/GMPLS Security framework
February 2007
implementations, and also due to differences between types or
numbers of filters deployed. For filtering to be useful, the
performance of the equipment has to be acceptable in the presence
of filters.
The precise definition of "acceptable" may vary from service
provider to service provider, and may depend upon the intended use
of the filters. For example, for some uses a filter may be turned
on all the time in order to set CoS, to prevent an attack, or to
mitigate the effect of a possible future attack. In this case it is
likely that the service provider will want the filter to have
minimal or no impact on performance. In other cases, a filter may
be turned on only in response to a major attack (such as a major
DDOS attack). In this case a greater performance impact may be
acceptable to some service providers.
A key consideration with the use of packet filters is that they can
provide few options for filtering packets carrying encrypted data.
Since the data itself is not accessible, only packet header
information or other unencrypted fields can be used for filtering.
5.3.2. Firewalls
Firewalls provide a mechanism for control over traffic passing
between different trusted zones in the MPLS/GMPLS model, or between
a trusted zone and an untrusted zone. Firewalls typically provide
much more functionality than filters, since they may be able to
apply detailed analysis and logical functions to flows, and not
just to individual packets. They may offer a variety of complex
services, such as threshold-driven denial-of-service attack
protection, virus scanning, acting as a TCP connection proxy, etc.
As with other access control techniques, the value of firewalls
depends on a clear understanding of the topologies of the
MPLS/GMPLS core network, the user networks, and the threat model.
Their effectiveness depends on a topology with a clearly defined
inside (secure) and outside (not secure).
Firewalls may be applied to help protect MPLS/GMPLS core network
functions from attacks originating from the Internet or from
MPLS/GMPLS user sites, but typically other defensive techniques
will be used for this purpose.
Where firewalls are employed as a service to protect user VPN sites
from the Internet, different VPN users, and even different sites of
a single VPN user, may have varying firewall requirements. The
Fang, et al. Informational 28
MPLS/GMPLS Security framework
February 2007
overall PPVPN logical and physical topology, along with the
capabilities of the devices implementing the firewall services,
will have a significant effect on the feasibility and manageability
of such varied firewall service offerings.
Another consideration with the use of firewalls is that they can
provide few options for handling packets carrying encrypted data.
Since the data itself is not accessible, only packet header
information, other unencrypted fields, or analysis of the flow of
encrypted packets can be used for making decisions on accepting or
rejecting encrypted traffic.
5.3.3. Access Control to management interfaces
Most of the security issues related to management interfaces can be
addressed through the use of authentication techniques as described
in the section on authentication. However, additional security may
be provided by controlling access to management interfaces in other
ways.
Management interfaces, especially console ports on MPLS/GMPLS
devices, may be configured so they are only accessible out-of-band,
through a system which is physically and/or logically separated
from the rest of the MPLS/GMPLS infrastructure.
Where management interfaces are accessible in-band within the
MPLS/GMPLS domain, filtering or firewalling techniques can be used
to restrict unauthorized in-band traffic from having access to
management interfaces. Depending on device capabilities, these
filtering or firewalling techniques can be configured either on
other devices through which the traffic might pass, or on the
individual MPLS/GMPLS devices themselves.
5.4. Use of Isolated Infrastructure
One way to protect the infrastructure used for support of
MPLS/GMPLS is to separate the resources for support of MPLS/GMPLS
services from the resources used for other purposes (such as
support of Internet services). In some cases this may make use of
physically separate equipment for VPN services, or even a
physically separate network.
For example, PE-based L3 VPNs may be run on a separate backbone not
connected to the Internet, or may make use of separate edge routers
from those used to support Internet service. Private IP addresses
Fang, et al. Informational 29
MPLS/GMPLS Security framework
February 2007
(local to the provider and non-routable over the Internet) are
sometimes used to provide additional separation.
5.5. Use of Aggregated Infrastructure
In general it is not feasible to use a completely separate set of
resources for support of each service. In fact, one of the main
reasons for MPLS/GMPLS enabled services is to allow sharing of
resources between multiple users, including multiple VPNs, etc.
Thus even if certain services make use of a separate network from
Internet services, nonetheless there will still be multiple
MPLS/GMPLS users sharing the same network resources. In some cases
MPLS/GMPLS services will share the use of network resources with
Internet services or other services.
It is therefore important for MPLS/GMPLS services to provide
protection between resource utilization by different users. Thus a
well-behaved MPLS/GMPLS user should be protected from possible
misbehavior by other users. This requires that limits are placed on
the amount of resources which can be used by any one VPN. For
example, both control traffic and user data traffic may be rate
limited. In some cases or in some parts of the network where a
sufficiently large number of queues are available each VPN (and
optionally each VPN and CoS within the VPN) may make use of a
separate queue. Control-plane resources such as link bandwidth as
well as CPU and memory resources may be reserved on a per-VPN
basis.
The techniques which are used to provision resource protection
between multiple users served by the same infrastructure can also
be used to protect MPLS/GMPLS networks and services from Internet
services.
In general the use of aggregated infrastructure allows the service
provider to benefit from stochastic multiplexing of multiple bursty
flows, and also may in some cases thwart traffic pattern analysis
by combining the data from multiple users.
5.6. Service Provider Quality Control Processes
Deployment of provider-provisioned VPN services in general requires
a relatively large amount of configuration by the service provider.
For example, the service provider needs to configure which VPN each
site belongs to, as well as QoS and SLA guarantees. This large
amount of required configuration leads to the possibility of
misconfiguration.
Fang, et al. Informational 30
MPLS/GMPLS Security framework
February 2007
It is important for the service provider to have operational
processes in place to reduce the potential impact of
misconfiguration. CE to CE authentication may also be used to
detect misconfiguration when it occurs.
5.7. Deployment of Testable MPLS/GMPLS Service.
This refers to solutions that can be readily tested to make sure
they are configured correctly. E.g. for a point-point connection,
checking that the intended connectivity is working pretty much
ensures that there is not connectivity to some unintended site.
6. Monitoring, Detection, and Reporting of Security Attacks
MPLS/GMPLS network and service may be subject to attacks from a
variety of security threats. Many threats are described in another
part of this document. Many of the defensive techniques described
in this document and elsewhere provide significant levels of
protection from a variety of threats. However, in addition to
silently employing defensive techniques to protect against attacks,
MPLS/GMPLS services can also add value for both providers and
customers by implementing security monitoring systems which detect
and report on any security attacks which occur, regardless of
whether the attacks are effective.
Attackers often begin by probing and analyzing defenses, so systems
which can detect and properly report these early stages of attacks
can provide significant benefits.
Information concerning attack incidents, especially if available
quickly, can be useful in defending against further attacks. It
can be used to help identify attackers and/or their specific
targets at an early stage. This knowledge about attackers and
targets can be used to further strengthen defenses against specific
attacks or attackers, or improve the defensive services for
specific targets on an as-needed basis. Information collected on
attacks may also be useful in identifying and developing defenses
against novel attack types.
Monitoring systems used to detect security attacks in MPLS/GMPLS
will typically operate by collecting information from the Provider
Edge (PE), Customer Edge (CE), and/or Provider backbone (P)
devices. Security monitoring systems should have the ability to
actively retrieve information from devices (e.g., SNMP get) or to
passively receive reports from devices (e.g., SNMP notifications).
The specific information exchanged will depend on the capabilities
of the devices and on the type of VPN technology. Particular care
Fang, et al. Informational 31
MPLS/GMPLS Security framework
February 2007
should be given to securing the communications channel between the
monitoring systems and the MPLS/GMPLS devices.
The CE, PE, and P devices should employ efficient methods to
acquire and communicate the information needed by the security
monitoring systems. It is important that the communication method
between MPLS/GMPLS devices and security monitoring systems be
designed so that it will not disrupt network operations. As an
example, multiple attack events may be reported through a single
message, rather than allowing each attack event to trigger a
separate message, which might result in a flood of messages,
essentially becoming a denial-of-service attack against the
monitoring system or the network.
The mechanisms for reporting security attacks should be flexible
enough to meet the needs of MPLS/GMPLS service providers,
MPLS/GMPLS customers, and regulatory agencies, if applicable. The
specific reports will depend on the capabilities of the devices,
the security monitoring system, the type of VPN, and the service
level agreements between the provider and customer.
7. Service Provider General Security Requirements
In this section, we discuss the security requirements that the
provider may have in order to secure its MPLS/GMPLS network
infrastructure, including LDP and RSVP-TE specific requirements.
The MPLS/GMPLS service provider requirements defined here are the
requirements for the MPLS/GMPLS core in the reference model. The
core network can be implemented with different types of network
technologies, and each core network may use different technologies
to provide the various services to users with different levels of
offered security. Therefore, a MPLS/GMPLS service provider may
fulfill any number of the security requirements listed in this
section. This document does not state that a MPLS/GMPLS network
must fulfill all of these requirements to be secure.
These requirements are focused on: 1) how to protect the MPLS/GMPLS
core from various attacks outside the core including network users,
both accidentally and maliciously, 2) how to protect the end users.
7.1. Protection within the Core Network
7.1.1. Control Plane Protection - General
- Protocol authentication within the core:
Fang, et al. Informational 32
MPLS/GMPLS Security framework
February 2007
The network infrastructure must support mechanisms for
authentication of the control plane. In MPLS/GMPLS core is used,
LDP sessions may be authenticated by use TCP MD5, in addition, IGP
and BGP authentication should also be considered. For a core
providing Layer 2 services, PE to PE authentication may also be
used via IPsec.
With the cost of authentication coming down rapidly, the
application of control plane authentication may not increase the
cost of implementation for providers significantly, and will help
to improve the security of the core. If the core is dedicated to
MPLS/GMPLS enabled services and without any interconnects to third
parties then this may reduce the requirement for authentication of
the core control plane.
- Elements protection
Here we discuss means to hide the provider's infrastructure nodes.
A MPLS/GMPLS provider may make the infrastructure routers (P and PE
routers) unreachable from outside users and unauthorized internal
users. For example, separate address space may be used for the
infrastructure loopbacks.
Normal TTL propagation may be altered to make the backbone look
like one hop from the outside, but caution needs to be taken for
loop prevention. This prevents the backbone addresses from being
exposed through trace route; however this must also be assessed
against operational requirements for end to end fault tracing.
An Internet backbone core may be re-engineered to make Internet
routing an edge function, for example, using MPLS label switching
for all traffic within the core and possibly make the Internet a
VPN within the PPVPN core itself. This helps to detach Internet
access from PPVPN services.
Separating control plane, data plane, and management plane
functionality in terms of hardware and software may be implemented
on the PE devices to improve security. This may help to limit the
problems when attacked in one particular area, and may allow each
plane to implement additional security measurement separately.
PEs are often more vulnerable to attack than P routers, since PEs
cannot be made unreachable to outside users by their very nature.
Access to core trunk resources can be controlled on a per user
basis by the application of inbound rate-limiting/shaping, this can
Fang, et al. Informational 33
MPLS/GMPLS Security framework
February 2007
be further enhanced on a per Class of Service basis (see section
8.2.3)
In the PE, using separate routing processes for different services,
for example, Internet and PPVPN service may help to improve the
PPVPN security and better protect VPN customers. Furthermore, if
the resources, such as CPU and Memory, may be further separated
based on applications, or even individual VPNs, it may help to
provide improved security and reliability to individual VPN
customers.
7.1.2. Control plane protection with RSVP-TE
- RSVP Security Tools
Isolation of the trusted domain is an important security mechanism
with respect to RSVP, to ensure that an untrusted element cannot
access a router of the trusted domain. Though isolation is limited
by the need to allow ASBR-ASBR communication for inter-AS LSPs.
Isolation mechanisms might be bypassed by Router Alert IP packets.
- A solution would consists in disabling the RSVP router alert mode
and dropping all IP packets with the router alert option, or also
to drop on an interface all incoming IP packets with port 46, which
requires an access-list at the IP port level) or spoofed IP packets
if anti-spoofing is not activated.
RSVP security can be strengthened by deactivating RSVP on
interfaces with neighbors who are not authorized to use RSVP, to
protect against adjacent CE-PE attacks. However, this does not
really protect against DoS attacks, and does not protect against
attacks on non-adjacent routers. It has been demonstrated that
substantial CPU resources are consumed simply by processing
received RSVP packets, even if the RSVP process is deactivated for
the specific interface on which the RSVP message is received.
RSVP neighbor filtering at the protocol level, to restrict the set
of neighbors that can send RSVP messages to a given router,
protects against non-adjacent attacks. However, this does not
protect against DoS attacks, and does not effectively protect
against spoofing of the source address of RSVP packets, if the
filter relies on the neighbor's address within the RSVP message.
RSVP neighbor filtering at the data plane level (access list to
accept IP packet with port 46, only for specific neighbors). This
requires Router Alert mode to be deactivated. This does not protect
against spoofing.
- Authentication for RSVP messages
Fang, et al. Informational 34
MPLS/GMPLS Security framework
February 2007
One of the most powerful tools for protection against RSVP-based
attacks is the use of authentication for RSVP messages, based on a
secure message hash using a key shared by RSVP neighbors. This
protects against LSP creation attacks, at the expense of consuming
significant CPU resources for digest computation. In addition, if
the neighboring RSVP speaker is compromised, it could be used to
launch attacks using authenticated RSVP messages.
Another valuable tool is RSVP message pacing, to limit the number
of RSVP messages sent to a given neighbor during a given period.
This allows blocking DoS attack propagation.
In order to ensure continued effective operation of the MPLS router
even in the case of an attack which is able to bypass packet
filtering mechanisms such as Access Control Lists in the data
plane, it is important that routers have some mechanisms to limit
the impact of the attack. There should be a mechanism to rate
limit the amount of control plane traffic addressed to the router,
per interface. This should be configurable on a per-protocol
basis, (and, ideally, on a per sender basis) to avoid an attacked
protocol, or a given sender blocking all communications. This
requires the ability to filter and limit the rate of incoming
messages of particular protocols, such as RSVP (filtering at the IP
port level), and particular senders). In addition, there should be
a mechanism to limit CPU and memory capacity allocated to RSVP, so
as to protect other control plane elements. In order to limit the
memory allocation, it will probably be necessary to limit the
number of LSPs which can be set up.
- limit the impact of an attack on control plane resources
In order to ensure continued effective operation of the MPLS router
even in the case of an attack which is able to bypass packet
filtering mechanisms such as Access Control Lists in the data
plane, it is important that routers have some mechanisms to limit
the impact of the attack. There should be a mechanism to rate
limit the amount of control plane traffic addressed to the router,
per interface. This should be configurable on a per-protocol
basis, (and, ideally, on a per sender basis) to avoid an attacked
protocol, or a given sender blocking all communications. This
requires the ability to filter and limit the rate of incoming
messages of particular protocols, such as RSVP (filtering at the IP
port level, and particular senders). In addition, there should be
a mechanism to limit CPU and memory capacity allocated to RSVP, so
as to protect other control plane elements. In order to limit the
memory allocation, it will probably be necessary to limit the
number of LSPs which can be set up.
Fang, et al. Informational 35
MPLS/GMPLS Security framework
February 2007
7.1.3. Control plane protection with LDP
The approaches to protect MPLS routers against LDP-based attacks
are very similar to those for RSVP, including isolation, protocol
deactivation on specific interfaces, filtering of LDP neighbors at
the protocol level, filtering of LDP neighbors at the data plane
level (access list that filter the TCP & UDP LDP ports),
authentication with message digest, rate limiting of LDP messages
per protocol per sender and limiting all resources which might be
allocated to LDP-related tasks.
7.1.4. Data Plane Protection
IPsec technologies can provide - encryption of secure provider or
user data.
In today's MPLS/GMPLS, ATM, or Frame Relay networks, encryption is
not provided as a basic feature. Mechanisms described in section 5
can be used to secure the MPLS data plane to secure the data
carried over MPLS core.
7.2. Protection on the User Access Link
Peer / Neighbor protocol authentication may be used to enhance
security. For example, BGP MD5 authentication may be used to
enhance security on PE-CE links using eBGP. In the case of Inter-
provider connection, authentication / encryption mechanisms between
ASes, such as IPsec, may be used.
WAN link address space separation for different services (e.g. VPN
and non-VPN) may be implemented to improve security in order to
protect each service if multiple services are provided on the same
PE platform.
Firewall / Filtering: access control mechanisms can be used to
filter out any packets destined for the service provider's
infrastructure prefix or eliminate routes identified as
illegitimate routes.
Rate limiting may be applied to the user interface/logical
interfaces against DDOS bandwidth attack. This is very helpful when
the PE device is supporting both multi-services, especially when
supporting VPN and Internet Services on the same physical
interfaces through different logical interfaces.
Fang, et al. Informational 36
MPLS/GMPLS Security framework
February 2007
7.2.1. Link Authentication
Authentication mechanisms can be employed to validate site access
to the network via fixed or logical (e.g. L2TP, IPsec) connections.
Where the user wishes to hold the 'secret' associated to acceptance
of the access and site into the VPN, then provider solutions
require the flexibility for either direct authentication by the PE
itself or interaction with a customer authentication server.
Mechanisms are required in the latter case to ensure that the
interaction between the PE and the customer authentication server
is controlled e.g. limiting it simply to an exchange in relation to
the authentication phase and with other attributes e.g. RADIUS
optionally being filtered.
7.2.2. Access Routing
Mechanisms may be used to provide control at a routing protocol
level e.g. RIP, OSPF, BGP between the CE and PE. Per neighbor and
per VPN routing policies may be established to enhance security and
reduce the impact of a malicious or non-malicious attack on the PE,
in particular the following mechanisms should be considered:
- Limiting the number of prefixes that may be advertised on
a per access basis into the PE. Appropriate action may be
taken should a limit be exceeded e.g. the PE shutting
down the peer session to the CE
- Applying route dampening at the PE on received routing
updates
- Definition of a per VPN prefix limit after which
additional prefixes will not be added to the VPN routing
table.
In the case of Inter-provider connection, access protection, link
authentication, and routing policies as described above may be
applied. Both inbound and outbound firewall/filtering mechanism
between ASes may be applied. Proper security procedures must be
implemented in Inter-provider VPN interconnection to protect the
providers' network infrastructure and their customer VPNs. This may
be custom designed for each Inter-Provider VPN peering connection,
and must be agreed by both providers.
7.2.3. Access QoS
MPLS/GMPLS providers offering QoS enabled services require
mechanisms to ensure that individual accesses are validated against
their subscribed QOS profile and as such gain access to core
resources that match their service profile. Mechanisms such as per
Class of service rate limiting/traffic shaping on ingress to the
MPLS/GMPLS core are one option in providing this level of control.
Fang, et al. Informational 37
MPLS/GMPLS Security framework
February 2007
Such mechanisms may require the per Class of Service profile to be
enforced either by marking, remarking or discard of traffic outside
of profile.
7.2.4. Customer service monitoring tools
End users requiring visibility of the specific statistics on the
core e.g. routing table, interface status, QoS statistics, impose
requirements for mechanisms at the PE to both validate the incoming
user and limit the views available to that particular user.
Mechanisms should also be considered to ensure that such access
cannot be used a means of a DOS attack (either malicious or
accidental) on the PE itself. This could be accomplished through
either separation of these resources within the PE itself or via
the capability to rate-limit on a per physical/logical connection
basis such traffic.
7.3. General Requirements for MPLS/GMPLS Providers
The MPLS/GMPLS providers must support the users' security
requirements as listed in Section 7. Depending on the technologies
used, these requirements may include:
- User control plane separation - routing isolation
- Protection against intrusion, DOS attacks and spoofing
- Access Authentication
- Techniques highlighted through this document identify
methodologies for the protection of resources and
MPLS/GMPLS infrastructure.
Equipment hardware/software bugs leading to breaches in security
are not within the scope of this document.
8. Inter-provider Security Requirements
This section discusses security capabilities that are important at
the MPLS/GMPLS Inter-provider connections, and at devices
(including ASBR routers) which support the Inter-provider
connections. The security capabilities stated in this section
should be considered as complementary to security considerations
addressed in the individual protocol specifications and/or security
frameworks.
Security vulnerabilities and exposures may be propagated across
multiple networks because of security vulnerabilities arising in
one peer's network. Threats to security originate from accidental,
Fang, et al. Informational 38
MPLS/GMPLS Security framework
February 2007
administrative and intentional sources. Intentional threats include
events such as spoofing and Denial of Service (DoS) attacks.
The level and nature of threats, as well as security and
availability requirements, may vary over time and from network to
network. This section therefore discusses capabilities that need to
be available in equipment deployed for support of the MPLS-ICI.
Whether any particular capability is used in any one specific
instance of the ICI is up to the service providers managing the
provider edge equipment offering/using the ICI services.
8.1. Control Plane Protection
This section discusses capabilities for control plane protection,
including protection of routing, signaling, and OAM capabilities.
8.1.1. Authentication of Signaling Sessions
Authentication of signaling sessions (i.e., BGP, LDP and RSVP-TE)
and routing sessions (e.g., BGP) as well as OAM sessions across
domain boundaries. Equipment must be able to support exchange of
all protocol messages over a single IPsec tunnel, with NULL
encryption and authentication, between the peering ASBRs. Support
for TCP MD5 authentication for LDP and BGP and for RSVP-TE
authentication must also be provided.
Mechanisms to authenticate and validate a dynamic setup request
MUST be available. For instance, if dynamic signaling of a TE-LSP
or PW is crossing a domain boundary, there must be a way to detect
whether the LSP source is who he claims to be and that he is
allowed to connect to the destination.
MD5 authentication support for all TCP-based protocols within the
scope of the MPLS-ICI (i.e., LDP signaling, and BGP routing) and
MD5 authentication for the RSVP-TE Integrity Object MUST be
provided to interoperate with current practices.
Equipment SHOULD be able to support exchange of all signaling and
routing (LDP, RSVP-TE, and BGP) protocol messages over a single
IPSec in tunnel or transport mode with authentication but with NULL
encryption, between the peering ASBRs. IPSec, if supported, must be
supported with HMAC-MD-5 and optionally SHA-1. It is expected that
authentication algorithms will evolve over time and support can be
updated as needed.
OAM Operations across the MPLS-ICI could also be the source of
security threats on the provider infrastructure as well as the
service offered over the MPLS-ICI. A large volume of OAM messages
could overwhelm the processing capabilities of an ASBR if the ASBR
is not probably protected. Maliciously-generated OAM messages could
Fang, et al. Informational 39
MPLS/GMPLS Security framework
February 2007
also be used to bring down an otherwise healthy service (e.g., MPLS
Pseudo Wire), and therefore effecting service security. MPLS-ping
does not support authentication today and that support should be
subject for future considerations. Bidirectional Forwarding
Detection (BFD) however, does have support for carrying an
authentication object. It also supports Time-To-Live (TTL)
processing as anti-replay measure. Implementations conformant to
this MPLS-ICI should support BFD authentication using MD-5 and must
support the procedures for TTL processing.
8.1.2. Protection against DoS attacks in the Control Plane
Ability to prevent signaling and routing DOS attacks on the control
plane per interface and provider. Such prevention may be provided
by rate-limiting signaling and routing messages that can be sent by
a peer provider according to a traffic profile and by guarding
against malformed packets.
Equipment MUST provide the ability to filter signaling, routing,
and OAM packets destined for the device, and MUST provide the
ability to rate limit such packets. Packet filters SHOULD be
capable of being separately applied per interface, and SHOULD have
minimal or no performance impact. For example, this allows an
operator to filter or rate-limit signaling, routing, and OAM
messages that can be sent by a peer provider and limit such traffic
to a traffic profile.
In the presence of a control plane DoS attack against an ASBR, the
router SHOULD guarantee sufficient resources to allow network
operators to execute network management commands to take corrective
action, such as turning on additional filters or disconnecting an
interface which is under attack. DoS attacks on the control plane
SHOULD NOT adversely affect data plane performance.
Equipment which supports BGP MUST support the ability to limit the
number of BGP routes received from any particular peer.
Furthermore, in the case of IPVPN, a router MUST be able to limit
the number of routes learned from a BGP peer per IPVPN. In the case
that a device has multiple BGP peers, it SHOULD be possible for the
limit to vary between peers.
8.1.3. Protection against Malformed Packets
Equipment SHOULD be robust in the presence of malformed protocol
packets. For example, malformed routing, signaling, and OAM packets
should be treated in accordance to the relevant protocol
specification.
Fang, et al. Informational 40
MPLS/GMPLS Security framework
February 2007
8.1.4. Ability to Enable/Disable Specific Protocols
Ability to drop any signaling or routing protocol messages when
these messages are to be processed by the ASBR but the
corresponding protocol is not enabled on that interface.
Equipment must allow an administrator to enable or disable a
protocol (default protocol is disabled unless administratively
enable) on an interface basis.
Equipment MUST be able to drop any signaling or routing protocol
messages when these messages are to be processed by the ASBR but
the corresponding protocol is not enabled on that interface. This
dropping SHOULD NOT adversely affect data plane or control plane
performance.
8.1.5. Protection Against Incorrect Cross Connection
Capability of detecting and locating faults in an LSP cross-connect
MUST be provided. Such faults cause security violations as they
result in directing traffic to the wrong destinations. This
capability may rely on OAM functions.
Equipment MUST support MPLS LSP Ping [RFC4379]. This MAY be used to
verify end to end connectivity for the LSP (e.g., PW, TE Tunnel,
VPN LSP, etc), and to verify PE to PE connectivity for L3 VPN
services.
When routing information is advertised from one domain to the
other, there MUST be mechanisms that enable operators to guard
against situations that result in traffic hijacking, black-holing,
resource stealing (e.g., number of routes), etc. For instance, in
the IPVPN case, an operator must be able to block routes based on
associated route target attributes. In addition, mechanisms must
exist to verify whether a route advertised by a peer for a given
VPN is actually a valid route and whether the VPN has a site
attached or reachable through that domain.
Equipment (ASBRs and RRs) which supports operation of BGP MUST
allow a means to restrict which Route Target attributes are sent to
and accepted from a BGP peer across an ICI. Equipment (ASBRs, RRs)
SHOULD also be able to inform the peer regarding which Route Target
attributes it will accept from the peer. This is due to the fact
that a peer which sends an incorrect Route Target can result in
incorrect cross-connection of VPNs. Also, sending inappropriate
route targets to a peer may disclose confidential information.
Further Security Consideration for inter-provider BGP/MPLS IPVPN
operations are discussed in the IPVPN Annex.
Fang, et al. Informational 41
MPLS/GMPLS Security framework
February 2007
8.1.6. Protection Against Spoofed Updates and Route
Advertisements
Equipment MUST support signaling and routing.
Equipment MUST support route filtering of routes received via a BGP
peer sessions by applying policies that include one or more the
following: AS path, BGP next hop, standard community and/or
extended community.
8.1.7. Protection of Confidential Information
Ability to identify and prohibit messages that can reveal
confidential information about network operation (e.g., performance
OAM messages, MPLS-ping messages). Service Providers must have the
flexibility of handling these messages at the ASBR.
Equipment SHOULD provide the ability to identify and prohibit
messages that can reveal confidential information about network
operation (e.g., performance OAM messages, LSP Traceroute
messages). Service Providers must have the flexibility of handling
these messages at the ASBR. For example, equipment supporting LSP
Traceroute MAY limit which addresses replies can be sent to.
Note: This capability should be used with care. For example, if a
service provider chooses to prohibit the exchange of LSP PING
messages at the ICI, it may make it more difficult to debug
incorrect cross-connection of LSPs or other problems.
A provider may decide to progress these messages if they are
incoming from a trusted provider and are targeted to specific
agreed-on addresses. Another provider may decide to traffic police,
reject or apply policies to these messages. Solutions must enable
providers to control the information that is relayed to another
provider about the path that an LSP takes. For example, in RSVP-TE
record route object or MPLS-ping trace, a provider must be able to
control the information contained in corresponding messages when
sent to another provider.
8.1.8. Protection Against over-provisioned number of RSVP-TE
LSPs and bandwidth reservation
In addition to the control plane protection mechanisms listed in
the previous section on Control plane protection with RSVP-TE, the
ASBR needs mechanisms to both limit the number of LSPs that can be
set up by other domains and to limit the amount of bandwidth that
can be reserved. A provider's ASBR may deny the LSPs set up request
or the bandwidth reservation request sent by another provider's the
limits are reached.
Fang, et al. Informational 42
MPLS/GMPLS Security framework
February 2007
8.2. Data Plane Protection
8.2.1. Protection against DoS in the Data Plane
This is provided earlier in this document.
8.2.2. Protection against Label Spoofing
Verification that a label received across an interconnect was
actually assigned to the provider across the interconnect. If the
label was not assigned to the provider, the packet MUST be dropped.
Equipment MUST be able to verify that a label received across an
interconnect was actually assigned to an LSP arriving from the
provider across that interconnect. If the label was not assigned to
an LSP which arrives at this router from the correct neighboring
provider, the packet MUST be dropped. This verification can be
applied to the top label only. The top label is the received top
label and every label that is exposed by label popping to be used
for forwarding decisions.
Equipment MUST provide the capability of dropping MPLS-labeled
packets if all labels in the stack are not processed. This
provides carriers the capability of guaranteeing that every label
that enters its domain from another carrier was actually assigned
to that carrier.
The following requirements are not directly reflected in this
document but must be used as guidance for addressing further work.
Solutions MUST NOT force operators to reveal reachability
information to routers within their domains. <note, it is believed
that this requirement is met via other requirements specified in
this section plus the normal operation of IP routing, which does
not reveal individual hosts.
Mechanisms to authenticate and validate a dynamic setup request
MUST be available. For instance, if dynamic signaling of a TE-LSP
or PW is crossing a domain boundary, there must be a way to detect
whether the LSP source is who he claims to be and that he is
allowed to connect to the destination.
8.2.3. Protection using ingress traffic policing and enforcement
In the following diagram, we use a simple diagram to illustrate a
potential security issue on the data plane issue across the MPLS
interconnect:
SP2 - ASBR2 - labeled path - ASBR1 - P1 - SP1's PSN - P2 - PE1
Fang, et al. Informational 43
MPLS/GMPLS Security framework
February 2007
| | | |
|< AS2 >|<MPLS interconnect>|< AS1 >|
Traffic flow direction is from SP2 to SP1
Usually, the transit label used by ASBR2 is allocated by ASBR1
which in turn advertises to ASB2 (downstream unsolicited or on-
demand) and this label is used for a service context (VPN label, PW
VC label, etc.) and this LSP is normally terminated at a forwarding
table belonging to the service instance on PE (PE1) in SP1.
In the example above, ASBR1 would not know if the label of an
incoming packet from ASBR2 over the interconnect is VPN label or
PSN label for AS1. So it is possible (though rare) that ASBR2 can
be tempered such that the incoming label could match a PSN label
(e.g., LDP) in AS1 - then this LSP would end up on the global plane
of an infrastructure router (P or PE1) - this could invite a
unidirectional attack on that P or PE1 the LSP terminates.
To mitigate this threat, we SHOULD be able to do a forwarding path
look-up for the label on an incoming packet from a interconnect in
a LFIB space that is only intended for its own service context or
provide a mechanism on the data plane that would ensure the
incoming labels are what ASBR1 has allocated and advertised.
Similar concept has been proposed in "Requirements for Multi-
Segment Pseudowire Emulation Edge-to-Edge (PWE3)" [PW-REQ].
9. Security Considerations
Security considerations constitute the sole subject of this memo
and hence are discussed throughout. Here we recap what has been
presented and explain at a very high level the role of each type of
consideration in an overall secure MPLS/GMPLS system.
The document describes a number of potential security threats.
Some of these threats have already been observed occurring in
running networks; others are largely theoretical at this time. DOS
attacks and intrusion
Attacks from the Internet against service provider infrastructure
have been seen to occur. DOS "attacks" (typically not malicious)
have also been seen in which CE equipment overwhelms PE equipment
with high quantities or rates of packet traffic or routing
Fang, et al. Informational 44
MPLS/GMPLS Security framework
February 2007
information. Operational/provisioning errors are cited by service
providers as one of their prime concerns.
The document describes a variety of defensive techniques that may
be used to counter the suspected threats. All of the techniques
presented involve mature and widely implemented technologies that
are practical to implement.
The document describes the importance of detecting, monitoring, and
reporting attacks, both successful and unsuccessful. These
activities are essential for "understanding one's enemy",
mobilizing new defenses, and obtaining metrics about how secure the
MPLS/GMPLS network is. As such they are vital components of any
complete PPVPN security system.
The document evaluates MPLS/GMPLS security requirements from a
customer perspective as well as from a service provider
perspective. These sections re-evaluate the identified threats
from the perspectives of the various stakeholders and are meant to
assist equipment vendors and service providers, who must ultimately
decide what threats to protect against in any given equipment or
service offering.
10. IANA Considerations
TBD.
11. Normative References
[RFC3031] E. Rosen, A. Viswanathan, R. Callon, "Multiprotocol Label
Switching Architecture", RFC 3031, January 2001.
[RFC3945] E. Mannie, "Generalized Multi-Protocol Label Switching
(GMPLS) Architecture", RFC 3945, October 2004.
[RFC3036] Andersson, et al., "LDP Specification", January 2001.
[RFC3209] Awduche, et al., "RSVP-TE: Extensions to RSVP for LSP
Tunnels", December 2001.
[RFC4301] S. Kent, K. Seo, "Security Architecture for the Internet
Protocol," December 2005.
[RFC4302] S. Kent, "IP Authentication Header," December 2005.
Fang, et al. Informational 45
MPLS/GMPLS Security framework
February 2007
[RFC4305] D. Eastlake 3rd, "Cryptographic Algorithm Implementation
Requirements for Encapsulating Security Payload (ESP) and
Authentication Header (AH)", December 2005.
[RFC4306] C. Kaufman, "Internet Key Exchange (IKEv2)
Protocol",December 2005.
[RFC4346] T. Dierks and E. Rescorla, "The Transport Layer Security
(TLS) Protocol, Version 1.1," April 2006.
[RFC4379] K. Kompella and G. Swallow, "Detecting Multi-Protocol
Label Switched (MPLS) Data Plane Failures", February 2006.
[RFC4447] Martini, et al., "Pseudowire Setup and Maintenance Using
the Label Distribution Protocol (LDP)", April 2006.
[STD62] "Simple Network Management Protocol, Version 3," RFCs 3411-
3418, December 2002.
[STD-8] J. Postel and J. Reynolds, "TELNET Protocol Specification",
STD 8, May 1983.
12. Informational References
[AES-CCM] Housley, R., "Using AES CCM Mode With IPsec ESP", draft-
ietf-ipsec-ciph-aes-ccm-05.txt, work in progress, November 2003.
[RFC 2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997
[Beard] D. Beard and Y. Yang, "Known Threats to Routing Protocols,"
draft-beard-rpsec-routing-threats-00.txt, Oct. 2002. (Note, this is
now approved as RFC, no number yet, http://www.ietf.org/internet-
drafts/draft-ietf-rpsec-routing-threats-06.txt.
[RFC2104] H. Krawczyk, M. Bellare, R. Canetti, "HMAC: Keyed-Hashing
for Message Authentication," February 1997.
[RFC2411] R. Thayer, N. Doraswamy, R. Glenn, "IP Security Document
Roadmap," November 1998.
[RFC3174] D. Eastlake, 3rd, and P. Jones, "US Secure Hash Algorithm
1 (SHA1)," September 2001.
Fang, et al. Informational 46
MPLS/GMPLS Security framework
February 2007
[RFC3985] S. Bryant and P. Pate, "Pseudo Wire Emulation Edge-to-
Edge (PWE3) Architecture", March 2005.
[RFC4111] L. Fang, "Security Framework of Provider Provisioned
VPN", RFC 4111, July 2005.
[RFC3631] S. Bellovin, C. Kaufman, J. Schiller, "Security
Mechanisms for the Internet," December 2003.
[RFC4110] R. Callon and M. Suzuki, "A Framework for Layer 3
Provider-Provisioned Virtual Private Networks (PPVPNs), July 2005.
[MFA MPLS ICI] N. Bitar, "MPLS InterCarrier Interconnect Technical
Specification", MFA2006.109.01, August 2006.
[opsec efforts] C. Lonvick and D. Spak, "Security Best Practices
Efforts and Documents", draft-ietf-opsec-efforts-05.txt, December
2006.
[PW-REQ] N. Bitar, M. Bocci, L. Martini, "Requirements for Multi-
Segment Pseudowire Emulation Edge-to-Edge", draft-ietf-pwe3-ms-pw-
requirements-04.txt.
13. Author's Addresses
Luyuan Fang
Cisco Systems, Inc.
300 Beaver Brook Road
Boxborough, MA 01719
USA
EMail: lufang@cisco.com
Michael Behringer
Cisco Systems, Inc.
Village d'Entreprises Green Side
400, Avenue Roumanille, Batiment T 3
06410 Biot, Sophia Antipolis
FRANCE
Email: mbehring@cisco.com
Ross Callon
Juniper Networks
10 Technology Park Drive
Westford, MA 01886
Fang, et al. Informational 47
MPLS/GMPLS Security framework
February 2007
USA
Email: rcallon@juniper.net
Jean-Louis Le Roux
France Telecom
2, avenue Pierre-Marzin
22307 Lannion Cedex
FRANCE
Email: jeanlouis.leroux@francetelecom.com
Raymond Zhang
British Telecom
2160 E. Grand Ave. El Segundo, CA 90025
USA
Email: raymond.zhang@bt.com
Paul Knight
Nortel
600 Technology Park Drive
Billerica, MA 01821
EMail: paul.knight@nortel.com
Yaakov (Jonathan) Stein
RAD Data Communications
24 Raoul Wallenberg St., Bldg C
Tel Aviv 69719
ISRAEL
Email: yaakov_s@rad.com
Intellectual Property
The IETF takes no position regarding the validity or scope of any
Intellectual Property Rights or other rights that might be claimed
to pertain to the implementation or use of the technology described
in this document or the extent to which any license under such
rights might or might not be available; nor does it represent that
it has made any independent effort to identify any such rights.
Information on the procedures with respect to rights in RFC
documents can be found in BCP 78 and BCP 79.
Fang, et al. Informational 48
MPLS/GMPLS Security framework
February 2007
Copies of IPR disclosures made to the IETF Secretariat and any
assurances of licenses to be made available, or the result of an
attempt made to obtain a general license or permission for the use
of such proprietary rights by implementers or users of this
specification can be obtained from the IETF on-line IPR repository
at http://www.ietf.org/ipr.
The IETF invites any interested party to bring to its attention any
copyrights, patents or patent applications, or other proprietary
rights that may cover technology that may be required to implement
this standard. Please address the information to the IETF at ietf-
ipr@ietf.org.
Full Copyright Statement
Copyright (C) The IETF Trust (2007).
This document is subject to the rights, licenses and restrictions
contained in BCP 78, and except as set forth therein, the authors
retain all their rights.
Disclaimer
This document and the information contained herein are provided on
an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE
REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE
IETF TRUST AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM 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.
14. Acknowledgement
Funding for the RFC Editor function is provided by the IETF
Administrative Support Activity (IASA).
Fang, et al. Informational 49