Request for Comments: 4381 Cisco Systems Inc
Category: Informational February 2006
Analysis of the Security of BGP/MPLS IP
Virtual Private Networks (VPNs)
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.
Copyright Notice
Copyright (C) The Internet Society (2006).
IESG Note
The content of this RFC was at one time considered by the IETF, and
therefore it may resemble a current IETF work in progress or a
published IETF work. This RFC is not a candidate for any level of
Internet Standard. The IETF disclaims any knowledge of the fitness
of this RFC for any purpose, and in particular notes that the
decision to publish is not based on IETF review for such things as
security, congestion control or inappropriate interaction with
deployed protocols. The RFC Editor has chosen to publish this
document at its discretion. Readers of this RFC should exercise
caution in evaluating its value for implementation and deployment.
See RFC 3932 for more information.
Abstract
This document analyses the security of the BGP/MPLS IP virtual
private network (VPN) architecture that is described in RFC 4364, for
the benefit of service providers and VPN users.
The analysis shows that BGP/MPLS IP VPN networks can be as secure as
traditional layer-2 VPN services using Asynchronous Transfer Mode
(ATM) or Frame Relay.
Table of Contents
1. Scope and Introduction ..........................................3
2. Security Requirements of VPN Networks ...........................4
2.1. Address Space, Routing, and Traffic Separation .............4
2.2. Hiding the Core Infrastructure .............................5
2.3. Resistance to Attacks ......................................5
2.4. Impossibility of Label Spoofing ............................6
3. Analysis of BGP/MPLS IP VPN Security ............................6
3.1. Address Space, Routing, and Traffic Separation .............6
3.2. Hiding of the BGP/MPLS IP VPN Core Infrastructure ..........7
3.3. Resistance to Attacks ......................................9
3.4. Label Spoofing ............................................11
3.5. Comparison with ATM/FR VPNs ...............................12
4. Security of Advanced BGP/MPLS IP VPN Architectures .............12
4.1. Carriers’ Carrier .........................................13
4.2. Inter-Provider Backbones ..................................14
5. What BGP/MPLS IP VPNs Do Not Provide ...........................16
5.1. Protection against Misconfigurations of the Core
and Attacks ’within’ the Core .............................16
5.2. Data Encryption, Integrity, and Origin Authentication .....17
5.3. Customer Network Security .................................17
6. Layer 2 Security Considerations ................................18
7. Summary and Conclusions ........................................19
8. Security Considerations ........................................20
9. Acknowledgements ...............................................20
10. Normative References ..........................................20
11. Informative References ........................................20
1. Scope and Introduction
As Multiprotocol Label Switching (MPLS) is becoming a more widespread
technology for providing IP virtual private network (VPN) services,
the security of the BGP/MPLS IP VPN architecture is of increasing
concern to service providers and VPN customers. This document gives
an overview of the security of the BGP/MPLS IP VPN architecture that
is described in RFC 4364 [1], and compares it with the security of
traditional layer-2 services such as ATM or Frame Relay.
The term "MPLS core" is defined for this document as the set of
Provider Edge (PE) and provider (P) routers that provide a BGP/MPLS
IP VPN service, typically under the control of a single service
provider (SP). This document assumes that the MPLS core network is
trusted and secure. Thus, it does not address basic security
concerns such as securing the network elements against unauthorised
access, misconfigurations of the core, or attacks internal to the
core. A customer that does not wish to trust the service provider
network must use additional security mechanisms such as IPsec over
the MPLS infrastructure.
This document analyses only the security features of BGP/MPLS IP
VPNs, not the security of routing protocols in general. IPsec
technology is also not covered, except to highlight the combination
of MPLS VPNs with IPsec.
The overall security of a system has three aspects: the architecture,
the implementation, and the operation of the system. Security issues
can exist in any of these aspects. This document analyses only the
architectural security of BGP/MPLS IP VPNs, not implementation or
operational security issues.
This document is targeted at technical staff of service providers and
enterprises. Knowledge of the basic BGP/MPLS IP VPN architecture as
described in RFC 4364 [1] is required to understand this document.
For specific Layer 3 VPN terminology and reference models refer to
[11].
Section 2 of this document specifies the typical VPN requirements a
VPN user might have, and section 3 analyses how RFC 4364 [1]
addresses these requirements. Section 4 discusses specific security
issues of multi-AS (Autonomous System) MPLS architectures, and
section 5 lists security features that are not covered by this
architecture and therefore need to be addressed separately. Section
6 highlights potential security issues on layer 2 that might impact
the overall security of a BGP/MPLS IP VPN service. The findings of
this document are summarized in section 7.
2. Security Requirements of VPN Networks
Both service providers offering any type of VPN services and
customers using them have specific demands for security. Mostly,
they compare MPLS-based solutions with traditional layer 2-based VPN
solutions such as Frame Relay and ATM, since these are widely
deployed and accepted. This section outlines the typical security
requirements for VPN networks. The following section discusses if
and how BGP/MPLS IP VPNs address these requirements, for both the
MPLS core and the connected VPNs.
2.1. Address Space, Routing, and Traffic Separation
Non-intersecting layer 3 VPNs of the same VPN service are assumed to
have independent address spaces. For example, two non-intersecting
VPNs may each use the same 10/8 network addresses without conflict.
In addition, traffic from one VPN must never enter another VPN. This
implies separation of routing protocol information, so that routing
tables must also be separate per VPN. Specifically:
o Any VPN must be able to use the same address space as any other
VPN.
o Any VPN must be able to use the same address space as the MPLS
core.
o Traffic, including routing traffic, from one VPN must never flow
to another VPN.
o Routing information, as well as distribution and processing of
that information, for one VPN instance must be independent from
any other VPN instance.
o Routing information, as well as distribution and processing of
that information, for one VPN instance must be independent from
the core.
From a security point of view, the basic requirement is to prevent
packets destined to a host a.b.c.d within a given VPN reaching a host
with the same address in another VPN or in the core, and to prevent
routing packets to another VPN even if it does not contain that
destination address.
Confidentiality, as defined in the L3VPN Security Framework [11], is
a requirement that goes beyond simple isolation of VPNs and provides
protection against eavesdropping on any transmission medium.
Encryption is the mechanism used to provide confidentiality. This
document considers confidentiality an optional VPN requirement, since
many existing VPN deployments do not encrypt transit traffic.
2.2. Hiding the Core Infrastructure
The internal structure of the core network (MPLS PE and P elements)
should not be externally visible. Whilst breaking this requirement
is not a security problem in itself, many service providers believe
it is advantageous if the internal addresses and network structure
are hidden from the outside world. An argument is that denial-of-
service (DoS) attacks against a core router are much easier to carry
out if an attacker knows the router addresses. Addresses can always
be guessed, but attacks are more difficult if addresses are not
known. The core should be as invisible to the outside world as a
comparable layer 2 infrastructure (e.g., Frame Relay, ATM). Core
network elements should also not be accessible from within a VPN.
Security should never rely entirely on obscurity, i.e., the hiding of
information. Services should be equally secure if the implementation
is known. However, there is a strong market perception that hiding
of details is advantageous. This point addresses that market
perception.
2.3. Resistance to Attacks
There are two basic types of attacks: DoS attacks, where resources
become unavailable to authorised users, and intrusions, where
resources become available to unauthorised users. BGP/MPLS IP VPN
networks must provide at least the same level of protection against
both forms of attack as current layer 2 networks.
For intrusions, there are two fundamental ways to protect the
network: first, to harden protocols that could be abused (e.g.,
Telnet into a router), and second, to make the network as
inaccessible as possible. This is achieved by a combination of
packet filtering / firewalling and address hiding, as discussed
above.
DoS attacks are easier to execute, since a single known IP address
might be enough information to attack a machine. This can be done
using normal "permitted" traffic, but using higher than normal packet
rates, so that other users cannot access the targeted machine. The
only way to be invulnerable to this kind of attack is to make sure
that machines are not reachable, again by packet filtering and
optionally by address hiding.
This document concentrates on protecting the core network against
attacks from the "outside", i.e., the Internet and connected VPNs.
Protection against attacks from the "inside", i.e., an attacker who
has logical or physical access to the core network, is not discussed
here.
2.4. Impossibility of Label Spoofing
Assuming the address and traffic separation discussed above, an
attacker might try to access other VPNs by inserting packets with a
label that he does not "own". This could be done from the outside,
i.e., another Customer Edge (CE) router or from the Internet, or from
within the MPLS core. The latter case (from within the core) will
not be discussed, since we assume that the core network is provided
securely. Should protection against an insecure core be required, it
is necessary to use security protocols such as IPsec across the MPLS
infrastructure, at least from CE to CE, since the PEs belong to the
core.
Depending on the way that CE routers are connected to PE routers, it
might be possible to intrude into a VPN that is connected to the same
PE, using layer 2 attack mechanisms such as 802.1Q-label spoofing or
ATM VPI/VCI spoofing. Layer 2 security issues will be discussed in
section 6.
It is required that VPNs cannot abuse the MPLS label mechanisms or
protocols to gain unauthorised access to other VPNs or the core.
3. Analysis of BGP/MPLS IP VPN Security
In this section, the BGP/MPLS IP VPN architecture is analysed with
respect to the security requirements listed above.
3.1. Address Space, Routing, and Traffic Separation
BGP/MPLS allows distinct IP VPNs to use the same address space, which
can also be private address space (RFC 1918 [2]). This is achieved
by adding a 64-bit Route Distinguisher (RD) to each IPv4 route,
making VPN-unique addresses also unique in the MPLS core. This
"extended" address is also called a "VPN-IPv4 address". Thus,
customers of a BGP/MPLS IP VPN service do not need to change their
current addressing plan.
Each PE router maintains a separate Virtual Routing and Forwarding
instance (VRF) for each connected VPN. A VRF includes the addresses
of that VPN as well as the addresses of the PE routers with which the
CE routers are peering. All addresses of a VRF, including these PE
addresses, belong logically to the VPN and are accessible from the
VPN. The fact that PE addresses are accessible to the VPN is not an
issue if static routing is used between the PE and CE routers, since
packet filters can be deployed to block access to all addresses of
the VRF on the PE router. If dynamic routing protocols are used, the
CE routers need to have the address of the peer PE router in the core
configured. In an environment where the service provider manages the
CE routers as CPE, this can be invisible to the customer. The
address space on the CE-PE link (including the peering PE address) is
considered part of the VPN address space. Since address space can
overlap between VPNs, the CE-PE link addresses can overlap between
VPNs. For practical management considerations, SPs typically address
CE-PE links from a global pool, maintaining uniqueness across the
core.
Routing separation between VPNs can also be achieved. Each VRF is
populated with routes from one VPN through statically configured
routes or through routing protocols that run between the PE and CE
router. Since each VPN is associated with a separate VRF there is no
interference between VPNs on the PE router.
Across the core to the other PE routers separation is maintained with
unique VPN identifiers in multiprotocol BGP, the Route Distinguishers
(RDs). VPN routes including the RD are exclusively exchanged between
PE routers by Multi-Protocol BGP (MP-BGP, RFC 2858 [8]) across the
core. These BGP routing updates are not re-distributed into the
core, but only to the other PE routers, where the information is kept
again in VPN-specific VRFs. Thus, routing across a BGP/MPLS network
is separate per VPN.
On the data plane, traffic separation is achieved by the ingress PE
pre-pending a VPN-specific label to the packets. The packets with
the VPN labels are sent through the core to the egress PE, where the
VPN label is used to select the egress VRF.
Given the addressing, routing, and traffic separation across an BGP/
MPLS IP VPN core network, it can be assumed that this architecture
offers in this respect the same security as a layer-2 VPN. It is not
possible to intrude from a VPN or the core into another VPN unless
this has been explicitly configured.
If and when confidentiality is required, it can be achieved in BGP/
MPLS IP VPNs by overlaying encryption services over the network.
However, encryption is not a standard service on BGP/MPLS IP VPNs.
See also section 5.2.
3.2. Hiding of the BGP/MPLS IP VPN Core Infrastructure
Service providers and end-customers do not normally want their
network topology revealed to the outside. This makes attacks more
difficult to execute: If an attacker doesn’t know the address of a
victim, he can only guess the IP addresses to attack. Since most DoS
attacks don’t provide direct feedback to the attacker it would be
difficult to attack the network. It has to be mentioned specifically
that information hiding as such does not provide security. However,
in the market this is a perceived requirement.
With a known IP address, a potential attacker can launch a DoS attack
more easily against that device. Therefore, the ideal is to not
reveal any information about the internal network to the outside
world. This applies to the customer network and the core. A number
of additional security measures also have to be taken: most of all,
extensive packet filtering.
For security reasons, it is recommended for any core network to
filter packets from the "outside" (Internet or connected VPNs)
destined to the core infrastructure. This makes it very hard to
attack the core, although some functionality such as pinging core
routers will be lost. Traceroute across the core will still work,
since it addresses a destination outside the core.
MPLS does not reveal unnecessary information to the outside, not even
to customer VPNs. The addressing of the core can be done with
private addresses (RFC 1918 [2]) or public addresses. Since the
interface to the VPNs as well as the Internet is BGP, there is no
need to reveal any internal information. The only information
required in the case of a routing protocol between PE and CE is the
address of the PE router. If no dynamic routing is required, static
routing on unnumbered interfaces can be configured between the PE and
CE. With this measure, the BGP/MPLS IP VPN core can be kept
completely hidden.
Customer VPNs must advertise their routes to the BGP/MPLS IP VPN core
(dynamically or statically), to ensure reachability across their VPN.
In some cases, VPN users prefer that the service provider have no
visibility of the addressing plan of the VPN. The following has to
be noted: First, the information known to the core is not about
specific hosts, but networks (routes); this offers a degree of
abstraction. Second, in a VPN-only BGP/MPLS IP VPN network (no
Internet access) this is equal to existing layer-2 models, where the
customer has to trust the service provider. Also, in a Frame Relay
or ATM network, routing and addressing information about the VPNs can
be seen on the core network.
In a VPN service with shared Internet access, the service provider
will typically announce the routes of customers who wish to use the
Internet to his upstream or peer providers. This can be done
directly if the VPN customer uses public address space, or via
Network Address Translation (NAT) to obscure the addressing
information of the customers’ networks. In either case, the customer
does not reveal more information than would be revealed by a general
Internet service. Core information will not be revealed, except for
the peering address(es) of the PE router(s) that hold(s) the peering
with the Internet. These addresses must be secured as in a
traditional IP backbone.
In summary, in a pure MPLS-VPN service, where no Internet access is
provided, information hiding is as good as on a comparable FR or ATM
network. No addressing information is revealed to third parties or
the Internet. If a customer chooses to access the Internet via the
BGP/MPLS IP VPN core, he will have to reveal the same information as
required for a normal Internet service. NAT can be used for further
obscurity. Being reachable from the Internet automatically exposes a
customer network to additional security threats. Appropriate
security mechanisms have to be deployed such as firewalls and
intrusion detection systems. This is true for any Internet access,
over MPLS or direct.
A BGP/MPLS IP VPN network with no interconnections to the Internet
has security equal to that of FR or ATM VPN networks. With an
Internet access from the MPLS cloud, the service provider has to
reveal at least one IP address (of the peering PE router) to the next
provider, and thus to the outside world.
3.3. Resistance to Attacks
Section 3.1 shows that it is impossible to directly intrude into
other VPNs. Another possibility is to attack the MPLS core and try
to attack other VPNs from there. As shown above, it is impossible to
address a P router directly. The only addresses reachable from a VPN
or the Internet are the peering addresses of the PE routers. Thus,
there are two basic ways that the BGP/MPLS IP VPN core can be
attacked:
1. By attacking the PE routers directly.
2. By attacking the signaling mechanisms of MPLS (mostly routing).
To attack an element of a BGP/MPLS IP VPN network, it is first
necessary to know the address of the element. As discussed in
section 3.2, the addressing structure of the BGP/MPLS IP VPN core is
hidden from the outside world. Thus, an attacker cannot know the IP
address of any router in the core to attack. The attacker could
guess addresses and send packets to these addresses. However, due to
the address separation of MPLS each incoming packet will be treated
as belonging to the address space of the customer. Thus, it is
impossible to reach an internal router, even by guessing IP
addresses. There is only one exception to this rule, which is the
peer interface of the PE router. This address of the PE is the only
attack point from the outside (a VPN or Internet).
The routing between a VPN and the BGP/MPLS IP VPN core can be
configured two ways:
1. Static: In this case, the PE routers are configured with static
routes to the networks behind each CE, and the CEs are configured
to statically point to the PE router for any network in other
parts of the VPN (mostly a default route). There are two sub-
cases: The static route can point to the IP address of the PE
router or to an interface of the CE router (e.g., serial0).
2. Dynamic: A routing protocol (e.g., Routing Information Protocol
(RIP), OSPF, BGP) is used to exchange routing information between
the CE and PE at each peering point.
In the case of a static route that points to an interface, the CE
router doesn’t need to know any IP addresses of the core network or
even of the PE router. This has the disadvantage of needing a more
extensive (static) configuration, but is the most secure option. In
this case, it is also possible to configure packet filters on the PE
interface to deny any packet to the PE interface. This protects the
router and the whole core from attack.
In all other cases, each CE router needs to know at least the router
ID (RID, i.e., peer IP address) of the PE router in the core, and
thus has a potential destination for an attack. One could imagine
various attacks on various services running on a router. In
practice, access to the PE router over the CE-PE interface can be
limited to the required routing protocol by using access control
lists (ACLs). This limits the point of attack to one routing
protocol, for example, BGP. A potential attack could be to send an
extensive number of routes, or to flood the PE router with routing
updates. Both could lead to a DoS, however, not to unauthorised
access.
To reduce this risk, it is necessary to configure the routing
protocol on the PE router to operate as securely as possible. This
can be done in various ways:
o By accepting only routing protocol packets, and only from the CE
router. The inbound ACL on each CE interface of the PE router
should allow only routing protocol packets from the CE to the PE.
o By configuring MD5 authentication for routing protocols. This is
available for BGP (RFC 2385 [6]), OSPF (RFC 2154 [4]), and RIP2
(RFC 2082 [3]), for example. This avoids packets being spoofed
from other parts of the customer network than the CE router. It
requires the service provider and customer to agree on a shared
secret between all CE and PE routers. It is necessary to do this
for all VPN customers. It is not sufficient to do this only for
the customer with the highest security requirements.
o By configuring parameters of the routing protocol to further
secure this communication. For example, the rate of routing
updates should be restricted where possible (in BGP through
damping); a maximum number of routes accepted per VRF and per
routing neighbor should be configured where possible; and the
Generalized TTL Security Mechanism (GTSM; RFC 3682 [10]) should be
used for all supported protocols.