Internet Draft R. Bonica
Expiration Date: August 2001 WorldCom
Tracing Requirements for Generic Tunnels
1. Status of this Memo
This document is an Internet-Draft and is in full conformance with
all provisions of Section 10 of [RFC-2026].
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
The list of Internet-Draft Shadow Directories can be accessed at
This document specifies requirements for a generic route tracing
application. The application must provide all functionality that
"traceroute" [RFC 2151] currently provides. It also must provide
enhanced capabilities with regard to tracing through tunnels (e.g.,
3. 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 [RFC-2119].
Currently, the IETF supports the following tunneling technologies:
Generic Routing Encapsulation (GRE)
Multiprotocol Label Switching (MPLS)
IP over Optical (IPO)
IP Security Protocol (IPSEC)
IP in IP
Layer 2 Tunneling Protocol (L2TP)
Although these tunneling technologies provide operators with many
useful features, they also present management challenges.
Specifically, operators require a generic route tracing application
that they can use to verify tunnel paths and diagnose tunnel faults.
This document specifies requirements for that generic route tracing
application. It also specifies requirements the protocol that will
5. Review of Existing Functionality
Currently, network operators use "traceroute" to identify the path
toward any destination in an IP network. Section 3.4 of [RFC-2151]
provides a thorough description of traceroute. Although traceroute
is very reliable and very widely deployed, it is deficient with
regard to tunnels.
Depending upon tunnel type, traceroute may display an entire tunnel
as a single IP hop, or it may display a tunnel as a collection of IP
hops, without indicating that they are part of a tunnel.
For example, assume that engineers are using IP tunnels in an IP
network. Assume also that they configure a tunnel so that the head-
end router does not copy the TTL value from the inner IP header to
outer IP header. Instead, the head-end router always sets the outer
TTL value to its maximum permitted value. When engineers trace
routes through the network, traceroute will always display the tunnel
as a single IP hop, hiding all components except the tail-end
Now assume that engineers are using MPLS to support an IP network.
Assume also that engineers configure an MPLS LSP so that the LSP
ingress router copies the TTL value from the IP header to the MPLS
header. When engineers trace routes through the network, traceroute
will always display the LSP as a series of IP hops, without
indicating that they are part of a tunnel.
6. Application Requirements
Network operators require a new route-tracing application. The new
application must provide all functionality that is currently provided
by traceroute. It also must provide enhanced tunnel tracing
The following list provides specific requirements for the new route-
1) Support in-line traces. An in-line trace reveals the path
between the host upon which the route-tracing application executes
and any interface in an IP network.
2) Support third party traces. A third party trace reveals the
path between any two points on a network. The application that
initiates a third party trace need not execute upon a host or
router that is part of the revealed path.
3) When tracing through a tunnel, either as part of an in-line
trace or a third party trace, display the tunnel either as a single
IP hop or in detail. The user's request determines how the application
4) When displaying a tunnel in detail, include the tunnel type
(e.g., GRE, MPLS), the tunnel name (if applicable) and the tunnel
identifier (if applicable). Also, include tunnel components and
round trip delay across each component.
5) Permit the application user to specify whether the application
should yield tunnel details or not.
6) If the user requests tunnel details, also allow the user to
specify a security token. Network elements will use this security
token to determine whether they will return tunnel details to that
7) Support the following tunneling technologies: GRE, MPLS, IPSEC,
IP/O, IP-in-IP, L2TP.
8) Be easily extensible to support new tunnel technologies.
9) When the tunneling technology isolates the user-plane from the
control-plane, do not rely upon the control plane to discover the
10) Support multiple levels of heterogeneous tunneling (e.g.,
IP-in-IP over MPLS).
11) Support tracing through unidirectional tunnels.
12) Terminate gracefully when tracing through a path that exceeds
a configurable maximum number of hops.
7. Protocol Requirements
Implementers require a new protocol that supports the application
described above. This protocol reveals the path between two points
in an IP network. When access policy permits, the protocol also
reveals tunnel details.
7.1. Trace-Response Information Requirements
The protocol elicits a series of trace-response messages. Each trace
response message represents a hop that connects the head-end of the
traced path to the tail-end of the traced path. A hop can be either
a top-level IP hop or lower-level hop along a tunnel.
Each trace-response message contains the following information:
1) Session Number - identifies the route-trace to which the
current trace-response is part
2) Requestor - identifies the device that hosts the route-tracing
application by IP address
3) Head-end - identifies the head-end of the traced path by IP
4) Responder - identifies the responding interface by IP address
5) Distance - specifies the number of top-level IP hops from the
head-end of the traced path to the Responder.
6) Timestamp - A timestamp copied from the message that elicited
the current trace response. The route-tracing application uses this
value to calculate round trip delay.
If the trace response represents a lower-level hop along a tunnel
path, the <<Responder>> field specifies the IP address of the top-
level IP hop that is directly upstream of the tunnel. The trace-
response message also contains one or more tunnel objects, with each
tunnel object representing a layer in the tunnel stack. For example,
assume that the trace-response represents an IP hop inside two nested
IP-in-IP tunnels. The trace-response would include two tunnel
objects, with one tunnel object representing each of the nested IP-
Each tunnel object contains the following information:
1) Depth - specifies depth in the tunnel stack.
2) Type - specifies technology that implements the tunnel (e.g.,
3) Name - specifies a potentially non-unique name that is
associated with the tunnel (e.g., LSP Name, tunnel name)
4) Identifier - specifies a unique identifier that is associated
with the tunnel (e.g., IP address, MPLS label). The identifier may
have local significance only.
5) Head-end - identifies the device that supports the head-end of
the tunnel by IP address
6) Responder - identifies the device that supports the tunnel hop
by IP address
7) Distance - Number of tunnel hops from the head-end of the tunnel.
7.2. Trace-Probe Information Requirements
A trace-probe message elicits trace-response messages.
The trace-probe message contains the following information:
1) Session Number - identifies the route trace to which the
current trace-response is part
2) Sequence Number - orders trace-probe messages within a
3) Requestor - identifies the device that hosts the route-tracing
application by IP address
4) Head-end - identifies the head-end of the traced path by IP
5) Tail-end - identifies the tail-end of the traced path by IP
6) Access control information - used by network elements to
determine whether or not tunnel details should be revealed.
7) Timestamp - a timestamp used to calculate round trip delay.
7.3. Network Layer Requirements
The Internet Protocol (IP) carries trace-probe and trace-response
messages to their destinations. IP datagrams that carry trace-probe
and trace-probe and trace-response response messages MUST NOT include
the loose source route, strict source route, or record route IP
7.4. Transport Layer Requirements
As the new protocol does not require reliable transport services, UDP
may carry trace-probe and trace-response messages to their
destinations. Trace-probe and trace-response messages may also ride
directly over IP.
7.5. Routing Requirements
The device that hosts the route-tracing application must maintain a
route to the head-end of the traced path. It need not maintain
routes to any other interface along the traced path.
In order for the trace-response message to reach its destination, the
device at the head-end of the traced path must maintain a route to
the device that hosts route-tracing application. No other device
along the traced path need maintain a route to that device.
Devices contained by tunnels must maintain routes to the head-end of
the tunnel in which they are contained. They need not maintain
routes to all devices upstream or downstream along the traced path.
Devices contained by the traced path, but not contained by tunnels,
must maintain a route to the head-end of the traced path.
7.6. Maintaining State
Because of the requirements specified in Section 7.5, devices that
support the head-end of a tunnel must relay trace-response messages
upstream through the traced path. Furthermore, devices that support
the head-end of a traced path must relay trace-response messages to
the device that hosts the route-tracing application.
The protocol may not require these devices to maintain any state
8. Path MTU Discovery (PMTU) [RFC-1191]
Existing network layer tunneling protocols, such as GRE [RFC-], may
not implement Path MTU discovery and hence may not set the Don't
Fragment bit in the encapsulating header. This can cause large
packets to become fragmented within the tunnel and reassembled at the
tunnel exit (independent of whether the payload packet is using
PMTU). This may or may not be a problem for some higher level
protocols, but the behavior of the packet network itself is not
incorrect in this case. However, if a tunnel entry point were to use
Path MTU discovery, that tunnel entry point would also need to relay
ICMP unreachable error messages (in particular the "fragmentation
needed and DF set" code) back to the originator of the packet, which
is not in general a requirement for network layer tunneling protocols
(and may in practice be difficult, as in the case of nested tunnels).
Note however that failure to properly relay Path MTU information to
an originator can result in the following behavior: the originator
sets the don't fragment bit, the packet gets dropped within the
tunnel, but since the originator doesn't receive proper feedback, it
retransmits with the same PMTU, causing subsequently transmitted
packets to be dropped. In this case the packet network does not
operate correctly. How do we want to handle this? Make this
protocol's tunnel ingress points maintain tunnel MTU?
9. IANA Guidelines
Protocol code points [RFC-2434].
10. Security Considerations
A configurable access control policy determines the degree to which
features described herein are delivered. The access control policy
requires user identification and authorization.
As stated above, the new protocol must not introduce security holes
nor consume excessive resources (e.g., CPU, bandwidth). It also must
not be exploitable by those launching DoS attacks.
[RFC-2026], Bradner, S., "Internet Standards Process Revision 3", RFC
2026, Harvard University, October 1996.
[RFC-2119], Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", RFC 2119, Harvard University, March 1997
[RFC-2151], Kessler, G., Shepard, S., A Primer On Internet and TCP/IP
Tools and Utilities, RFC 2151, Hill Associates, Inc., June 1997
[RFC-2434] T. Narten and H. Alvestrand, "Guidelines for Writing an
IANA Considerations Section in RFCs", RFC 2434, October, 1998.
[RFC-2637] Hamzeh, K. et. al., "Point-to-Point Tunneling Protocol
(PPTP)", RFC 2637, July, 1999.
Thanks to Randy Bush and Steve Bellovin for their comments.
13. Author's Addresses
Ronald P. Bonica
22001 Loudoun County Pkwy
Ashburn, Virginia, 20147
Phone: 703 886 1681
Juniper Networks, Inc.
1194 N. Mathilda Ave.
Sunnyvale, California 94089
170 Tasman Drive
San Jose, California 94025
14. Full Copyright Statement
Copyright (C) The Internet Society (2000). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.