wdiff draft-ietf-dnsop-dnssec-trust-anchor-03.txt draft-ietf-dnsop-dnssec-trust-anchor-04.txt

Intended Status: Informational M. Larson
DNS Operations VeriSign
Internet-Draft O. Gudmundsson
Expires: September 10, 2009 OGUD Consulting LLC
March 9, 2009 April 26, 2011 Shinkuro Inc.
October 23, 2010

DNSSEC Trust Anchor Configuration and Maintenance
draft-ietf-dnsop-dnssec-trust-anchor-03
draft-ietf-dnsop-dnssec-trust-anchor-04

Abstract

This document recommends a preferred format for specifying trust
anchors in DNSSEC validating security-aware resolvers and describes
how such a resolver should initialize trust anchors for use. This
document also describes different mechanisms for keeping trust
anchors up to date over time.

Status of this Memo

This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79. This document may contain material
from IETF Documents or IETF Contributions published or made publicly
available before November 10, 2008. The person(s) controlling the
copyright in some of this material may not have granted the IETF
Trust the right to allow modifications of such material outside the
IETF Standards Process. Without obtaining an adequate license from
the person(s) controlling the copyright in such materials, this
document may not be modified outside the IETF Standards Process, and
derivative works of it may not be created outside the IETF Standards
Process, except to format it for publication as an RFC or to
translate it into languages other than English.

Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts.
Drafts is at http://datatracker.ietf.org/drafts/current/.

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.

This Internet-Draft will expire on September 10, 2009. April 26, 2011.

This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document (http://trustee.ietf.org/license-info). document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document.

Abstract

This Code Components extracted from this document recommends a preferred format for specifying trust
anchors must
include Simplified BSD License text as described in DNSSEC validating security-aware resolvers Section 4.e of
the Trust Legal Provisions and describes
how such a resolver should initialize trust anchors for use. This
document also describes different mechanisms for keeping trust
anchors up to date over time. are provided without warranty as
described in the Simplified BSD License.

Table of Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 3
2. Trust Anchor Format and Storage . . . . . . . . . . . . . . . 4
2.1. Trust Anchor Storage . . . . . . 5 . . . . . . . . . . . . . 4
3. Trust Anchor Priming . . . . . . . . . . . . . . . . . . . . . 6
4. Trust Anchor Maintenance . . . . . . . . . . . . . . . . . . . 8
5. Security considerations . . . . . . . . . . . . . . . . . . . 10
6. IANA considerations . . . . . . . . . . . . . . . . . . . . . 11
7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 12
8. References . . . . . . . . . . . . . . . . . . . . . . . . . . 13
8.1. Normative References . . . . . . . . . . . . . . . . . . . 13
8.2. Informative References . . . . . . . . . . . . . . . . . . 13
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 14

1. Introduction

The DNSSEC standards documents ([RFC4033], [RFC4034] and [RFC4035])
describe the need for trust anchors and how they are used. A
validating security-aware resolver (subsequently referred to as a
"validating resolver") needs to be configured with one or more trust
anchors, which specify the public keys of signed zones. To
authenticate DNS data, a validating resolver builds a chain of trust
from a configured trust anchor to that data.

In a widespread public DNSSEC deployment, the

The DNS root zone would be is signed and a validating resolver would need needs to be
configured with at least the root zone's trust anchor. A validating
resolver might need additional trust anchors configured to
accommodate islands of security. (An island of security is a signed,
delegated zone that does not have an authentication chain from its
delegating parent.)
For example, consider Consider the situation where now that the root zone is
signed but when a given top-level domain (TLD) zone is not. not signed.
Various second-level zones under this unsigned TLD might be signed
and resolver operators might want to validate responses from those
zones, requiring a validating resolver to be configured with those
zones' trust anchors. Note islands of security can appear at any
depth in the DNS tree.

Because many different validating resolvers would need be configured with trust
anchors in a widespread DNSSEC deployment, there
is a benefit to creating a common trust anchor format. A similar
situation has occurred with the "root hints", the list of root name
server names and IP addresses: this information is distributed in
standard master file format and many resolver implementations support
this common format.

To simplify this trust anchor configuration process that will occur
on a large number of resolvers, this document offers guidance to
validating resolver implementers by specifying a standardized format
for describing trust anchors. The document also describes how a
validating resolver should initialize or "prime" trust anchors before
first use. Finally, the document lists options for keeping trust
anchor information current over time.

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 [RFC2119].

2. Trust Anchor Format and Storage

A trust anchor is a DNSSEC public key configured in a validating
resolver. A validating resolver's configuration MUST allow one or
more trust anchors to be specified. According to the definition in
Section 2 of RFC 4033 [RFC4033], a trust anchor can be specified as
either a public key from a DNSKEY resource record (RR) or the hash of
a public key as found in a DS RR. (DS records are defined in Section
5 of RFC 4034 [RFC4034].)

This document RECOMMENDS that a trust anchor be specified using the
hash of a public key rather than the key itself, i.e., the fields
from a DS record rather than from a DNSKEY record. A trust anchor
specified in this manner will use all the fields from the
corresponding key's DS record, including the owner name to indicate
which zone the trust anchor corresponds to as well as the various
fields from the DS RDATA. The digest algorithm SHOULD be SHA-256
[RFC4509], which is DS digest type 2. DS records using SHA-1 (DS
digest type 1) to specify trust anchors are NOT RECOMMENDED: RFC 4509
encourages the use of DS RRs using SHA-256 over those using SHA-1.

Specifying a trust anchor using a DS format instead of a DNSKEY
format offers a slight an advantage because it forces the resolver to make a
DNS query to obtain the trust anchor's complete DNSKEY RRSet during a
priming operation (described below). If only a DNSKEY record were
specified, resolver implementers could conceivably avoid priming the
trust anchor. But priming is desirable because it causes the
resolver to retrieve an up-to-date version of a zone's DNSKEY RRSet
from one of the zone's authoritative servers. It should be noted
that in practice, priming is almost always required because frequently required, when the data in
the trust anchor zone will usually be is signed with a different key than the one
configured as the trust anchor, thus requiring the
validating resolver to obtain all keys in the DNSKEY RRSet. anchor.

Using a DS format is also recommended because it is smaller than the
DNSKEY format and is easier to enter compare manually, either by typing or
cutting and pasting.

2.1. Trust Anchor Storage

For trust anchors to be useful the validating resolver needs to be
able to read a file with the trust anchors. This document recommends
that all resolvers be able to read trust anchors specified in a file
in the following format:

ZoneName [DS] KeyTag DNSKEY-Algorithm Digest-type Digest

Any truncated digest SHOULD be ignored. The text "DS" in input is
optional. The input format assumes that the trust anchor is either
in the IN class or is valid in all classes.

Validating resolvers ought to be able write out a list of current
trust anchors in the format above. Validating resolvers that perform
trust anchor maintenance MUST be able to update their trust anchor
storage.

Example: (ID width rules force text onto two lines)

. 19036 8 2
49AAC11D7B6F6446702E54A1607371607A1A41855200FD2CE1CDDE32F24E8FB5

Note: Trust anchor maintenance [RFC5011] and other schemas may
require a different format as timers and other meta data is needed.

3. Trust Anchor Priming

A validating resolver needs to obtain and validate the DNSKEY RRSet
corresponding to a configured DS for that trust anchor to be usable
in DNSSEC validation. This process is called "priming" the trust
anchor. Priming can occur when the validating resolver starts, but a
validating resolver SHOULD may want to defer priming of individual trust
anchors until each is first needed for verification. This priming on
demand is especially important when a validating resolver is
configured with a large number of trust anchors to avoid sending a
large number of DNS queries on start-up. startup. This section adds additional
details to the discussion of trust anchors in Section 5 of RFC 4035
[RFC4035].

Following are the steps a validating resolver SHOULD take to prime a
configured trust anchor:

1. Read the trust anchor's information (corresponding to the fields
in a DS record) record as descried above) from the validating resolver's
configuration (e.g., a text file).

2. Look up the DNSKEY RRSet corresponding to the owner name of the
trust anchor. (The validating resolver can either perform
iterative resolution or request recursive service from a
recursive name server, depending on its capabilities.)

3. Verify that one of the DNSKEY RR corresponding RR(s) correspond to one the
configured trust
anchor anchor(s) (i.e., one of the DNSKEY whose hash is
configured) appears in the DNSKEY RRSet and that this DNSKEY RR
has the Zone Key Flag (DNSKEY RDATA bit 7) set. (This bit only
indicates that the DNSKEY is allowed to sign the zone. zone data. This
DNSKEY may or may not be a zone signing key.) key (ZSK) as defined in
RFC 4641 [RFC4641].)

4. Verify that the DNSKEY RRSet is signed by one of the DNSKEYs
found in the previous step, i.e., that there exists a valid RRSIG
(cryptographically and temporally) for the DNSKEY RRSet generated
with the private key corresponding to the DNSKEY found in the
previous step.

If the validating resolver can successfully complete the steps above,
all DNSKEY RRs in the RRSet ought to be considered authenticated and
can be used to authenticate RRSets at or below the trust anchor.
There is one exception: if the revoke bit used by the trust anchor
automated update protocol RFC 5011 [RFC5011] is set, the trust anchor
MUST be removed and not used.

If any of the steps above result in an error, the validating resolver
SHOULD log them and abort the verification as specified in section 5
5.5 of RFC 4035 [RFC4035].

If there are multiple trust anchors configured for a zone, any one of
them is sufficient to validate data in the zone. For this reason,
old trust anchors SHOULD be removed from a validating resolver's
trust anchor list soon after the corresponding keys are no longer
used by the zone. If there are multiple trust anchors configured for
a zone, any one of them is sufficient to validate data as described in the zone.
For this reason, old trust anchors SHOULD be removed from RFC 5011 [RFC5011]. Even if a
validating resolver's
trust anchor list soon after the corresponding
keys are no longer is not used by the zone, in resolution, a validating resolver needs
to query for it frequently enough to detect changes as described prescribed in RFC 5011
[RFC5011].
RFC5011.

If a validating resolver is unable to retrieve a signed DNSKEY RRSet
corresponding to a trust anchor (i.e., prime the trust anchor), it
SHOULD log this condition as an error. Inability to prime a zone's
trust anchor results in the validating resolver's inability to
validate data from the corresponding zone. The validating resolver
MUST treat this zone as bogus, until such time it is able to get a
DNSKEY set validated by a Trust anchor. The processing of trust
anchor and DS from parent errors MUST follow the same rules. anchor.

4. Trust Anchor Maintenance

Trust anchors usually correspond to zones' key signing keys and these
keys do change in the course of normal operation. It is up to
validating resolver operators to ensure that configured trust anchor
information remains current and does not go stale: each configured
trust anchor SHOULD correspond to a DNSKEY RR in the trust anchor
zone's apex DNSKEY RRSet. This process is called trust anchor
maintenance. (Initial trust anchor configuration requires human
intervention to verify the trust anchor's authenticity using out-of-band out-of-
band means and is outside the scope of this document.)

This section provides a brief overview of some possible mechanisms to
keep trust anchor information current:

Manual configuration: The validating resolver operator MAY choose to
maintain trust anchor information completely manually. In this
case, the operator assumes responsibility for noticing stale trust
anchor information (i.e., DS records that no longer point to a
corresponding DNSKEY RR in the trust anchor zone's apex DNSKEY
RRSet) and updating that information. This process MAY require
the operator to use the same out-of-band verification mechanism as
used for initial configuration to ensure that the new trust anchor
DS record is trustworthy. Because manual maintenance is
burdensome and prone to error, and because other automated trust
anchor maintenance processes either exist or are in development,
manual trust anchor maintenance is NOT RECOMMENDED.

DNSSEC In-band Update: The IETF DNS Extensions Working Group has
developed a protocol to automatically update DNSSEC trust anchors,
which is described in RFC 5011 [RFC5011]. [RFC5011] defines an automated way
keep DNSSEC trust anchors updated. This protocol relies on a
small DNSSEC protocol change (an additional flag in the DNSKEY
record) and can be implemented either in a validating resolver
itself or in an external program with access to the validating
resolver's trust anchor configuration data.

Trusted update mechanism: Updated trust anchor information MAY be
obtained via a trusted non-DNS update mechanism. One possibility
is the operating system update mechanism provided by most software
vendors. Operators already place considerable trust in this
mechanism, so it is reasonable to extend this trust to allow
distribution and update of DNSSEC public key material. Another
possibility is to obtain trust anchor configuration directly from
the validating resolver software vendor. This A possible error
condition in this mechanism is
realistically only feasible for updating that a small number of machine is brought up with
an "old" trust
anchors, such as for the top-level domains. In configuration, like when a public DNSSEC
deployment, machine is configured
from an old media or brought out of storage. The machines ought
to be able to detect the root zone would fact the list of trust anchors is "out-
of-date" and fetch a more recent update. During this process it
may be signed necessary to disable DNSSEC and only depend on the root's
trust anchor would need updating. keys for
the update mechanism to authorize the changes to the
configuration.

Combination of update mechanisms: It is possible that for a given
validating resolver, different trust anchors will be maintained by
different mechanisms. For example, some trust anchors might be
kept up to date by a trusted update mechanism and others
maintained by some site-specific mechanism. In this case, it is
important that the mechanisms maintain a mutually exclusive set of
trust anchors.

The out-of-sync errors described above in the "Trusted update
mechanism" section can occur if the system the validating resolver is
offline or in storage for an extend period or reinstalled.

Trust Anchor Repositories (TAR) are sometimes mentioned at the same
time as a trust anchor configuration. TARs are in essence an
outsourced trust anchor maintenance mechanism, where the user can
avoid maintaining a large set of trust anchors by only configuring
the root zone's key and the TAR key.

5. Security considerations

This document proposes a standard format for documenting DNSSEC trust
anchors. Configuration of trust anchors, especially those obtained
from third parties as part of an automated process, is a critical
security operation. The procedures listed above describe the minimal
checks that should be performed and reporting that should be done
when configuring trust anchors.

In a widespread DNSSEC deployment, the

The root zone is now signed and many TLD zones
would be signed, thus TLD's are planning DNSSEC
deployment. This state of affairs greatly reducing reduces the number of
trust anchors that validating resolvers would need to store configure and keep track of.
maintain.

If multiple mechanisms are updating the trust anchor list then there
is the possibility of conflict, such as one mechanism reinserting an
expired trust anchor.

Trust anchors are configuration information. A validating resolver
ought to treat this information differently than DNS data obtained
over the network and never use the configured trust anchors as part
of an answer.

A signed zone that plans to transition to an unsigned state must
first give a warning that it is going insecure, such as using the
technique described in RFC 5011 [RFC5011]. Failure to do so will
cause all validating resolvers that keep a trust anchor for the zone
configured to treat responses from the zone as bogus, causing
resolution failures.

6. IANA considerations

This document does not have any IANA actions.

7. Acknowledgments

This work was undertaken at the suggestion of the DNSSEC Deployment
working group (www.dnssec-deployment.org). Following The following people are
acknowledged for contributing to this document, document: Alfred Hoenes, Edward
Lewis, Wes Hardaker, Geoff Huston, Paul Hoffman, Matthijs Mekking,
Scott Rose Rose, Paul Wouters.

8. References

8.1. Normative References

[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.

[RFC4033] Arends, R., Austein, R., Larson, M., Massey, D., and S.
Rose, "DNS Security Introduction and Requirements",
RFC 4033, March 2005.

[RFC4034] Arends, R., Austein, R., Larson, M., Massey, D., and S.
Rose, "Resource Records for the DNS Security Extensions",
RFC 4034, March 2005.

[RFC4035] Arends, R., Austein, R., Larson, M., Massey, D., and S.
Rose, "Protocol Modifications for the DNS Security
Extensions", RFC 4035, March 2005.

[RFC4509] Hardaker, W., "Use of SHA-256 in DNSSEC Delegation Signer
(DS) Resource Records (RRs)", RFC 4509, May 2006.

[RFC5011] StJohns, M., "Automated Updates of DNS Security (DNSSEC)
Trust Anchors", RFC 5011, September 2007.

8.2. Informative References

[RFC4641] Kolkman, O. and R. Gieben, "DNSSEC Operational Practices",
RFC 4641, September 2006.

Authors' Addresses

Matt Larson
VeriSign, Inc.
21345 Ridgetop Circle
Dulles, VA 20166-6503
USA

Email: mlarson@verisign.com

Olafur Gudmundsson
OGUD Consulting LLC
3821 Village Park Drive
Chevy Chase,
Shinkuro Inc.
4922 Fairmont Av, Suite 250
Bethsda, MD 20815 20814
USA

Email: ogud@ogud.com