4.3. Parent Policies
4.3.1. Initial Key Exchanges and Parental Policies Considerations
The initial key exchange is always subject to the policies set by the
parent. It is specifically important in a registry-registrar-
registrant model where a registry maintains the parent zone, and the
registrant (the user of the child-domain name) deals with the
registry through an intermediary called a registrar (see [RFC3375]
for a comprehensive definition). The key material is to be passed
from the DNS operator to the parent via a registrar, where both the
DNS operator and registrar are selected by the registrant and might
be different organizations. When designing a key exchange policy,
one should take into account that the authentication and
authorization mechanisms used during a key exchange should be as
strong as the authentication and authorization mechanisms used for
the exchange of delegation information between the parent and child.
That is, there is no implicit need in DNSSEC to make the
authentication process stronger than it is for regular DNS.
Using the DNS itself as the source for the actual DNSKEY material has
the benefit that it reduces the chances of user error. A DNSKEY
query tool can make use of the SEP bit [RFC4035] to select the proper
key(s) from a DNSSEC key set, thereby reducing the chance that the
wrong DNSKEY is sent. It can validate the self-signature over a key,
thereby verifying the ownership of the private key material.
Fetching the DNSKEY from the DNS ensures that the chain of trust
remains intact once the parent publishes the DS RR indicating that
the child is secure.
Note: Out-of-band verification is still needed when the key material
is fetched for the first time, even via DNS. The parent can never be
sure whether or not the DNSKEY RRs have been spoofed.
With some types of key rollovers, the DNSKEY is not pre-published,
and a DNSKEY query tool is not able to retrieve the successor key.
In this case, the out-of-band method is required. This also allows
the child to determine the digest algorithm of the DS record.
4.3.2. Storing Keys or Hashes?
When designing a registry system, one should consider whether to
store the DNSKEYs and/or the corresponding DSs. Since a child zone
might wish to have a DS published using a message digest algorithm
not yet understood by the registry, the registry can't count on being
able to generate the DS record from a raw DNSKEY. Thus, we suggest
that registry systems should be able to store DS RRs, even if they
also store DNSKEYs (see also "DNSSEC Trust Anchor Configuration and
The storage considerations also relate to the design of the customer
interface and the method by which data is transferred between the
registrant and registry: Will the child-zone administrator be able to
upload DS RRs with unknown hash algorithms, or does the interface
only allow DNSKEYs? When registries support the Extensible
Provisioning Protocol (EPP) [RFC5910], that can be used for
registrar-registry interactions, since that protocol allows the
transfer of both DS and, optionally, DNSKEY RRs. There is no
standardized way to move the data between the customer and the
registrar. Different registrars have different mechanisms, ranging
from simple web interfaces to various APIs. In some cases, the use
of the DNSSEC extensions to EPP may be applicable.
Having an out-of-band mechanism such as a registry directory (e.g.,
Whois) to find out which keys are used to generate DS Resource
Records for specific owners and/or zones may also help with
4.3.3. Security Lameness
Security lameness is defined as the state whereby the parent has a DS
RR pointing to a nonexistent DNSKEY RR. Security lameness may occur
temporarily during a Double-DS rollover scheme. However, care should
be taken that not all DS RRs are pointing to a nonexistent DNSKEY RR,
which will cause the child's zone to be marked Bogus by verifying DNS
As part of a comprehensive delegation check, the parent could, at key
exchange time, verify that the child's key is actually configured in
the DNS. However, if a parent does not understand the hashing
algorithm used by the child, the parental checks are limited to only
comparing the key id.
Child zones should be very careful in removing DNSKEY material --
specifically, SEP keys -- for which a DS RR exists.
Once a zone is "security lame", a fix (e.g., removing a DS RR) will
take time to propagate through the DNS.
4.3.4. DS Signature Validity Period
Since the DS can be replayed as long as it has a valid signature, a
short signature validity period for the DS RRSIG minimizes the time
that a child is vulnerable in the case of a compromise of the child's
KSK(s). A signature validity period that is too short introduces the
possibility that a zone is marked Bogus in the case of a
configuration error in the signer. There may not be enough time to
fix the problems before signatures expire (this is a generic
argument; also see Section 4.4.2). Something as mundane as zone
administrator unavailability during weekends shows the need for DS
signature validity periods longer than two days. Just like any
signature validity period, we suggest an absolute minimum for the DS
signature validity period of a few days.
The maximum signature validity period of the DS record depends on how
long child zones are willing to be vulnerable after a key compromise.
On the other hand, shortening the DS signature validity period
increases the operational risk for the parent. Therefore, the parent
may have a policy to use a signature validity period that is
considerably longer than the child would hope for.
A compromise between the policy/operational constraints of the parent
and minimizing damage for the child may result in a DS signature
validity period somewhere between a week and several months.
In addition to the signature validity period, which sets a lower
bound on the number of times the zone administrator will need to sign
the zone data and an upper bound on the time that a child is
vulnerable after key compromise, there is the TTL value on the DS
RRs. Shortening the TTL reduces the damage of a successful replay
attack. It does mean that the authoritative servers will see more
queries. But on the other hand, a short TTL lowers the persistence
of DS RRsets in caches, thereby increasing the speed with which
updated DS RRsets propagate through the DNS.
4.3.5. Changing DNS Operators
The parent-child relationship is often described in terms of a
registry-registrar-registrant model, where a registry maintains the
parent zone and the registrant (the user of the child-domain name)
deals with the registry through an intermediary called a registrar
[RFC3375]. Registrants may outsource the maintenance of their DNS
system, including the maintenance of DNSSEC key material, to the
registrar or to another third party, referred to here as the DNS
For various reasons, a registrant may want to move between DNS
operators. How easy this move will be depends principally on the DNS
operator from which the registrant is moving (the losing operator),
as the losing operator has control over the DNS zone and its keys.
The following sections describe the two cases: where the losing
operator cooperates with the new operator (the gaining operator), and
where the two do not cooperate.
22.214.171.124. Cooperating DNS Operators
In this scenario, it is assumed that the losing operator will not
pass any private key material to the gaining operator (that would
constitute a trivial case) but is otherwise fully cooperative.
In this environment, the change could be made with a Pre-Publish ZSK
rollover, whereby the losing operator pre-publishes the ZSK of the
gaining operator, combined with a Double-Signature KSK rollover where
the two registrars exchange public keys and independently generate a
signature over those key sets that they combine and both publish in
their copy of the zone. Once that is done, they can use their own
private keys to sign any of their zone content during the transfer.
In this figure, A denotes the losing operator and B the gaining
operator. RRSIG_Z is the RRSIG produced by a ZSK, RRSIG_K is
produced with a KSK, and the appended A or B indicates the producers
of the key pair. "Child at A" is how the zone content is represented
by the losing DNS operator, and "Child at B" is how the zone content
is represented by the gaining DNS operator.
The zone is initially delegated from the parent to the name servers
of operator A. Operator A uses his own ZSK and KSK to sign the zone.
The cooperating operator A will pre-publish the new NS record and the
ZSK and KSK of operator B, including the RRSIG over the DNSKEY RRset
generated by the KSK of operator B. Operator B needs to publish the
same DNSKEY RRset. When that DNSKEY RRset has populated the caches,
the re-delegation can be made, which involves adjusting the NS and DS
records in the parent zone to point to operator B. And after all
DNSSEC records related to operator A have expired from the caches,
operator B can stop publishing the keys and signatures belonging to
operator A, and vice versa.
The requirement to exchange signatures has a couple of drawbacks. It
requires more operational overhead, because not only do the operators
have to exchange public keys but they also have to exchange the
signatures of the new DNSKEY RRset. This drawback does not exist if
the Double-Signature KSK rollover is replaced with a Double-DS KSK
rollover. See Figure 15 in Appendix D for the diagram.
Thus, if the registry and registrars allow DS records to be published
that do not point to a published DNSKEY in the child zone, the
Double-DS KSK rollover is preferred (see Figure 5), in combination
with the Pre-Publish ZSK rollover. This does not require sharing the
KSK signatures between the operators, but both operators still have
to publish each other's ZSKs.
126.96.36.199. Non-Cooperating DNS Operators
In the non-cooperating case, matters are more complicated. The
losing operator may not cooperate and leave the data in the DNS as
is. In extreme cases, the losing operator may become obstructive and
publish a DNSKEY RR with a high TTL and corresponding signature
validity period so that registrar A's DNSKEY could end up in caches
for (in theory at least) decades.
The problem arises when a validator tries to validate with the losing
operator's key and there is no signature material produced with the
losing operator available in the delegation path after re-delegation
from the losing operator to the gaining operator has taken place.
One could imagine a rollover scenario where the gaining operator
takes a copy of all RRSIGs created by the losing operator and
publishes those in conjunction with its own signatures, but that
would not allow any changes in the zone content. Since a
re-delegation took place, the NS RRset has by definition changed, so
such a rollover scenario will not work. Besides, if zone transfers
are not allowed by the losing operator and NSEC3 is deployed in the
losing operator's zone, then the gaining operator's zone will not
have certainty that all of the losing operator's RRSIGs have been
The only viable operation for the registrant is to have his zone go
Insecure for the duration of the change. The registry should be
asked to remove the DS RR pointing to the losing operator's DNSKEY
and to change the NS RRset to point to the gaining operator. Once
this has propagated through the DNS, the registry should be asked to
insert the DS record pointing to the (newly signed) zone at
Note that some behaviors of resolver implementations may aid in the
process of changing DNS operators:
o TTL sanity checking, as described in RFC 2308 [RFC2308], will
limit the impact of the actions of an obstructive losing operator.
Resolvers that implement TTL sanity checking will use an upper
limit for TTLs on RRsets in responses.
o If RRsets at the zone cut (are about to) expire, the resolver
restarts its search above the zone cut. Otherwise, the resolver
risks continuing to use a name server that might be un-delegated
by the parent.
o Limiting the time that DNSKEYs that seem to be unable to validate
signatures are cached and/or trying to recover from cases where
DNSKEYs do not seem to be able to validate data also reduce the
effects of the problem of non-cooperating registrars.
However, there is no operational methodology to work around this
business issue, and proper contractual relationships between all
involved parties seem to be the only solution to cope with these
problems. It should be noted that in many cases, the problem with
temporary broken delegations already exists when a zone changes from
one DNS operator to another. Besides, it is often the case that when
operators are changed, the services that are referenced by that zone
also change operators, possibly involving some downtime.
In any case, to minimize such problems, the classic configuration is
to have relatively short TTLs on all involved Resource Records. That
will solve many of the problems regarding changes to a zone,
regardless of whether DNSSEC is used.
4.4. Time in DNSSEC
Without DNSSEC, all times in the DNS are relative. The SOA fields
REFRESH, RETRY, and EXPIRATION are timers used to determine the time
that has elapsed after a slave server synchronized with a master
server. The TTL value and the SOA RR minimum TTL parameter [RFC2308]
are used to determine how long a forwarder should cache data (or
negative responses) after it has been fetched from an authoritative
server. By using a signature validity period, DNSSEC introduces the
notion of an absolute time in the DNS. Signatures in DNSSEC have an
expiration date after which the signature is marked as invalid and
the signed data is to be considered Bogus.
The considerations in this section are all qualitative and focused on
the operational and managerial issues. A more thorough quantitative
analysis of rollover timing parameters can be found in "DNSSEC Key
Timing Considerations" [DNSSEC-KEY-TIMING].
4.4.1. Time Considerations
Because of the expiration of signatures, one should consider the
o We suggest that the Maximum Zone TTL value of your zone data be
smaller than your signature validity period.
If the TTL duration was similar to that of the signature
validity period, then all RRsets fetched during the validity
period would be cached until the signature expiration time.
Section 8.1 of RFC 4033 [RFC4033] suggests that "the resolver
may use the time remaining before expiration of the signature
validity period of a signed RRset as an upper bound for the
TTL". As a result, the query load on authoritative servers
would peak at the signature expiration time, as this is also
the time at which records simultaneously expire from caches.
Having a TTL that is at least a few times smaller than your
signature validity period avoids query load peaks.
o We suggest that the signature publication period end at least one
Maximum Zone TTL duration (but preferably a minimum of a few days)
before the end of the signature validity period.
Re-signing a zone shortly before the end of the signature
validity period may cause the simultaneous expiration of data
from caches. This in turn may lead to peaks in the load on
authoritative servers. To avoid this, schemes are deployed
whereby the zone is periodically visited for a re-signing
operation, and those signatures that are within a so-called
Refresh Period from signature expiration are recreated. Also
see Section 4.4.2 below.
In the case of an operational error, you would have one Maximum
Zone TTL duration to resolve the problem. Re-signing a zone a
few days before the end of the signature validity period
ensures that the signatures will survive at least a (long)
weekend in case of such operational havoc. This is called the
Refresh Period (see Section 4.4.2).
o We suggest that the Minimum Zone TTL be long enough to both fetch
and verify all the RRs in the trust chain. In workshop
environments, it has been demonstrated [NIST-Workshop] that a low
TTL (under 5 to 10 minutes) caused disruptions because of the
following two problems:
1. During validation, some data may expire before the validation
is complete. The validator should be able to keep all data
until it is completed. This applies to all RRs needed to
complete the chain of trust: DS, DNSKEY, RRSIG, and the final
answers, i.e., the RRset that is returned for the initial
2. Frequent verification causes load on recursive name servers.
Data at delegation points, DS, DNSKEY, and RRSIG RRs benefits
from caching. The TTL on those should be relatively long.
Data at the leaves in the DNS tree has less impact on
recursive name servers.
o Slave servers will need to be able to fetch newly signed zones
well before the RRSIGs in the zone served by the slave server pass
their signature expiration time.
When a slave server is out of synchronization with its master
and data in a zone is signed by expired signatures, it may be
better for the slave server not to give out any answer.
Normally, a slave server that is not able to contact a master
server for an extended period will expire a zone. When that
happens, the server will respond differently to queries for
that zone. Some servers issue SERVFAIL, whereas others turn
off the AA bit in the answers. The time of expiration is set
in the SOA record and is relative to the last successful
refresh between the master and the slave servers. There exists
no coupling between the signature expiration of RRSIGs in the
zone and the expire parameter in the SOA.
If the server serves a DNSSEC-secured zone, then it may happen
that the signatures expire well before the SOA expiration timer
counts down to zero. It is not possible to completely prevent
this by modifying the SOA parameters.
However, the effects can be minimized where the SOA expiration
time is equal to or shorter than the Refresh Period (see
The consequence of an authoritative server not being able to
update a zone for an extended period of time is that signatures
may expire. In this case, non-secure resolvers will continue
to be able to resolve data served by the particular slave
servers, while security-aware resolvers will experience
problems because of answers being marked as Bogus.
We suggest that the SOA expiration timer be approximately one
third or a quarter of the signature validity period. It will
allow problems with transfers from the master server to be
noticed before signatures time out.
We also suggest that operators of name servers that supply
secondary services develop systems to identify upcoming
signature expirations in zones they slave and take appropriate
action where such an event is detected.
When determining the value for the expiration parameter, one
has to take the following into account: What are the chances
that all secondaries expire the zone? How quickly can the
administrators of the secondary servers be reached to load a
valid zone? These questions are not DNSSEC-specific but may
influence the choice of your signature validity periods.
4.4.2. Signature Validity Periods
188.8.131.52. Maximum Value
The first consideration for choosing a maximum signature validity
period is the risk of a replay attack. For low-value, long-term
stable resources, the risks may be minimal, and the signature
validity period may be several months. Although signature validity
periods of many years are allowed, the same "operational habit"
arguments as those given in Section 3.2.2 play a role: When a zone is
re-signed with some regularity, then zone administrators remain
conscious of the operational necessity of re-signing.
184.108.40.206. Minimum Value
The minimum value of the signature validity period is set for the
time by which one would like to survive operational failure in
provisioning: At what time will a failure be noticed, and at what
time is action expected to be taken? By answering these questions,
availability of zone administrators during (long) weekends or time
taken to access backup media can be taken into account. The result
could easily suggest a minimum signature validity period of a few
Note, however, that the argument above is assuming that zone data has
just been signed and published when the problem occurred. In
practice, it may be that a zone is signed according to a frequency
set by the Re-Sign Period, whereby the signer visits the zone content
and only refreshes signatures that are within a given amount of time
(the Refresh Period) of expiration. The Re-Sign Period must be
smaller than the Refresh Period in order for zone data to be signed
in a timely fashion.
If an operational problem occurs during re-signing, then the
signatures in the zone to expire first are the ones that have been
generated longest ago. In the worst case, these signatures are the
Refresh Period minus the Re-Sign Period away from signature
To make matters slightly more complicated, some signers vary the
signature validity period over a small range (the jitter interval) so
that not all signatures expire at the same time.
In other words, the minimum signature validity period is set by first
choosing the Refresh Period (usually a few days), then defining the
Re-Sign Period in such a way that the Refresh Period minus the
Re-Sign Period, minus the maximum jitter sets the time in which
operational havoc can be resolved.
The relationship between signature times is illustrated in Figure 11.
Inception Signing Expiration
time time time
| | | | |
| | | | |
| Inception offset | |
|<---------------->| Validity Period |
Inception Signing Reuse Reuse Reuse New Expiration
time time RRSIG time
| | | | | | |
| | | | | | |
<-----> <-----> <-----> <----->
| Refresh |
| Period |
Figure 11: Signature Timing Parameters
Note that in the figure the validity of the signature starts shortly
before the signing time. That is done to deal with validators that
might have some clock skew. This is called the inception offset, and
it should be chosen so that false negatives are minimized to a
220.127.116.11. Differentiation between RRsets
It is possible to vary signature validity periods between signatures
over different RRsets in the zone. In practice, this could be done
when zones contain highly volatile data (which may be the case in
dynamic-update environments). Note, however, that the risk of replay
(e.g., by stale secondary servers) should be the leading factor in
determining the signature validity period, since the TTLs on the data
itself are still the primary parameter for cache expiry.
In some cases, the risk of replaying existing data might be different
from the risk of replaying the denial of data. In those cases, the
signature validity period on NSEC or NSEC3 records may be tweaked
When a zone contains secure delegations, then a relatively short
signature validity period protects the child against replay attacks
in the case where the child's key is compromised (see Section 4.3.4).
Since there is a higher operational risk for the parent registry when
choosing a short validity period and a higher operational risk for
the child when choosing a long validity period, some (price)
differentiation may occur for validity periods between individual DS
RRs in a single zone.
There seem to be no other arguments for differentiation in validity
5. "Next Record" Types
One of the design tradeoffs made during the development of DNSSEC was
to separate the signing and serving operations instead of performing
cryptographic operations as DNS requests are being serviced. It is
therefore necessary to create records that cover the very large
number of nonexistent names that lie between the names that do exist.
There are two mechanisms to provide authenticated proof of
nonexistence of domain names in DNSSEC: a clear-text one and an
obfuscated-data one. Each mechanism:
o includes a list of all the RRTYPEs present, which can be used to
prove the nonexistence of RRTYPEs at a certain name;
o stores only the name for which the zone is authoritative (that is,
glue in the zone is omitted); and
o uses a specific RRTYPE to store information about the RRTYPEs
present at the name: The clear-text mechanism uses NSEC, and the
obfuscated-data mechanism uses NSEC3.
5.1. Differences between NSEC and NSEC3
The clear-text mechanism (NSEC) is implemented using a sorted linked
list of names in the zone. The obfuscated-data mechanism (NSEC3) is
similar but first hashes the names using a one-way hash function,
before creating a sorted linked list of the resulting (hashed)
The NSEC record requires no cryptographic operations aside from the
validation of its associated signature record. It is human readable
and can be used in manual queries to determine correct operation.
The disadvantage is that it allows for "zone walking", where one can
request all the entries of a zone by following the linked list of
NSEC RRs via the "Next Domain Name" field. Though all agree that DNS
data is accessible through query mechanisms, for some zone
administrators this behavior is undesirable for policy, regulatory,
or other reasons.
Furthermore, NSEC requires a signature over every RR in the zone
file, thereby ensuring that any denial of existence is
cryptographically signed. However, in a large zone file containing
many delegations, very few of which are to signed zones, this may
produce unacceptable additional overhead, especially where insecure
delegations are subject to frequent updates (a typical example might
be a TLD operator with few registrants using secure delegations).
NSEC3 allows intervals between two secure delegations to "opt out",
in which case they may contain one or more insecure delegations, thus
reducing the size and cryptographic complexity of the zone at the
expense of the ability to cryptographically deny the existence of
names in a specific span.
The NSEC3 record uses a hashing method of the requested name. To
increase the workload required to guess entries in the zone, the
number of hashing iterations can be specified in the NSEC3 record.
Additionally, a salt can be specified that also modifies the hashes.
Note that NSEC3 does not give full protection against information
leakage from the zone (you can still derive the size of the zone,
which RRTYPEs are in there, etc.).
5.2. NSEC or NSEC3
The first motivation to deploy NSEC3 -- prevention of zone
enumeration -- only makes sense when zone content is not highly
structured or trivially guessable. Highly structured zones, such as
in-addr.arpa., ip6.arpa., and e164.arpa., can be trivially enumerated
using ordinary DNS properties, while for small zones that only
contain records in the apex of the zone and a few common names such
as "www" or "mail", guessing zone content and proving completeness is
also trivial when using NSEC3. In these cases, the use of NSEC is
preferred to ease the work required by signers and validating
For large zones where there is an implication of "not readily
available" names, such as those where one has to sign a
non-disclosure agreement before obtaining it, NSEC3 is preferred.
The second reason to consider NSEC3 is "Opt-Out", which can reduce
the number of NSEC3 records required. This is discussed further
below (Section 5.3.4).
5.3. NSEC3 Parameters
NSEC3 is controlled by a number of parameters, some of which can be
varied: This section discusses the choice of those parameters.
5.3.1. NSEC3 Algorithm
The NSEC3 hashing algorithm is performed on the Fully Qualified
Domain Name (FQDN) in its uncompressed form. This ensures that brute
force work done by an attacker for one FQDN cannot be reused for
another FQDN attack, as these entries are by definition unique.
At the time of this writing, there is only one NSEC3 hash algorithm
defined. [RFC5155] specifically states: "When specifying a new hash
algorithm for use with NSEC3, a transition mechanism MUST also be
defined". Therefore, this document does not consider NSEC3 hash
5.3.2. NSEC3 Iterations
One of the concerns with NSEC3 is that a pre-calculated dictionary
attack could be performed in order to assess whether or not certain
domain names exist within a zone. Two mechanisms are introduced in
the NSEC3 specification to increase the costs of such dictionary
attacks: iterations and salt.
The iterations parameter defines the number of additional times the
hash function has been performed. A higher value results in greater
resiliency against dictionary attacks, at a higher computational cost
for both the server and resolver.
RFC 5155Section 10.3 [RFC5155] considers the tradeoffs between
incurring cost during the signing process and imposing costs to the
validating name server, while still providing a reasonable barrier
against dictionary attacks. It provides useful limits of iterations
for a given RSA key size. These are 150 iterations for 1024-bit
keys, 500 iterations for 2048-bit keys, and 2,500 iterations for
4096-bit keys. Choosing a value of 100 iterations is deemed to be a
sufficiently costly, yet not excessive, value: In the worst-case
scenario, the performance of name servers would be halved, regardless
of key size [NSEC3-HASH-PERF].
5.3.3. NSEC3 Salt
While the NSEC3 iterations parameter increases the cost of hashing a
dictionary word, the NSEC3 salt reduces the lifetime for which that
calculated hash can be used. A change of the salt value by the zone
administrator would cause an attacker to lose all pre-calculated work
for that zone.
There must be a complete NSEC3 chain using the same salt value, that
matches the salt value in the NSEC3PARAM record. NSEC3 salt changes
do not need special rollover procedures. Since changing the salt
requires that all the NSEC3 records be regenerated and thus requires
generating new RRSIGs over these NSEC3 records, it makes sense to
align the change of the salt with a change of the Zone Signing Key,
as that process in itself already usually requires that all RRSIGs be
regenerated. If there is no critical dependency on incremental
signing and the zone can be signed with little effort, there is no
need for such alignment.
The Opt-Out mechanism was introduced to allow for a gradual
introduction of signed records in zones that contain mostly
delegation records. The use of the Opt-Out flag changes the meaning
of the NSEC3 span from authoritative denial of the existence of names
within the span to proof that DNSSEC is not available for the
delegations within the span. This allows for the addition or removal
of the delegations covered by the span without recalculating or
re-signing RRs in the NSEC3 RR chain.
Opt-Out is specified to be used only over delegation points and will
therefore only bring relief to zones with a large number of insecure
delegations. This consideration typically holds for large TLDs and
similar zones; in most other circumstances, Opt-Out should not be
deployed. Further considerations can be found in Section 12.2 of
RFC 5155 [RFC5155].