A Common Operational Problem in DNS Servers: Failure to Communicate
Internet Systems ConsortiumPO Box 360NewmarketNH03857United States of Americamarka@isc.orgInternet Systems ConsortiumPO Box 360NewmarketNH03857United States of Americaray@isc.org
The DNS is a query/response protocol. Failing to respond
to queries, or responding incorrectly, causes both immediate
operational problems and long-term problems with protocol
development.
This document identifies a number of common kinds of queries
to which some servers either fail to respond or respond
incorrectly. This document also suggests procedures for
zone operators to apply to identify and remediate the problem.
The document does not look at the DNS data itself, just the
structure of the responses.
IntroductionThe DNS is a query/response protocol.
Failing to respond to queries or responding incorrectly causes both
immediate operational problems and long-term problems with protocol
development.
Failure to respond to a query is indistinguishable from
packet loss without doing an analysis of query-response
patterns. Additionally, failure to respond results in
unnecessary queries being made by DNS clients and
introduces delays to the resolution process.
Due to the inability to distinguish between packet loss and
nameservers or middleboxes dropping Extension Mechanisms for DNS
(EDNS)
queries, packet loss is sometimes misclassified as lack of EDNS
support, which can lead to DNSSEC validation failures.
The existence of servers that fail to respond to queries
results in developers being hesitant to deploy new standards.
Such servers need to be identified and remediated.
The DNS has response codes that cover almost any conceivable
query response. A nameserver should be able to respond to
any conceivable query using them. There should be no need
to drop queries because a nameserver does not understand
them.
Unless a nameserver is under attack, it should respond to
all DNS requests directed to it. When a nameserver is under
attack, it may wish to drop packets. A common attack is to
use a nameserver as an amplifier by sending spoofed packets.
This is done because response packets are bigger than the
queries and large amplification factors are available, especially
if EDNS is supported. Limiting the rate of responses is
reasonable when this is occurring, and the client should
retry. However, this only works if legitimate clients are
not being forced to guess whether or not EDNS queries are accepted. As long as there is still a pool of servers that don't
respond to EDNS requests, clients have no way to know if
the lack of response is due to packet loss, EDNS packets
not being supported, or rate limiting due to the server being
under attack. Misclassification of server behaviour is
unavoidable when rate limiting is used until the population
of servers that fail to respond to well-formed queries
drops to near zero.
Nameservers should respond to queries even if the queried
name is not for any name the server is configured to answer
for. Misconfigured nameservers are a common occurrence in the
DNS, and receiving queries for zones that the server is not
configured for is not necessarily an indication that the
server is under attack. Parent zone operators are advised
to regularly check that the delegating NS records are
consistent with those of the delegated zone and to correct
them when they are not (,
Paragraph 3).
Doing this regularly should reduce the instances
of broken delegations.
This document does not try to identify all possible errors
nor does it supply an exhaustive list of tests.
Consequences
Failure to follow the guidance in relevant DNS RFCs has multiple adverse
consequences. Some are caused directly by the non-compliant
behaviour and others as a result of workarounds forced on
recursive servers. Addressing known issues now will reduce
future interoperability issues as the DNS protocol continues
to evolve and clients make use of newly introduced DNS
features. In particular, the base DNS specification
and the
EDNS specification , when implemented,
need to be followed.
Some examples of known consequences include the following:
The AD (Authenticated Data) bit in a response
cannot be trusted to mean anything, as some servers
incorrectly copy the flag bit from the request to the
response .
The use of the AD bit in requests is defined in
.
Widespread non-response to EDNS queries has led to
recursive servers having to assume that EDNS is not
supported and that fallback to plain DNS is required,
potentially causing DNSSEC validation failures.
Widespread non-response to EDNS options requires
recursive servers to decide whether to probe
to see if it is the specific EDNS option or the use of
EDNS in general that is causing the non-response. In
the limited amount of time required to resolve a query
before the client times out, this is not possible.
Incorrectly returning FORMERR to an EDNS option being
present leads to the recursive server not being able
to determine if the server is just broken in the handling
of the EDNS option or if it doesn't support EDNS at all.
Mishandling of unknown query types has contributed
to the abandonment of the transition of the SPF type.
Mishandling of unknown query types has slowed up the
development of DNS-Based Authentication of Named Entities (DANE)
and resulted in additional rules
being specified to reduce the probability of interacting
with a broken server when making TLSA queries.
The consequences of servers not following the RFCs will
only grow if measures are not put in place to remove non-compliant
servers from the ecosystem. Working around issues
due to non-compliance with RFCs is not sustainable.
Most (if not all) of these consequences could have been
avoided if action had been taken to remove non-compliant
servers as soon as people were aware of them, i.e., to actively
seek out broken implementations and servers and inform their
developers and operators that they need to fix their servers.
Common Kinds of Queries That Result in No or Bad Responses
This section is broken down into Basic DNS requests and EDNS requests.
Basic DNS QueriesZone Existence
If a zone is delegated to a server, that server should
respond to a SOA query for that zone with an SOA record.
Failing to respond at all is always incorrect, regardless
of the configuration of the server. Responding with
anything other than an SOA record in the answer section
indicates a bad delegation.
Unknown/Unsupported Type Queries
Some servers fail to respond to unknown or unsupported
types. If a server receives a query for a type that
it doesn't recognise, or doesn't implement, it is
expected to return the appropriate response as if it
did recognise the type but does not have any data for
that type, i.e., either NOERROR or NXDOMAIN. The exceptions
to this are queries for Meta-RR types, which may return
NOTIMP.
DNS Flags
Some servers fail to respond to DNS queries with various
DNS flags set, regardless of whether they are defined or
still reserved. At the time of writing, there are servers
that fail to respond to queries with the AD flag set to 1
and servers that fail to respond to queries with the last
reserved flag set.
Servers should respond to such queries. If the server does not
know the meaning of a flag, it must not copy it to the response
(). If
the server does not understand the meaning of a request, it should
reply with a FORMERR response with unknown flags set to zero.Recursive Queries
A non-recursive server is supposed to respond to recursive
queries as if the Recursion Desired (RD) bit is not set .
Unknown DNS Opcodes
The use of previously undefined opcodes is to be expected.
Since the DNS was first defined, two new opcodes have been
added, UPDATE and NOTIFY.
NOTIMP is the expected rcode to an unknown or unimplemented
opcode.
TCP Queries
All DNS servers are supposed to respond to queries over
TCP . While firewalls should not
block TCP connection attempts, those that do should
cleanly terminate the connection by sending TCP RESET
or sending ICMP/ICMPv6 Administratively Prohibited
messages. Dropping TCP connections introduces excessive
delays to the resolution process.
EDNS Queries
EDNS queries are specified in .
EDNS Queries: Version Independent
Identifying servers that fail to respond to EDNS queries
can be done by first confirming that the server responds
to regular DNS queries, followed by a series of otherwise
identical queries using EDNS, then making the original
query again. A series of EDNS queries is needed, as at
least one DNS implementation responds to the first EDNS
query with FORMERR but fails to respond to subsequent
queries from the same address for a period until a
regular DNS query is made. The EDNS query should specify
a UDP buffer size of 512 bytes to avoid false classification
of not supporting EDNS due to response packet size.
If the server responds to the first and last queries
but fails to respond to most or all of the EDNS queries,
it is probably faulty. The test should be repeated a
number of times to eliminate the likelihood of a false
positive due to packet loss.
Firewalls may also block larger EDNS responses, but there
is no easy way to check authoritative servers to see
if the firewall is misconfigured.
EDNS Queries: Version Specific
Some servers respond correctly to EDNS version 0 queries
but fail to respond to EDNS queries with version numbers
that are higher than zero. Servers should respond with
BADVERS to EDNS queries with version numbers that they
do not support.
Some servers respond correctly to EDNS version 0 queries
but fail to set QR=1 when responding to EDNS versions
they do not support. Such responses may be discarded
as invalid (as QR is not 1) or treated as requests (when the
source port of the original request was port 53).
EDNS Options
Some servers fail to respond to EDNS queries with EDNS
options set. The original EDNS specification left this
behaviour undefined , but the
correct behaviour was clarified in .
Unknown EDNS options are supposed to be ignored by the server.
EDNS Flags
Some servers fail to respond to EDNS queries with EDNS
flags set. Servers should ignore EDNS flags they do not
understand and must not add them to the response .
Truncated EDNS Responses
Some EDNS-aware servers fail to include an OPT record when a
truncated response is sent. An OPT record is supposed to be
included in a truncated response .
Some EDNS-aware servers fail to honour the advertised EDNS
UDP buffer size and send oversized responses
. Servers must send UDP responses
no larger than the advertised EDNS UDP buffer size.
DO=1 Handling
Some nameservers incorrectly only return an EDNS response
when the DNSSEC OK (DO) bit is 1 in the
query. Servers that support EDNS should always respond to
EDNS requests with EDNS responses.
Some nameservers fail to copy the DO bit to the response
despite clearly supporting DNSSEC by returning an RRSIG
records to EDNS queries with DO=1. Nameservers that support
DNSSEC are expected to copy the DO bit from the request to
the response.
EDNS over TCP
Some EDNS-aware servers incorrectly limit the TCP response
sizes to the advertised UDP response size. This breaks DNS
resolution to clients where the response sizes exceed the
advertised UDP response size despite the server and the client
being capable of sending and receiving larger TCP responses,
respectively. It effectively defeats setting TC=1 in UDP
responses.
Firewalls and Load Balancers
Firewalls and load balancers can affect the externally
visible behaviour of a nameserver. Tests for conformance
should to be done from outside of any firewall so that the
system is tested as a whole.
Firewalls and load balancers should not drop DNS packets
that they don't understand. They should either pass
the packets or generate an appropriate error response.
Requests for unknown query types are normal client behaviour
and should not be construed as an attack. Nameservers have
always been expected to be able to handle such queries.
Requests for unknown query classes are normal client behaviour
and should not be construed as an attack. Nameservers have
always been expected to be able to handle such queries.
Requests with unknown opcodes are normal client behaviour
and should not be construed as an attack. Nameservers have
always been expected to be able to handle such queries.
Requests with unassigned flags set (DNS or EDNS) are expected client
behaviour and should not be construed as an attack. The behaviour for
unassigned flags is to ignore them in the request and to not set them in
the response. Dropping DNS/EDNS packets with unassigned flags makes it
difficult to deploy extensions that make use of them due to the need to
reconfigure and update firewalls.
Requests with unknown EDNS options are expected client
behaviour and should not be construed as an attack. The
correct behaviour for unknown EDNS options is to ignore
their presence when constructing a reply.
Requests with unknown EDNS versions are expected client
behaviour and should not be construed as an attack. The
correct behaviour for unknown EDNS versions is to return
BADVERS along with the highest EDNS version the server
supports. Dropping EDNS packets breaks EDNS version
negotiation.
Firewalls should not assume that there will only be a single
response message to a request. There have been proposals
to use EDNS to signal that multiple DNS messages be returned
rather than a single UDP message that is fragmented at the
IP layer.
DNS, and EDNS in particular, are designed to allow clients to
be able to use new features against older servers without
having to validate every option. Indiscriminate blocking
of messages breaks that design.
However, there may be times when a nameserver mishandles
messages with a particular flag, EDNS option, EDNS version
field, opcode, type or class field, or combination thereof
to the point where the integrity of the nameserver is
compromised. Firewalls should offer the ability to selectively
reject messages using an appropriately constructed response
based on all these fields while awaiting a fix from the
nameserver vendor. Returning FORMERR or REFUSED are two
potential error codes to return.
Packet Scrubbing Services
Packet scrubbing services are used to filter out undesired traffic,
including but not limited to denial-of-service traffic. This is
often done using heuristic analysis of the traffic.
Packet scrubbing services can affect the externally visible behaviour
of a nameserver in a similar way to firewalls. If an operator
uses a packet scrubbing service, they should check that legitimate
queries are not being blocked.
Packet scrubbing services, unlike firewalls, are also turned on
and off in response to denial-of-service attacks. One needs
to take care when choosing a scrubbing service.
Ideally, operators should run these tests against a packet scrubbing
service to ensure that these tests are not seen as attack
vectors.
Whole Answer Caches
Whole answer caches take a previously constructed answer
and return it to a subsequent query for the same question.
However, they can return the wrong response if they do not
take all of the relevant attributes of the query into account.
In addition to the standard tuple of <qname,qtype,qclass>,
a non-exhaustive set of attributes that must be considered
include: RD, AD, CD, OPT record, DO, EDNS buffer size, EDNS
version, EDNS options, and transport.
Response Code Selection
Choosing the correct response code when responding to DNS
queries is important. Response codes should be chosen
considering how clients will handle them.
For unimplemented opcodes, NOTIMP is the expected response code.
Note: newly implemented opcodes may change the message format by
extending the header, changing the structure of the records,
etc. Servers are not expected to be able to parse these
and should respond with a response code of NOTIMP rather
than FORMERR (which would be expected if there was a parse error
with a known opcode).
For unimplemented type codes, and in the absence of other
errors, the only valid response is NOERROR if the qname
exists and NXDOMAIN otherwise.
For Meta-RRs, NOTIMP may be returned instead.
If a zone cannot be loaded because it contains unimplemented type
codes that are not encoded as unknown record types according to , then the expected response is
SERVFAIL, as the whole zone should be rejected (). If a zone loads, then applies.
If the server supports EDNS and receives a query with an
unsupported EDNS version, the correct response is BADVERS
.
If the server does not support EDNS at all, FORMERR is
the expected error code. That said, a minimal EDNS
server implementation requires parsing the OPT records
and responding with an empty OPT record in the additional
section in most cases. There is no need to interpret any
EDNS options present in the request, as unsupported EDNS
options are expected to be ignored .
Additionally, EDNS flags can be ignored. The only part of the
OPT record that needs to be examined is the version field
to determine if BADVERS needs to be sent or not.
Testing
Testing is divided into two sections: "Basic DNS", which all
servers should meet, and "Extended DNS", which should be met
by all servers that support EDNS (a server is deemed to
support EDNS if it gives a valid EDNS response to any EDNS
query). If a server does not support EDNS, it should still
respond to all the tests, albeit with error responses.
These tests query for records at the apex of a zone that the
server is nominally configured to serve. All tests should
use the same zone.
It is advisable to run all of the tests below in parallel
so as to minimise the delays due to multiple timeouts when
the servers do not respond. There are 16 queries directed
to each nameserver (assuming no packet loss) testing different
aspects of Basic DNS and Extended DNS.
The tests below use dig from BIND 9.11.0 .
Replace $zone with the name of the zone being used for testing.
Replace $server with the name or address of the server being tested.
When testing, recursive servers set RD=1 and choose a zone
name that is known to exist and is not being served by the
recursive server. The root zone (".") is often a good
candidate, as it is DNSSEC signed. RD=1, rather than RD=0,
should be present in the responses for all test involving
the opcode QUERY. Non-authoritative answers (AA=0) are
expected when talking to a recursive server. AD=1 is only
expected if the server is validating responses and
one or both AD=1 or DO=1 is set in the request, otherwise
AD=0 is expected.
Testing: Basic DNS
This first set of tests cover Basic DNS server behaviour
and all servers should pass these tests.
Is the server configured for the zone?
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set and without
EDNS.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the Authoritative Answer (AA) and Query/Response (QR) bits to be
set in the header; the Recursion Available (RA) bits may also be
set . We
do not expect an OPT record to be returned .
Verify the server is configured for the zone:
Testing Unknown Types
Identifying servers that fail to respond to unknown or
unsupported types can be done by making an initial DNS
query for an A record, making a number of queries for an
unallocated type, then making a query for an A record
again. IANA maintains a registry of allocated types .
If the server responds to the first and last queries but
fails to respond to the queries for the unallocated type,
it is probably faulty. The test should be repeated a
number of times to eliminate the likelihood of a false
positive due to packet loss.
Ask for the TYPE1000 RRset at the configured zone's
name. This query is made with no DNS flag bits set and
without EDNS. TYPE1000 has been chosen for this purpose,
as IANA is unlikely to allocate this type in the near
future and it is not in a range reserved for private
use . Any unallocated type code
could be chosen for this test.
We expect no records to be returned in the answer
section, the rcode to be set to NOERROR, and the AA and
QR bits to be set in the header; RA may also be set
. We do not expect an OPT record
to be returned .
Check that queries for an unknown type work:
Testing Header BitsTesting CD=1 Queries
Ask for the SOA record of the configured zone.
This query is made with only the CD DNS flag bit set,
with all other DNS bits clear, and without EDNS.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR,
and the AA and QR bits to be set in the header. We
do not expect an OPT record to be returned.
If the server supports DNSSEC, CD should be set in
the response ; otherwise, CD
should be clear .
Check that queries with CD=1 work:
Testing AD=1 QueriesAsk for the SOA record of the configured zone. This query is
made with only the AD DNS flag bit set, with all other DNS bits clear,
and without EDNS.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR,
and the AA and QR bits to be set in the header. We
do not expect an OPT record to be returned. The
purpose of this query is to detect blocking of queries
with the AD bit present, not the specific value of
AD in the response.
Check that queries with AD=1 work:
AD use in queries is defined in .
Testing Reserved Bit
Ask for the SOA record of the configured zone. This
query is made with only the final reserved DNS flag
bit set, with all other DNS bits clear, and without EDNS.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR,
and the AA and QR bits to be set in the header; RA
may be set. The final reserved bit must not be set
. We do not expect an OPT
record to be returned .
Check that queries with the last unassigned DNS
header flag work and that the flag bit is not
copied to the response:
MBZ (Must Be Zero) is a dig-specific
indication that the flag bit has been incorrectly copied. See
:
"Z
Reserved for future use. Must be zero in all queries and responses."
Testing Recursive Queries
Ask for the SOA record of the configured zone. This
query is made with only the RD DNS flag bit set and
without EDNS.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA, QR and RD bits to be set in the header; RA
may also be set . We do not
expect an OPT record to be returned .
Check that recursive queries work:
Testing Unknown Opcodes
Construct a DNS message that consists of only a DNS
header with opcode set to 15 (currently not allocated),
no DNS header bits set, and empty question, answer,
authority, and additional sections.
Check that new opcodes are handled:
Testing TCP
Whether a server accepts TCP connections can be tested
by first checking that it responds to UDP queries to
confirm that it is up and operating, then attempting the
same query over TCP. An additional query should be made
over UDP if the TCP connection attempt fails to confirm
that the server under test is still operating.
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set and without
EDNS. This query is to be sent using TCP.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA and QR bits to be set in the header; RA may
also be set . We do not expect
an OPT record to be returned .
Check that TCP queries work:
The requirement that TCP be supported is defined
in .
Testing: Extended DNS
The next set of tests cover various aspects of EDNS
behaviour. If any of these tests succeed (indicating
at least some EDNS support), then all of them should
succeed. There are servers that support EDNS but fail
to handle plain EDNS queries correctly, so a plain EDNS
query is not a good indicator of lack of EDNS support.
Testing Minimal EDNS
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set. EDNS version
0 is used without any EDNS options or EDNS flags set.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA and QR bits to be set in the header; RA may
also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response. The EDNS version field
should be 0, and there should be no EDNS options present
.
Check that plain EDNS queries work:
+nocookie disables sending an EDNS COOKIE option, which
is otherwise enabled by default in BIND 9.11.0 (and
later).
Testing EDNS Version Negotiation
Ask for the SOA record of a zone the server is nominally
configured to serve. This query is made with no DNS flag
bits set. EDNS version 1 is used without any EDNS options
or EDNS flags set.
We expect the SOA record for the zone to NOT be returned
in the answer section with the extended rcode set to
BADVERS and the QR bit to be set in the header; RA
may also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response. The EDNS version field
should be 0 in the response, as no other EDNS version
has as yet been specified .
Check that EDNS version 1 queries work (EDNS supported):
+noednsneg has been set, as dig supports EDNS version
negotiation, and we want to see only the response to the
initial EDNS version 1 query.
Testing Unknown EDNS Options
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set. EDNS version
0 is used without any EDNS flags. An EDNS option is
present with a value that has not yet been assigned by
IANA. We have picked an unassigned code of 100 for
the example below. Any unassigned EDNS option code
could have been chosen for this test.
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA and QR bits to be set in the header; RA may
also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response. The EDNS version field
should be 0, as EDNS versions other than 0 are yet to
be specified, and there should be no EDNS options present,
as unknown EDNS options are supposed to be ignored by the
server ().
Check that EDNS queries with an unknown option work (EDNS supported):
Testing Unknown EDNS Flags
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set. EDNS version
0 is used without any EDNS options. An unassigned EDNS
flag bit is set (0x40 in this case).
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA and QR bits to be set in the header; RA may
also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response, as unknown EDNS flags are
supposed to be ignored. The EDNS version field
should be 0, and there should be no EDNS options present
.
Check that EDNS queries with unknown flags work (EDNS supported):
MBZ (Must Be Zero) is a dig-specific indication that
a flag bit has been incorrectly copied, as per
.
Testing EDNS Version Negotiation with Unknown EDNS Flags
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set. EDNS version
1 is used without any EDNS options. An unassigned EDNS
flag bit is set (0x40 in this case).
We expect the SOA record for the zone to NOT be returned
in the answer section with the extended rcode set to
BADVERS and the QR bit to be set in the header; RA
may also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response, as unknown EDNS flags are
supposed to be ignored. The EDNS version field should
be 0, as EDNS versions other than 0 are yet to be
specified, and there should be no EDNS options present
.
Check that EDNS version 1 queries with unknown flags work (EDNS supported):
Testing EDNS Version Negotiation with Unknown EDNS Options
Ask for the SOA record of the configured zone. This
query is made with no DNS flag bits set. EDNS version
1 is used. An unknown EDNS option is present. We have
picked an unassigned code of 100 for the example below.
Any unassigned EDNS option code could have been chosen for
this test.
We expect the SOA record for the zone to NOT be returned
in the answer section with the extended rcode set to
BADVERS and the QR bit to be set in the header; RA
may also be set . We expect
an OPT record to be returned. There should be no EDNS
flags present in the response. The EDNS version field
should be 0, as EDNS versions other than 0 are yet
to be specified, and there should be no EDNS options
present .
Check that EDNS version 1 queries with unknown options work (EDNS supported):
Testing Truncated Responses
Ask for the DNSKEY records of the configured zone, which
must be a DNSSEC signed zone. This query is made with
no DNS flag bits set. EDNS version 0 is used without
any EDNS options. The only EDNS flag set is DO. The
EDNS UDP buffer size is set to 512. The intention of
this query is to elicit a truncated response from the
server. Most signed DNSKEY responses are bigger than
512 bytes. This test will not give a valid result if
the zone is not signed.
We expect a response, the rcode to be set to NOERROR, and the AA
and QR bits to be set. AD may be set in the response if the server
supports DNSSEC; otherwise it should be clear; TC and RA may also be
set . We expect an OPT record to be present in the
response. There should be no EDNS flags other than DO present in
the response. The EDNS version field should be 0, and there should
be no EDNS options present .
If TC is not set, it is not possible to confirm that the
server correctly adds the OPT record to the truncated
responses or not.
Testing DO=1 Handling
Ask for the SOA record of the configured zone, which
does not need to be DNSSEC signed. This query is made
with no DNS flag bits set. EDNS version 0 is used
without any EDNS options. The only EDNS flag set is
DO.
We expect the SOA record for the zone to be returned in the
answer section, the rcode to be set to NOERROR, and the AA and QR
bits to be set in the response. AD may be set in the response if the
server supports DNSSEC, otherwise it should be clear; RA may also be
set . We expect an OPT
record to be returned. There should be no EDNS flags other than DO
present in the response, which should be present if the server
supports DNSSEC. The EDNS version field should be 0, and there
should be no EDNS options present .
Check that DO=1 queries work (EDNS supported):
Testing EDNS Version Negotiation with DO=1
Ask for the SOA record of the configured zone, which does
not need to be DNSSEC signed. This query is made with no
DNS flag bits set. EDNS version 1 is used without any EDNS
options. The only EDNS flag set is DO.
We expect the SOA record for the zone NOT to be returned in the answer
section, the extended rcode to be set to BADVERS, and the QR bit to be set in
the header; RA may also be set .
We expect an OPT record to be returned. There
should be no EDNS flags other than DO present in the response, which
should be there if the server supports DNSSEC. The EDNS version
field should be 0, and there should be no EDNS options present .
Check that EDNS version 1, DO=1 queries work (EDNS supported):
Testing with Multiple Defined EDNS OptionsAsk for the SOA record of the configured zone. This query is
made with no DNS flag bits set. EDNS version 0 is used. A number
of defined EDNS options are present (NSID , DNS COOKIE , EDNS Client Subnet , and EDNS Expire ).
We expect the SOA record for the zone to be returned
in the answer section, the rcode to be set to NOERROR, and
the AA and QR bits to be set in the header; RA may
also be set . We expect an OPT
record to be returned. There should be no EDNS flags
present in the response. The EDNS version field should
be 0. Any of the requested EDNS options supported
by the server and permitted server configuration may
be returned .
Check that EDNS queries with multiple defined EDNS options work:
When EDNS Is Not Supported
If EDNS is not supported by the nameserver, we expect a
response to each of the above queries. That response may
be a FORMERR error response, or the OPT record may just
be ignored.
Some nameservers only return an EDNS response when a
particular EDNS option or flag (e.g., DO=1) is present in
the request. This behaviour is not compliant behaviour
and may hide other incorrect behaviour from the above
tests. Retesting with the triggering option/flag
present will expose this misbehaviour.
Remediation
Nameserver operators are generally expected to test their
own infrastructure for compliance to standards. The above
tests should be run when new systems are brought online
and should be repeated periodically to ensure continued
interoperability.
Domain registrants who do not maintain their own DNS
infrastructure are entitled to a DNS service that conforms
to standards and interoperates well. Registrants who become
aware that their DNS operator does not have a well-maintained
or compliant infrastructure should insist that their service
provider correct issues and switch providers if they do
not.
In the event that an operator experiences problems due to
the behaviour of nameservers outside their control, the
above tests will help in narrowing down the precise issue(s),
which can then be reported to the relevant party.
If contact information for the operator of a misbehaving
nameserver is not already known, the following methods of
communication could be considered:
the RNAME of the zone authoritative for the name of the
misbehaving server
the RNAME of zones for which the offending server is
authoritative
administrative or technical contacts listed in the
registration information for the parent domain of the
name of the misbehaving server or for zones for which
the nameserver is authoritative
the registrar or registry for such zones
DNS-specific, operational fora (e.g., mailing lists)
Operators of parent zones may wish to regularly test the
authoritative nameservers of their child zones. However,
parent operators can have widely varying capabilities in
terms of notification or remediation depending on whether
they have a direct relationship with the child operator.
Many Top-Level Domain (TLD) registries, for example, cannot directly contact
their registrants and may instead need to communicate through
the relevant registrar. In such cases, it may be most
efficient for registrars to take on the responsibility for
testing the nameservers of their registrants, since they
have a direct relationship.
When notification is not effective at correcting problems
with a misbehaving nameserver, parent operators can choose
to remove NS record sets (and glue records below) that refer
to the faulty server until the servers are fixed. This should
only be done as a last resort and with due consideration, as
removal of a delegation can have unanticipated side effects.
For example, other parts of the DNS tree may depend on names
below the removed zone cut, and the parent operator may find
themselves responsible for causing new DNS failures to occur.
Security Considerations
Testing protocol compliance can potentially result in false
reports of attempts to attack services from Intrusion Detection
Services and firewalls. All of the tests are well-formed
(though not necessarily common) DNS queries. None of the tests
listed above should cause any harm to a protocol-compliant
server.
Relaxing firewall settings to ensure EDNS compliance could
potentially expose a critical implementation flaw in the
nameserver. Nameservers should be tested for conformance
before relaxing firewall settings.
When removing delegations for non-compliant servers, there
can be a knock-on effect on other zones that require these
zones to be operational for the nameservers addresses to be
resolved.
IANA ConsiderationsThis document has no IANA actions.ReferencesNormative ReferencesInformative ReferencesInternet Systems ConsortuimDomain Name System (DNS) ParametersIANAAcknowledgementsThe contributions of and
are gratefully acknowledged.