draft-ietf-dnsop-serverid-06.txt

来自「非常好的dns解析软件」· 文本 代码 · 共 619 行 · 第 1/2 页

Network Working Group                                           S. Woolf
Internet-Draft                         Internet Systems Consortium, Inc.
Expires: September 6, 2006                                     D. Conrad
                                                           Nominum, Inc.
                                                           March 5, 2006


    Requirements for a Mechanism Identifying a Name Server Instance
                      draft-ietf-dnsop-serverid-06

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   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
   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 6, 2006.

Copyright Notice

   Copyright (C) The Internet Society (2006).

Abstract

   With the increased use of DNS anycast, load balancing, and other
   mechanisms allowing more than one DNS name server to share a single
   IP address, it is sometimes difficult to tell which of a pool of name
   servers has answered a particular query.  A standardized mechanism to
   determine the identity of a name server responding to a particular
   query would be useful, particularly as a diagnostic aid for
   administrators.  Existing ad hoc mechanisms for addressing this need



Woolf & Conrad          Expires September 6, 2006               [Page 1]

Internet-Draft                  Serverid                      March 2006


   have some shortcomings, not the least of which is the lack of prior
   analysis of exactly how such a mechanism should be designed and
   deployed.  This document describes the existing convention used in
   some widely deployed implementations of the DNS protocol, including
   advantages and disadvantages, and discusses some attributes of an
   improved mechanism.













































Woolf & Conrad          Expires September 6, 2006               [Page 2]

Internet-Draft                  Serverid                      March 2006


1.  Introduction and Rationale

   Identifying which name server is responding to queries is often
   useful, particularly in attempting to diagnose name server
   difficulties.  This is most obviously useful for authoritative
   nameservers in the attempt to diagnose the source or prevalence of
   inaccurate data, but can also conceivably be useful for caching
   resolvers in similar and other situations.  Furthermore, the ability
   to identify which server is responding to a query has become more
   useful as DNS has become more critical to more Internet users, and as
   network and server deployment topologies have become more complex.

   The traditional means for determining which of several possible
   servers is answering a query has traditionally been based on the use
   of the server's IP address as a unique identifier.  However, the
   modern Internet has seen the deployment of various load balancing,
   fault-tolerance, or attack-resistance schemes such as shared use of
   unicast IP addresses as documented in [RFC3258].  An unfortunate side
   effect of these schemes has been to make the use of IP addresses as
   identifiers somewhat problematic.  Specifically, a dedicated DNS
   query may not go to the same server as answered a previous query,
   even though sent to the same IP address.  Non-DNS methods such as
   ICMP ping, TCP connections, or non-DNS UDP packets (such as those
   generated by tools like "traceroute"), etc., may well be even less
   certain to reach the same server as the one which receives the DNS
   queries.

   There is a well-known and frequently-used technique for determining
   an identity for a nameserver more specific than the possibly-non-
   unique "server that answered the query I sent to IP address XXX".
   The widespread use of the existing convention suggests a need for a
   documented, interoperable means of querying the identity of a
   nameserver that may be part of an anycast or load-balancing cluster.
   At the same time, however, it also has some drawbacks that argue
   against standardizing it as it's been practiced so far.
















Woolf & Conrad          Expires September 6, 2006               [Page 3]

Internet-Draft                  Serverid                      March 2006


2.  Existing Conventions

   For some time, the commonly deployed Berkeley Internet Name Domain
   implementation of the DNS protocol suite from the Internet Systems
   Consortium [BIND] has supported a way of identifying a particular
   server via the use of a standards-compliant, if somewhat unusual, DNS
   query.  Specifically, a query to a recent BIND server for a TXT
   resource record in class 3 (CHAOS) for the domain name
   "HOSTNAME.BIND." will return a string that can be configured by the
   name server administrator to provide a unique identifier for the
   responding server.  (The value defaults to the result of a
   gethostname() call).  This mechanism, which is an extension of the
   BIND convention of using CHAOS class TXT RR queries to sub-domains of
   the "BIND." domain for version information, has been copied by
   several name server vendors.

   A refinement to the BIND-based mechanism, which dropped the
   implementation-specific string, replaces ".BIND" with ".SERVER".
   Thus the query string to learn the unique name of a server may be
   queried as "ID.SERVER".

   (For reference, the other well-known name used by recent versions of
   BIND within the CHAOS class "BIND." domain is "VERSION.BIND."  A
   query for a CHAOS TXT RR for this name will return an
   administratively defined string which defaults to the version of the
   server responding.  This is, however, not generally implemented by
   other vendors.)

2.1.  Advantages

   There are several valuable attributes to this mechanism, which
   account for its usefulness.

   1.  The "HOSTNAME.BIND" or "ID.SERVER" query response mechanism is
       within the DNS protocol itself.  An identification mechanism that
       relies on the DNS protocol is more likely to be successful
       (although not guaranteed) in going to the same system as a
       "normal" DNS query.

   2.  Since the identity information is requested and returned within
       the DNS protocol, it doesn't require allowing any other query
       mechanism to the server, such as holes in firewalls for
       otherwise-unallowed ICMP Echo requests.  Thus it is likely to
       reach the same server over a path subject to the same routing,
       resource, and security policy as the query, without any special
       exceptions to site security policy.





Woolf & Conrad          Expires September 6, 2006               [Page 4]

Internet-Draft                  Serverid                      March 2006


   3.  It is simple to configure.  An administrator can easily turn on
       this feature and control the results of the relevant query.

   4.  It allows the administrator complete control of what information
       is given out in the response, minimizing passive leakage of
       implementation or configuration details.  Such details are often
       considered sensitive by infrastructure operators.

   5.  Hypothetically, since it's an ordinary DNS record and the
       relevant DNSSEC RRs are class independent, the id.server response
       RR could be signed, which has the advantages described in
       [RFC4033].

2.2.  Disadvantages

   At the same time, there are some serious drawbacks to the CHAOS/TXT
   query mechanism that argue against standardizing it as it currently
   operates.

   1.  It requires an additional query to correlate between the answer
       to a DNS query under normal conditions and the supposed identity
       of the server receiving the query.  There are a number of
       situations in which this simply isn't reliable.

   2.  It reserves an entire class in the DNS (CHAOS) for what amounts
       to one zone.  While CHAOS class is defined in [RFC1034] and
       [RFC1035], it's not clear that supporting it solely for this
       purpose is a good use of the namespace or of implementation
       effort.

   3.  The initial and still common form, using .BIND, is implementation
       specific.  BIND is one DNS implementation.  At the time of this
       writing, it is probably the most prevalent for authoritative
       servers.  This does not justify standardizing on its ad hoc
       solution to a problem shared across many operators and
       implementors.  Meanwhile, the proposed refinement changes the
       string but preserves the ad hoc CHAOS/TXT mechanism.

   4.  There is no convention or shared understanding of what
       information an answer to such a query for a server identity could
       or should include, including a possible encoding or
       authentication mechanism.

   The first of the listed disadvantages may be technically the most
   serious.  It argues for an attempt to design a good answer to the
   problem that "I need to know what nameserver is answering my
   queries", not simply a convenient one.




Woolf & Conrad          Expires September 6, 2006               [Page 5]

Internet-Draft                  Serverid                      March 2006


2.3.  Characteristics of an Implementation Neutral Convention

   The discussion above of advantages and disadvantages to the
   HOSTNAME.BIND mechanism suggest some requirements for a better
   solution to the server identification problem.  These are summarized
   here as guidelines for any effort to provide appropriate protocol
   extensions:

   1.  The mechanism adopted must be in-band for the DNS protocol.  That
       is, it needs to allow the query for the server's identifying
       information to be part of a normal, operational query.  It should
       also permit a separate, dedicated query for the server's
       identifying information.  But it should preserve the ability of
       the CHAOS/TXT query-based mechanism to work through firewalls and
       in other situations where only DNS can be relied upon to reach
       the server of interest.

   2.  The new mechanism should not require dedicated namespaces or
       other reserved values outside of the existing protocol mechanisms
       for these, i.e. the OPT pseudo-RR.  In particular, it should not
       propagate the existing drawback of requiring support for a CLASS
       and top level domain in the authoritative server (or the querying
       tool) to be useful.

   3.  Support for the identification functionality should be easy to
       implement and easy to enable.  It must be easy to disable and