RFC 9144 PROPOSED STANDARD NETWORK MANAGEMENT · YANG · NETCONF

Comparison of Network Management Datastore Architecture (NMDA) Datastores

A NETCONF and RESTCONF operation that asks a device to diff two of its datastores, such as intended versus operational.

On rfc-editor.org ↗
PUBLISHED
December 2021
AUTHORS
A. Clemm, Y. Qu, J. Tantsura, A. Bierman
STREAM
IETF
OPERATION
compare (RPC)
YANG MODULE
ietf-nmda-compare
MODULE PREFIX
cmp
OUTPUT FORMAT
YANG Patch (RFC 8072)
TRANSPORTS
NETCONF or RESTCONF
DOI
10.17487/RFC9144

RFC text

Contents
Original text · RFC Editor View on rfc-editor.org ↗
Internet Engineering Task Force (IETF)                          A. Clemm
Request for Comments: 9144                                         Y. Qu
Category: Standards Track                                      Futurewei
ISSN: 2070-1721                                              J. Tantsura
                                                               Microsoft
                                                              A. Bierman
                                                               YumaWorks
                                                           December 2021


     Comparison of Network Management Datastore Architecture (NMDA)
                               Datastores

Abstract

   This document defines a Remote Procedure Call (RPC) operation to
   compare management datastores that comply with the Network Management
   Datastore Architecture (NMDA).

Status of This Memo

   This is an Internet Standards Track document.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Further information on
   Internet Standards is available in Section 2 of RFC 7841.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   https://www.rfc-editor.org/info/rfc9144.

Copyright Notice

   Copyright (c) 2021 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (https://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Revised BSD License text as described in Section 4.e of the
   Trust Legal Provisions and are provided without warranty as described
   in the Revised BSD License.

Table of Contents

   1.  Introduction
   2.  Key Words
   3.  Data Model Overview
   4.  YANG Data Model
   5.  Example
   6.  Performance Considerations
   7.  IANA Considerations
     7.1.  Update to the IETF XML Registry
     7.2.  Update to the YANG Module Names Registry
   8.  Security Considerations
   9.  References
     9.1.  Normative References
     9.2.  Informative References
   Appendix A.  Possible Future Extensions
   Acknowledgments
   Authors' Addresses

1.  Introduction

   The revised NMDA [RFC8342] introduces a set of new datastores that
   each hold YANG-defined data [RFC7950] and represent a different
   "viewpoint" on the data that is maintained by a server.  New YANG
   datastores that are introduced include <intended>, which contains
   validated configuration data that a client application intends to be
   in effect, and <operational>, which contains operational state data
   (such as statistics) as well as configuration data that is actually
   in effect.

   NMDA introduces, in effect, a concept of "lifecycle" for management
   data, distinguishing between data that is part of a configuration
   that was supplied by a user, configuration data that has actually
   been successfully applied and that is part of the operational state,
   and the overall operational state that includes applied configuration
   data as well as status and statistics.

   As a result, data from the same management model can be reflected in
   multiple datastores.  Clients need to specify the target datastore to
   be specific about which viewpoint of the data they want to access.
   For example, a client application can differentiate whether they are
   interested in the configuration that is supplied to a server and is
   supposed to be in effect or the configuration that has been applied
   and is actually in effect on the server.

   Due to the fact that data can propagate from one datastore to
   another, it is possible for differences between datastores to occur.
   Some of this is entirely expected, as there may be a time lag between
   when a configuration is given to the device and reflected in
   <intended> until when it actually takes effect and is reflected in
   <operational>.  However, there may be cases when a configuration item
   that was to be applied may not actually take effect at all or needs
   an unusually long time to do so.  This can be the case due to certain
   conditions not being met, certain parts of the configuration not
   propagating because they are considered inactive, resource
   dependencies not being resolved, or even implementation errors in
   corner conditions.

   When the configuration that is in effect is different from the
   configuration that was applied, many issues can result.  It becomes
   more difficult to operate the network properly due to limited
   visibility of the actual operational status, which makes it more
   difficult to analyze and understand what is going on in the network.
   Services may be negatively affected (for example, degrading or
   breaking a customer service), and network resources may be
   misallocated.

   Applications can potentially analyze any differences between two
   datastores by retrieving the contents from both datastores and
   comparing them.  However, in many cases, this will be both costly and
   extremely wasteful.

   This document introduces a YANG data model that defines RPCs intended
   to be used in conjunction with NETCONF [RFC6241] or RESTCONF
   [RFC8040].  These RPCs allow a client to request a server to compare
   two NMDA datastores and report any differences.

2.  Key Words

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
   "OPTIONAL" in this document are to be interpreted as described in
   BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
   capitals, as shown here.

3.  Data Model Overview

   The core of the solution is a new management operation, <compare>,
   that compares the data tree contents of two datastores.  The
   operation checks whether there are any differences in values or in
   data nodes that are contained in either datastore and returns any
   differences as output.  The output is returned in the format
   specified in YANG Patch [RFC8072].

   The YANG data model defines the <compare> operation as a new RPC.
   The operation takes the following input parameters:

   source:  The source identifies the datastore to serve as the
      reference for the comparison -- for example, <intended>.

   target:  The target identifies the datastore to compare against the
      source -- for example, <operational>.

   filter-spec:  This is a choice between different filter constructs to
      identify the parts of the datastore to be retrieved.  It acts as a
      node selector that specifies which data nodes are within the scope
      of the comparison and which nodes are outside the scope.  This
      allows a comparison operation to be applied only to a specific
      part of the datastore that is of interest, such as a particular
      subtree.  Note that the filter does not allow expressions that
      match against data node values, since this may incur
      implementation difficulties and is not required for normal use
      cases.

   all:  When set, this parameter indicates that all differences should
      be included, including differences pertaining to schema nodes that
      exist in only one of the datastores.  When this parameter is not
      included, a prefiltering step is automatically applied to exclude
      data from the comparison that does not pertain to both datastores:
      if the same schema node is not present in both datastores, then
      all instances of that schema node and all its descendants are
      excluded from the comparison.  This allows client applications to
      focus on the differences that constitute true mismatches of
      instance data without needing to specify more complex filter
      constructs.

   report-origin:  When set, this parameter indicates that origin
      metadata should be included as part of RPC output.  When this
      parameter is omitted, origin metadata in comparisons that involve
      <operational> is by default omitted.  Note that origin metadata
      only applies to <operational>; it is therefore also omitted in
      comparisons that do not involve <operational> regardless of
      whether or not the parameter is set.

   The operation provides the following output parameter:

   differences:  This parameter contains the list of differences.  Those
      differences are encoded per the YANG Patch data model defined in
      [RFC8072].  When a datastore node in the source of the comparison
      is not present in the target of the comparison, this can be
      indicated either as a "delete" or as a "remove" in the patch as
      there is no differentiation between those operations for the
      purposes of the comparison.  The YANG Patch data model is
      augmented to indicate the value of source datastore nodes in
      addition to the patch itself that would need to be applied to the
      source to produce the target.  When the target datastore is
      <operational> and the input parameter "report-origin" is set,
      origin metadata is included as part of the patch.  Including
      origin metadata can help explain the cause of a difference in some
      cases -- for example, when a data node is part of <intended> but
      the origin of the same data node in <operational> is reported as
      "system".

   The data model is defined in the ietf-nmda-compare YANG module.  Its
   structure is shown in the following figure.  The notation syntax
   follows [RFC8340].

   module: ietf-nmda-compare
     rpcs:
       +---x compare
          +---w input
          |  +---w source            identityref
          |  +---w target            identityref
          |  +---w all?              empty
          |  +---w report-origin?    empty
          |  +---w (filter-spec)?
          |     +--:(subtree-filter)
          |     |  +---w subtree-filter?
          |     +--:(xpath-filter)
          |        +---w xpath-filter?     yang:xpath1.0 {nc:xpath}?
          +--ro output
             +--ro (compare-response)?
                +--:(no-matches)
                |  +--ro no-matches?    empty
                +--:(differences)
                   +--ro differences
                      +--ro yang-patch
                         +--ro patch-id    string
                         +--ro comment?    string
                         +--ro edit* [edit-id]
                            +--ro edit-id         string
                            +--ro operation       enumeration
                            +--ro target          target-resource-offset
                            +--ro point?          target-resource-offset
                            +--ro where?          enumeration
                            +--ro value?
                            +--ro source-value?

                  Figure 1: Structure of ietf-nmda-compare

4.  YANG Data Model

   This YANG module includes references to [RFC6991], [RFC8342],
   [RFC8072], and [RFC6241].

   <CODE BEGINS> file "ietf-nmda-compare@2021-12-10.yang"
   module ietf-nmda-compare {
     yang-version 1.1;
     namespace "urn:ietf:params:xml:ns:yang:ietf-nmda-compare";
     prefix cmp;

     import ietf-yang-types {
       prefix yang;
       reference
         "RFC 6991: Common YANG Data Types";
     }
     import ietf-datastores {
       prefix ds;
       reference
         "RFC 8342: Network Management Datastore
          Architecture (NMDA)";
     }
     import ietf-yang-patch {
       prefix ypatch;
       reference
         "RFC 8072: YANG Patch Media Type";
     }
     import ietf-netconf {
       prefix nc;
       reference
         "RFC 6241: Network Configuration Protocol (NETCONF)";
     }

     organization
       "IETF NETMOD (Network Modeling) Working Group";
     contact
       "WG Web:   <https://datatracker.ietf.org/wg/netmod/>
        WG List:  <mailto:netmod@ietf.org>

        Author: Alexander Clemm
                <mailto:ludwig@clemm.org>

        Author: Yingzhen Qu
                <mailto:yqu@futurewei.com>

        Author: Jeff Tantsura
                <mailto:jefftant.ietf@gmail.com>

        Author: Andy Bierman
                <mailto:andy@yumaworks.com>";
     description
       "The YANG data model defines a new operation, <compare>, that
        can be used to compare NMDA datastores.

        Copyright (c) 2021 IETF Trust and the persons identified as
        authors of the code.  All rights reserved.

        Redistribution and use in source and binary forms, with or
        without modification, is permitted pursuant to, and subject to
        the license terms contained in, the Revised BSD License set
        forth in Section 4.c of the IETF Trust's Legal Provisions
        Relating to IETF Documents
        (https://trustee.ietf.org/license-info).

        This version of this YANG module is part of RFC 9144; see the
        RFC itself for full legal notices.";

     revision 2021-12-10 {
       description
         "Initial revision.";
       reference
         "RFC 9144: Comparison of Network Management Datastore
          Architecture (NMDA) Datastores";
     }

     /* RPC */

     rpc compare {
       description
         "NMDA datastore compare operation.";
       input {
         leaf source {
           type identityref {
             base ds:datastore;
           }
           mandatory true;
           description
             "The source datastore to be compared.";
         }
         leaf target {
           type identityref {
             base ds:datastore;
           }
           mandatory true;
           description
             "The target datastore to be compared.";
         }
         leaf all {
           type empty;
           description
             "When this leaf is provided, all data nodes are compared,
              whether their schema node pertains to both datastores or
              not.  When this leaf is omitted, a prefiltering step is
              automatically applied that excludes data nodes from the
              comparison that can occur in only one datastore but not
              the other.  Specifically, if one of the datastores
              (source or target) contains only configuration data and
              the other datastore is <operational>, data nodes for
              the config that is false are excluded from the
              comparison.";
         }
         leaf report-origin {
           type empty;
           description
             "When this leaf is provided, origin metadata is
              included as part of RPC output.  When this leaf is
              omitted, origin metadata in comparisons that involve
              <operational> is by default omitted.";
         }
         choice filter-spec {
           description
             "Identifies the portions of the datastores to be
                  compared.";
           anydata subtree-filter {
             description
               "This parameter identifies the portions of the
                target datastore to retrieve.";
             reference
               "RFC 6241, Section 6.";
           }
           leaf xpath-filter {
             if-feature "nc:xpath";
             type yang:xpath1.0;
             description
               "This parameter contains an XPath expression
                identifying the portions of the target
                datastore to retrieve.";
             reference
               "RFC 6991: Common YANG Data Types";
           }
         }
       }
       output {
         choice compare-response {
           description
             "Comparison results.";
           leaf no-matches {
             type empty;
             description
               "This leaf indicates that the filter did not match
                anything and nothing was compared.";
           }
           container differences {
             description
               "The list of differences, encoded per RFC 8072 with an
                augmentation to include source values where applicable.
                When a datastore node in the source is not present in
                the target, this can be indicated either as a 'delete'
                or as a 'remove' as there is no difference between
                them for the purposes of the comparison.";
             uses ypatch:yang-patch {
               augment "yang-patch/edit" {
                 description
                   "Provides the value of the source of the patch,
                    respectively of the source of the comparison, in
                    addition to the target value, where applicable.";
                 anydata source-value {
                   when "../operation = 'delete'"
                      + "or ../operation = 'merge'"
                      + "or ../operation = 'move'"
                      + "or ../operation = 'replace'"
                      + "or ../operation = 'remove'";
                   description
                     "The anydata 'value' is only used for 'delete',
                      'move', 'merge', 'replace', and 'remove'
                      operations.";
                 }
                 reference
                   "RFC 8072: YANG Patch Media Type";
               }
             }
           }
         }
       }
     }
   }
   <CODE ENDS>

5.  Example

   The following example compares the difference between <operational>
   and <intended> for a subtree under "interfaces".  The subtree
   contains a subset of objects that are defined in a YANG data model
   for the management of interfaces defined in [RFC8343].  For the
   purposes of understanding the subsequent example, the following
   excerpt of the data model whose instantiation is the basis of the
   comparison is provided:

   container interfaces {
     description
       "Interface parameters.";
     list interface {
       key "name";
       leaf name {
         type string;
         description
           "The name of the interface.";
       }
       leaf description {
         type string;
         description
           "A textual description of the interface.";
       }
       leaf enabled {
         type boolean;
         default "true";
         description
           "This leaf contains the configured, desired state of the
            interface.";
       }
     }
   }

   The contents of <intended> and <operational> datastores in XML
   [W3C.REC-xml-20081126]:

   <!--INTENDED-->
   <interfaces xmlns="urn:ietf:params:xml:ns:yang:ietf-interfaces">
     <interface>
       <name>eth0</name>
       <enabled>false</enabled>
       <description>ip interface</description>
     </interface>
   </interfaces>

   <!--OPERATIONAL-->
   <interfaces
       xmlns="urn:ietf:params:xml:ns:yang:ietf-interfaces"
       xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin">
     <interface or:origin="or:learned">
       <name>eth0</name>
       <enabled>true</enabled>
     </interface>
   </interfaces>

   <operational> does not contain an instance for leaf "description"
   that is contained in <intended>.  Another leaf, "enabled", has
   different values in the two datastores, being "true" in <operational>
   and "false" in <intended>.  A third leaf, "name", is the same in both
   cases.  The origin of the leaf instances in <operational> is
   "learned", which may help explain the discrepancies.

   RPC request to compare <operational> (source of the comparison) with
   <intended> (target of the comparison):

   <rpc message-id="101"
       xmlns="urn:ietf:params:xml:ns:netconf:base:1.0">
     <compare xmlns="urn:ietf:params:xml:ns:yang:ietf-nmda-compare"
         xmlns:ds="urn:ietf:params:xml:ns:yang:ietf-datastores">
       <source>ds:operational</source>
       <target>ds:intended</target>
       <report-origin/>
       <xpath-filter
           xmlns:if="urn:ietf:params:xml:ns:yang:ietf-interfaces">
         /if:interfaces
       </xpath-filter>
     </compare>
   </rpc>

   RPC reply when a difference is detected:

   <rpc-reply
        xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"
        message-id="101">
     <differences
        xmlns="urn:ietf:params:xml:ns:yang:ietf-nmda-compare"
        xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin">
       <yang-patch>
         <patch-id>interface status</patch-id>
         <comment>
           diff between operational (source) and intended (target)
         </comment>
         <edit>
           <edit-id>1</edit-id>
           <operation>replace</operation>
           <target>/ietf-interfaces:interface=eth0/enabled</target>
           <value>
             <if:enabled>false</if:enabled>
           </value>
           <source-value>
              <if:enabled or:origin="or:learned">true</if:enabled>
           </source-value>
         </edit>
         <edit>
           <edit-id>2</edit-id>
           <operation>create</operation>
           <target>/ietf-interfaces:interface=eth0/description</target>
           <value>
             <if:description>ip interface</if:description>
           </value>
         </edit>
       </yang-patch>
     </differences>
   </rpc-reply>

   The same request in RESTCONF (using JSON format [RFC7951]):

   POST /restconf/operations/ietf-nmda-compare:compare HTTP/1.1
   Host: example.com
   Content-Type: application/yang-data+json
   Accept: application/yang-data+json

   { "ietf-nmda-compare:input" : {
      "source" : "ietf-datastores:operational",
      "target" : "ietf-datastores:intended",
      "report-origin" : null,
      "xpath-filter" : "/ietf-interfaces:interfaces"
      }
   }

   The same response in RESTCONF (using JSON format):

   HTTP/1.1 200 OK
   Date: Thu, 24 Jan 2019 20:56:30 GMT
   Server: example-server
   Content-Type: application/yang-data+json

   { "ietf-nmda-compare:output" : {
       "differences" : {
         "ietf-yang-patch:yang-patch" : {
           "patch-id" : "interface status",
           "comment" : "diff between intended (source) and operational",
           "edit" : [
             {
               "edit-id" : "1",
               "operation" : "replace",
               "target" : "/ietf-interfaces:interface=eth0/enabled",
               "value" : {
                  "ietf-interfaces:interface/enabled" : "false"
               },
               "source-value" : {
                  "ietf-interfaces:interface/enabled" : "true",
                  "@ietf-interfaces:interface/enabled" : {
                    "ietf-origin:origin" : "ietf-origin:learned"
                  }
                }
             },
             {
               "edit-id" : "2",
               "operation" : "create",
               "target" : "/ietf-interfaces:interface=eth0/description",
               "value" : {
                 "ietf-interface:interface/description" : "ip interface"
               }
             }
           ]
         }
       }
     }
   }

6.  Performance Considerations

   The <compare> operation can be computationally expensive.  While
   responsible client applications are expected to use the operation
   responsibly and sparingly only when warranted, implementations need
   to be aware of the fact that excessive invocation of this operation
   will burden system resources and need to ensure that system
   performance will not be adversely impacted.  One possibility for an
   implementation to mitigate against this is to limit the number of
   requests that are served to a client, or to any number of clients, in
   any one time interval, by rejecting requests made at a higher
   frequency than the implementation can reasonably sustain.

   While useful, tools such as YANG data models that allow for the
   monitoring of server resources, system performance, and statistics
   about RPCs and RPC rates are outside the scope of this document.
   When defined, any such model should be general in nature and not
   limited to the RPC operation defined in this document.

7.  IANA Considerations

7.1.  Update to the IETF XML Registry

   IANA has registered the following URI in the "IETF XML Registry"
   [RFC3688]:

   URI:  urn:ietf:params:xml:ns:yang:ietf-nmda-compare
   Registrant Contact:  The IESG.
   XML:  N/A; the requested URI is an XML namespace.

7.2.  Update to the YANG Module Names Registry

   IANA has registered the following YANG module in the "YANG Module
   Names" registry [RFC6020]:

   name:  ietf-nmda-compare
   namespace:  urn:ietf:params:xml:ns:yang:ietf-nmda-compare
   prefix:  cmp
   reference:  RFC 9144

8.  Security Considerations

   The YANG module specified in this document defines a schema for data
   that is designed to be accessed via network management protocols such
   as NETCONF [RFC6241] or RESTCONF [RFC8040].  The lowest NETCONF layer
   is the secure transport layer, and the mandatory-to-implement secure
   transport is Secure Shell (SSH) [RFC6242].  The lowest RESTCONF layer
   is HTTPS, and the mandatory-to-implement secure transport is TLS
   [RFC8446].

   The Network Configuration Access Control Model (NACM) [RFC8341]
   provides the means to restrict access for particular NETCONF or
   RESTCONF users to a preconfigured subset of all available NETCONF or
   RESTCONF protocol operations and content.

   NACM specifies access for the server in its entirety, and the same
   access rules apply to all datastores.  Any subtrees to which a
   requestor does not have read access are silently skipped and not
   included in the comparison.

   The RPC operation defined in this YANG module, <compare>, may be
   considered sensitive or vulnerable in some network environments.  It
   is thus important to control access to this operation.  This is the
   sensitivity/vulnerability of RPC operation <compare>:

   Comparing datastores for differences requires a certain amount of
   processing resources at the server.  An attacker could attempt to
   attack a server by making a high volume of comparison requests.
   Server implementations can guard against such scenarios in several
   ways.  For one, they can implement the NACM in order to require
   proper authorization for requests to be made.  Second, server
   implementations can limit the number of requests that they serve to a
   client in any one time interval, rejecting requests made at a higher
   frequency than the implementation can reasonably sustain.

9.  References

9.1.  Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119,
              DOI 10.17487/RFC2119, March 1997,
              <https://www.rfc-editor.org/info/rfc2119>.

   [RFC3688]  Mealling, M., "The IETF XML Registry", BCP 81, RFC 3688,
              DOI 10.17487/RFC3688, January 2004,
              <https://www.rfc-editor.org/info/rfc3688>.

   [RFC6020]  Bjorklund, M., Ed., "YANG - A Data Modeling Language for
              the Network Configuration Protocol (NETCONF)", RFC 6020,
              DOI 10.17487/RFC6020, October 2010,
              <https://www.rfc-editor.org/info/rfc6020>.

   [RFC6241]  Enns, R., Ed., Bjorklund, M., Ed., Schoenwaelder, J., Ed.,
              and A. Bierman, Ed., "Network Configuration Protocol
              (NETCONF)", RFC 6241, DOI 10.17487/RFC6241, June 2011,
              <https://www.rfc-editor.org/info/rfc6241>.

   [RFC6242]  Wasserman, M., "Using the NETCONF Protocol over Secure
              Shell (SSH)", RFC 6242, DOI 10.17487/RFC6242, June 2011,
              <https://www.rfc-editor.org/info/rfc6242>.

   [RFC6991]  Schoenwaelder, J., Ed., "Common YANG Data Types",
              RFC 6991, DOI 10.17487/RFC6991, July 2013,
              <https://www.rfc-editor.org/info/rfc6991>.

   [RFC7950]  Bjorklund, M., Ed., "The YANG 1.1 Data Modeling Language",
              RFC 7950, DOI 10.17487/RFC7950, August 2016,
              <https://www.rfc-editor.org/info/rfc7950>.

   [RFC7951]  Lhotka, L., "JSON Encoding of Data Modeled with YANG",
              RFC 7951, DOI 10.17487/RFC7951, August 2016,
              <https://www.rfc-editor.org/info/rfc7951>.

   [RFC8040]  Bierman, A., Bjorklund, M., and K. Watsen, "RESTCONF
              Protocol", RFC 8040, DOI 10.17487/RFC8040, January 2017,
              <https://www.rfc-editor.org/info/rfc8040>.

   [RFC8072]  Bierman, A., Bjorklund, M., and K. Watsen, "YANG Patch
              Media Type", RFC 8072, DOI 10.17487/RFC8072, February
              2017, <https://www.rfc-editor.org/info/rfc8072>.

   [RFC8174]  Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
              2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
              May 2017, <https://www.rfc-editor.org/info/rfc8174>.

   [RFC8340]  Bjorklund, M. and L. Berger, Ed., "YANG Tree Diagrams",
              BCP 215, RFC 8340, DOI 10.17487/RFC8340, March 2018,
              <https://www.rfc-editor.org/info/rfc8340>.

   [RFC8341]  Bierman, A. and M. Bjorklund, "Network Configuration
              Access Control Model", STD 91, RFC 8341,
              DOI 10.17487/RFC8341, March 2018,
              <https://www.rfc-editor.org/info/rfc8341>.

   [RFC8342]  Bjorklund, M., Schoenwaelder, J., Shafer, P., Watsen, K.,
              and R. Wilton, "Network Management Datastore Architecture
              (NMDA)", RFC 8342, DOI 10.17487/RFC8342, March 2018,
              <https://www.rfc-editor.org/info/rfc8342>.

   [RFC8446]  Rescorla, E., "The Transport Layer Security (TLS) Protocol
              Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018,
              <https://www.rfc-editor.org/info/rfc8446>.

   [W3C.REC-xml-20081126]
              Bray, T., Paoli, J., Sperberg-McQueen, M., Maler, E., and
              F. Yergeau, "Extensible Markup Language (XML) 1.0 (Fifth
              Edition)", World Wide Web Consortium Recommendation REC-
              xml-20081126, November 2008,
              <https://www.w3.org/TR/2008/REC-xml-20081126>.

9.2.  Informative References

   [RFC8343]  Bjorklund, M., "A YANG Data Model for Interface
              Management", RFC 8343, DOI 10.17487/RFC8343, March 2018,
              <https://www.rfc-editor.org/info/rfc8343>.

Appendix A.  Possible Future Extensions

   It is conceivable to extend the <compare> operation with a number of
   possible additional features in the future.

   Specifically, it is possible to define an extension with an optional
   feature for dampening.  This will allow clients to specify a minimum
   time period for which a difference must persist for it to be
   reported.  This will enable clients to distinguish between
   differences that are only fleeting from ones that are not and that
   may represent a real operational issue and inconsistency within the
   device.

   For this purpose, an additional input parameter can be added to
   specify the dampening period.  Only differences that pertain for at
   least the dampening time are reported.  A value of 0 or omission of
   the parameter indicates no dampening.  Reporting of differences MAY
   correspondingly be delayed by the dampening period from the time the
   request is received.

   To implement this feature, a server implementation might run a
   comparison when the RPC is first invoked and temporarily store the
   result.  Subsequently, it could wait until after the end of the
   dampening period to check whether the same differences are still
   observed.  The differences that still persist are then returned.

Acknowledgments

   We thank Rob Wilton, Martin Bjorklund, Mahesh Jethanandani, Lou
   Berger, Kent Watsen, Phil Shafer, Ladislav Lhotka, Tim Carey, and
   Reshad Rahman for their valuable feedback and suggestions.

Authors' Addresses

   Alexander Clemm
   Futurewei
   2330 Central Expressway
   Santa Clara, CA 95050
   United States of America

   Email: ludwig@clemm.org


   Yingzhen Qu
   Futurewei
   2330 Central Expressway
   Santa Clara, CA 95050
   United States of America

   Email: yqu@futurewei.com


   Jeff Tantsura
   Microsoft

   Email: jefftant.ietf@gmail.com


   Andy Bierman
   YumaWorks

   Email: andy@yumaworks.com

AI ELI5

On this page
PROBLEMWith NMDA, the same data can sit in several datastores, and what a user intended can differ from what is really in effect. Fetching both datastores and comparing them is costly and wasteful.
MECHANISMA new YANG RPC, compare, takes a source and target datastore plus optional filter, all and report-origin parameters. The server returns the differences as a YANG Patch.
WHY IT MATTERSOperators can see where applied configuration departs from intended configuration without moving whole datastores. The operation can be expensive, so access control and rate limiting are advised.

§1Why Compare Datastores

The Network Management Datastore Architecture (NMDA, RFC 8342) introduces datastores that each give a different viewpoint on YANG-defined data. <intended> holds validated configuration a client wants in effect. <operational> holds operational state, such as statistics, plus configuration that is actually in effect.

Data moves from one datastore to another, so differences can appear. Some are expected, because there can be a time lag between configuration reaching <intended> and taking effect. Others come from unmet conditions, inactive parts of the configuration, unresolved resource dependencies or implementation errors. They can hide the true status of the network, harm services and misallocate resources.

A client could fetch both datastores and compare them itself, but the RFC calls that costly and extremely wasteful. RFC 9144 instead lets the server do it, with RPCs used over NETCONF (RFC 6241) or RESTCONF (RFC 8040).

Imagine writing a to-do list and a helper writing what they actually did. This lets you ask for just the differences instead of reading both lists.

§3The compare Operation

The core is one new RPC, <compare>. It checks for differences in values or in data nodes present in either datastore. The input parameters are:

  • source: the reference datastore, for example <intended>.
  • target: the datastore compared against it, for example <operational>.
  • filter-spec: a subtree filter or an XPath filter that limits the comparison to part of the datastore. It cannot match on data node values.
  • all: include all differences, including schema nodes present in only one datastore. Without it, the server first drops schema nodes not present in both.
  • report-origin: include origin metadata. It is omitted by default and applies only when <operational> is involved.

The prefiltering when all is absent lets clients focus on real mismatches of instance data without writing more complex filters.

You tell the device which two lists to compare, and optionally which part of them. By default it ignores items that could only ever exist in one list.

§4Output and the YANG Module

STRUCTURE OF IETF-NMDA-COMPARE (SIMPLIFIED FROM FIGURE 1)
  • compare (rpc)
    • input
      • source
      • target
      • all?
      • report-origin?
      • filter-spec? subtree-filter or xpath-filter
    • output
      • no-matches?
      • differences yang-patch with edits

The output is either no-matches, meaning the filter matched nothing so nothing was compared, or differences. Differences use the YANG Patch format of RFC 8072, so the result is the patch that would turn the source into the target. The data model augments each edit with a source-value so the reader also sees the source's value.

A node in the source that is missing from the target may be reported as either delete or remove, since the comparison does not distinguish them. When the target is <operational> and report-origin is set, origin metadata is included. This can explain a difference, for example when a node in <intended> has origin system in <operational>.

The module is named ietf-nmda-compare, with prefix cmp, and its tree notation follows RFC 8340.

The answer is a list of edits, like tracked changes in a document, that would make the first list match the second.

§5A Worked Example

The RFC compares an interface in <operational> (the source) with <intended> (the target). In <intended>, interface eth0 has enabled set to false and a description. In <operational> it is true with no description, and its origin is learned.

The reply holds two edits. Edit 1 is a replace of enabled with the target value false, plus a source-value of true with its origin. Edit 2 is a create of description with the value ip interface. The RFC shows the exchange in NETCONF XML and in RESTCONF JSON.

One list says the lamp is off and has a label. The other says the lamp is on and has no label. The answer says: turn it off and add the label.

RESTCONF REQUEST FROM RFC 9144
Compare request (JSON)
POST /restconf/operations/ietf-nmda-compare:compare HTTP/1.1
Host: example.com
Content-Type: application/yang-data+json
Accept: application/yang-data+json

{ "ietf-nmda-compare:input" : {
   "source" : "ietf-datastores:operational",
   "target" : "ietf-datastores:intended",
   "report-origin" : null,
   "xpath-filter" : "/ietf-interfaces:interfaces"
   }
}

§6Performance Costs

The operation can be computationally expensive. Clients are expected to use it sparingly, and implementations must make sure that heavy use does not hurt system performance. One suggested mitigation is to limit how many requests are served to a client, or to all clients, in a time interval, rejecting requests that come too often.

Models for monitoring server resources and RPC rates are outside the scope of this document. Any such model should be general and not limited to this RPC.

Comparing big lists takes effort. If too many people ask at once, the device should say no to some so it does not get tired out.

§8Access Control and Abuse

The module is accessed through NETCONF or RESTCONF. Their mandatory-to-implement secure transports are SSH (RFC 6242) for NETCONF and TLS (RFC 8446) for RESTCONF. The Network Configuration Access Control Model (NACM, RFC 8341) can restrict users to a subset of operations and content. NACM rules are per server, so they apply equally to all datastores, and subtrees the requestor cannot read are silently skipped in the comparison.

The <compare> RPC may be sensitive. An attacker could flood a server with comparison requests. Servers can guard against this by requiring authorization through NACM and by limiting the number of requests served to a client in a time interval.

Only people with permission should be allowed to ask, and anything they are not allowed to see is left out of the answer.

§7Possible Future Extensions

Appendix A sketches one extension, dampening. A client would give a minimum period for which a difference must persist before it is reported. That separates fleeting differences from ones that may show a real inconsistency inside the device.

An extra input parameter would carry the dampening period. A value of 0, or omitting it, means no dampening, and reports MAY be delayed by the period. A server could run a comparison when the RPC arrives, store the result, and after the period check which differences remain.

Dampening means waiting a moment before reporting a difference, to be sure it is not just a blink that goes away on its own.

§8References