One document matched: draft-bogdanovic-netmod-yang-model-classification-03.xml
<?xml version="1.0" encoding="UTF-8"?>
<!-- This is built from a template for a generic Internet Draft. Suggestions for
improvement welcome - write to Brian Carpenter, brian.e.carpenter @ gmail.com -->
<!-- This can be converted using the Web service at http://xml.resource.org/experimental.html
(which supports the latest, sometimes undocumented and under-tested, features.) -->
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
]>
<?rfc toc="yes"?> <!-- You want a table of contents -->
<?rfc symrefs="yes"?> <!-- Use symbolic labels for references -->
<?rfc sortrefs="yes"?> <!-- This sorts the references -->
<?rfc iprnotified="no" ?> <!-- Change to "yes" if someone has disclosed IPR for the draft -->
<?rfc compact="yes"?>
<!-- This defines the specific filename and version number of your draft (and inserts the appropriate IETF boilerplate -->
<rfc ipr="trust200902" docName="draft-bogdanovic-netmod-yang-model-classification-03" category="info" >
<front>
<title abbrev="YANG model classification">YANG model classification</title>
<author fullname="Dean Bogdanovic" initials="D." surname="Bogdanovic">
<organization>Juniper Networks</organization>
<address>
<email>deanb@juniper.net</email>
</address>
</author>
<author fullname="Benoit Claise" initials="B." surname="Claise">
<organization>Cisco Systems, Inc.</organization>
<address>
<email>bclaise@cisco.com</email>
</address>
</author>
<author fullname="Carl Moberg" initials="C." surname="Moberg">
<organization>Cisco Systems, Inc.</organization>
<address>
<email>camoberg@cisco.com</email>
</address>
</author>
<date day="3" month="June" year="2015" />
<area>Operations and Management</area>
<workgroup>NETMOD</workgroup>
<abstract>
<t>The YANG <xref target="RFC6020"/> data modeling language is currently
being considered for a wide variety of applications throughout the
networking industry at large. Many standards organizations and open
source projects are using YANG to develop and publish models of
configuration, state data and operations for a wide variety of
networking applications. At the same time, there is currently no
well-known terminology to categorize the various types of YANG models
that are being worked on.</t>
<t>A consistent terminology would help with the categorization of models,
assist in the analysis the YANG data modeling efforts in the IETF and
in other places, and bring clarity to the YANG-related discussions
between the different groups.</t>
<t>This document describes a set of concepts and associated terms to
support consistent classification of YANG models.</t>
</abstract>
</front>
<middle>
<section anchor="introduction" title="Introduction">
<t>The Internet Engineering Steering Group (IESG) has been actively
encouraging IETF working groups to use the NETCONF
<xref target="RFC6241"/> and YANG standards for configuration
management purposes, especially in new charters
<xref target="Writable-MIB-Module-IESG-Statement"/>.</t>
<t>YANG is also gaining wide acceptance as the de-facto standard modeling
language in the broader industry. This extends beyond the IETF,
including many standard development organizations, industry consortia,
ad hoc groups, and open source projects.</t>
<t>There are no clear guidelines on how to classify the layering of YANG
models according to abstraction, or how to classify models along the
continuum spanning industry standard and vendor-specific models.</t>
<t>This document presents a set of concepts and terms to form a useful
taxonomy for consistent classification of YANG models in two dimensions:
<list style="symbols">
<t>Based on the level of abstraction in the model</t>
<t>Based on the applicability of the model</t>
</list>
The two categories are covered in the next two sections.
</t>
</section> <!-- intro -->
<section anchor="firstdimension" title="First Dimension: YANG Data Model Layering">
<t>Model developers have taken two approaches to development: top-down
and bottom-up. The top-down approach starts with high level
abstractions modeling business or customer requirements and maps them to
specific networking technologies. The bottom-up approach starts with
fundamental networking technologies and maps them into more abstract
constructs.</t>
<t>There are currently no specific requirements on, or well-defined best
practices around the develoment of models. For the purpose of this
document we assume that both approaches (bottom-up and top-down) will
be used as they both provide benefits that appeals to different groups.
</t>
<t>For layering purposes, this documents suggests the classification of
data models into two distinct abstraction layers:</t>
<t>
<list style="symbols">
<t>Network Element YANG Models describe the configuration, state
data and operations of a specific device-centric technology or
feature.</t>
<t>Network Service YANG Models describes the configuration, state
data and operations of an abstract representation of a service
implemented on one or multiple network elements</t>
</list>
</t>
<t>Figure 1 illustrates the application of YANG models at different
layers of abstraction. Layering of models allow for reusability of
existing lower layer models in higher level models while limiting
duplication of features across layers.</t>
<t>For model developers, per-layer modeling allows for separation of
concern across editing teams focusing on specific areas.</t>
<t>As an example, experience from the IETF shows that creating useful
network element YANG models for e.g routing or switching protocols
requires teams that include developers with experience from
implementing those protocols.</t>
<t>On the other hand, network service models are best developed by people
experienced in defining network services for consumption by programmers
developing e.g. flow-through provisioning systems or self-service
portals.</t>
<figure align="center">
<artwork align="center"><![CDATA[
+-----------------------+
| |
| OSS/BSS |
| |
+-----------------------+
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Network Service YANG data models
+------------+ +-------------+ +-------------+
| | | | | |
| - VPWS | | - VPLS | | L3VPN |
| - L2VPN | | - L2VPN | | |
| | | | | |
+------------+ +-------------+ +-------------+
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Network Element YANG data models
+------------+ +-----------+ +-------------+ +-----------+
| | | | | | | |
| MPLS | | BGP | | Interface | | Routing |
| | | | | | | |
+------------+ +-----------+ +-------------+ +-----------+
Fig. 1 YANG Model Layers
]]></artwork>
</figure> <!-- YANG model layers -->
<section anchor="networkservice" title="Network Service YANG Data Models">
<t>Network Service YANG Data Models are created to describe the
characteristics of a service, as agreed upon with consumers of that
service. That is, a service model does not expose the detailed
configuration parameters of all participating network elements, but
describes an abstract model that allows instances of the service to be
decomposed into instance data according to the Network Element data
models of the participating network elements. The service-to-element
decomposition is a separate process with details depending on how the
network operator chooses to realize the service.</t>
<t>As an example, the Network Service model included in
http://datatracker.ietf.org/doc/draft-l3vpn-service-yang/ provides an
abstracted view of a Layer 3 IP VPN service configuration components.
An orchestrator receives operations on instances according to the
service model as input and decomposes the data into specific Network
Element models to configure the participating network elements to
perform the service.</t>
<t>Network Service YANG models defines complete service to be consumed by
external systems. These models are commonly designed, developed and
deployed by network infrastructure teams.</t>
<t>YANG allows for different design patterns to describe network services,
ranging from monolithic to component-based approaches.</t>
<t>The monolithic approach captures the entire service in a single model
and does not put focus on reusability of internal data definitions and
groupings. The monolithic approach has the advantages of single-purpose
development including speed at the expense of reusability.</t>
<t>The component-based approach captures device-centric features (e.g.
the definition of a VRF, routing protocols, or packet filtering) in a
vendor-independent manner. The components are designed for reuse across
many services. The set of components required for a specific service is
then composed into the higher-level service. The component-based
approach has the advantages of modular development including a higher
degree of reusability at the expense of initial speed.</t>
<t>As an example, an L2VPN service can be built on many different types of
transport network technologies, including e.g. MPLS or carrier ethernet.
A component-based approach would allow for reuse of e.g. UNI-interface
definitions independent of the underlying transport network (e.g. MEF
UNI interface or MPLS interface). The monolithic approach would assume
a specific set of transport technologies and interface definitions.</t>
</section> <!-- networkservice -->
<section anchor="networkElement" title="Network Element YANG Data models">
<t>Network Element YANG Data Models describe the configuration, state
data and operations of a network device as defined by the vendor of
that device. The models are commonly structured around features of the
device, e.g. interface configuration <xref target='RFC7223'/>, OSPF
configuration <xref target='I-D.ietf-ospf-yang'/>, and firewall rules
definitions <xref target='I-D.ietf-netmod-acl-model'/>. The model
provides a coherent representation of what is commonly a very
mixed software environment that consists of the operating system and
applications running on the device.</t>
<t>The decomposition, ordering and execution of changes to the operating
system, and application configuration is the task of the management
framework that implements the YANG model.</t>
</section> <!-- protocol/feature YANG data models -->
</section> <!-- firstdimension -->
<section title="Second Dimension: Model Type">
<t>At very high level, models can be divided into proprietary and standard.
Each vendor, consortium, open source project can publish their models and
those are considered proprietary models. When an SDO, such as IETF or IEEE,
publishes an accepted model document, then this is a standard model. There
are use cases where a consortium has published work which de facto became
standard, such as Linux kernel, but for the clarity in this document,
authors are making a separation between models based on the above
description.
</t>
<t><list style="symbol">
<t>Standard YANG Model: YANG model defined by an Standard Development
organization (SDO), e.g. IETF, IEEE. </t>
<t>Standard Extension YANG Model: YANG Model that describes a standard
extension, example route filter, to standard filter YANG model.</t>
<t>Proprietary Extension to Standard YANG Model: As the Standard YANG Models
contains a subset of all the Vendor Configuration Models, proprietary
extensions must complement the Standard YANG Models to represent a Vendor
Configuration Model.</t>
<t>Proprietary YANG Model: A non Standard YANG Model.</t>
<t>Vendor Configuration Model: It describes all configurable capabilities of
the device and what device vendor exposes for configuration. The vendor
configuration model can be CLI or YANG-based.</t>
</list></t>
<t>As mentioned earlier in this document, there are two ways of designing
models, top down and bottom up with one restriction. Everything is dependent
on the vendor data model. That model describes all the possibilities and if
model developers prefers, they can use vendor model only to design service
components, network service and business service. Using vendor model provides
all capabilities today, but it comes with restrictions of portability between
vendors and to certain extent devices.
On the other hand, only standard models and standard extensions can be used,
but this might result in less feature rich or less efficient services.
Service model developer has a choice to reuse service components or write
a model completely based on vendor data model.
</t>
<section anchor="standard" title="Standard YANG model">
<t>With YANG we have a common language, that enables different
communities to express data models that are widely understandable
without lot of additional explanation. This enables different groups,
such as IETF, to standardize data models, defined as an IETF RFC,
and vendors to support them, which will make it easier to for network
operators to manage their network configuration programmatically.
The Standard YANG Models can distinguished between the core YANG models,
such as the YANG Data Model for Interface Management <xref target="RFC7223"/>,
and the technology specific YANG models, such as the Configuration Data
Model for the IP Flow Information Export (IPFIX) and Packet Sampling (PSAMP)
Protocols <xref target="RFC6728"/>.
</t>
</section> <!-- standard -->
<section anchor="standardextension" title="Standard Extension YANG Model">
<t>Standard Extension is the conditional portion of a Standard YANG
Model, expressed with the feature, if-feature, augment YANG
statements <xref target="RFC6020"/>. An example of such standard
extension is policy based routing (PBR). PBR is found in many vendor
implementations and have many common features, but not all vendors
support PBR on all of their devices.</t>
</section> <!-- standardextension -->
<section anchor="proprietaryextensiontostandard" title="Proprietary Extension to Standard YANG Model">
<t>Proprietary extension is a conditional portion of a Standard YANG
Model, expressed with feature, if-feature, augment YANG statements
<xref target="RFC6020"/>. Proprietary extensions are required as the
Standard YANG model will not cover all the possible configuration
parameters of the different vendors. Proprietary extension can be a feature
depending on hardware platform capabilities and it is not available by
other vendors. Such an example could be match condition for packet
classification used for PBR.
</t>
</section> <!-- proprietaryextensiontostandard -->
<section anchor="vendor" title="Vendor configuration model">
<t>Base model for all other models is the vendor configuration model.
It describes all configurable capabilities of the device and what
device vendor exposes for configuration.</t>
<t>The standard configuration model is a subset of vendor configuration
model. The standard configuration model can be broken into base model
and standard extension models, where the base is common data model and
standard extensions are standard features that are not implemented
by all vendors. Example of standard base model is Access
Control List and routing filter is a standard extension on ACL. Or
another example: encryption algorithm is standard feature, but the
different types, like md5, hmac-md5, hmac-sha1, etc are standard
extensions, as it is not that all vendors have all encryption
algorithm types implemented.</t>
<t>Although all vendors provide very similar functionality using
standards, implementations are different. One of basic examples are
dynamic routing protocols. We can see today two main types of routing
protocol configuration.
</t>
<t><list style="empty">
<t>protocol centric - all the protocol related config is contained
with the protocol itself. Especially in case of multiple instances
of the routing protocol running in different routing-instances
(routing-instance as described in core routing model <xref target="I-D.ietf-netmod-routing-cfg"/>),
all the routing-instance protocol config is contained in the
default routing instance.
</t>
</list>
</t>
<figure>
<artwork>
Router ospf 10
Default-metric 100
Address-family ipv4 vrf VRF1
Network x.x.x.x area 0
Address-family ipv4 vrf VRF2
Network x.x.x.x area 0
Address-family ipv4
Network x.x.x.x area 1
</artwork>
</figure>
<t><list style="empty">
<t>In term of YANG model, the routing protocol configuration will be
defined within the default routing-instance and the routing-protocol
config will contain multiple instances referring to other
routing-instances.
</t>
<t>VRF centric - All the protocol related config for a routing-instance
is contained within this routing-instance.
</t>
</list></t>
<figure>
<artwork>
Routing-instance VRF1 {
Protocols isis {
}
}
Routing-instance VRF2 {
Protocols isis {
}
}
</artwork>
</figure>
<t><list style="empty">
<t>In terms of YANG model, the routing protocol configuration for a
routing-instance will be defined within the associated
routing-instance.
</t>
<t>The bottom line message is that, even if YANG models are standardized,
they will provide different CLI outcomes, simply because the CLI among
vendors is not standardized.
</t>
</list>
</t>
</section> <!--vendor -->
<section anchor="proprietary" title="Proprietary YANG Model">
<t>While waiting for the Standard YANG Models to be published,
the different vendors might offer Proprietary YANG Models.
</t>
</section> <!-- proprietary -->
</section> <!-- Second Dimension -->
<section title="Typical Architecture">
<figure align="center">
<artwork align="center"> <![CDATA[
+--------------------------------------------------------+
| OSS/BSS |
+--------------------------------------------------------+
+--------------------------------------------------------+
| Orchestrator |
| +------------------------------------------------+ |
| | Network Service Model | |
| | | |
| +------------------------------------------------+ |
+--------------------------------------------------------+
+--------------------------------------------------------+
| Network Element |
| | |
| +-----------------------+ | +-------------------+ |
| | Standard YANG model | | | Proprietary | |
| | | | | YANG Model | |
| +-----------------------+ | | | |
| | | | |
| +-----------------------+ | | | |
| | Proprietary Extension | | | | |
| | To YANG Standard | | | | |
| | Model | | | | |
| +-----------------------+ | +-------------------+ |
| | |
| +-------------------------------------------------+ |
| | Vendor Configuration Model | |
| +-------------------------------------------------+ |
+--------------------------------------------------------+
Fig. 2 Typical Architecture
]]></artwork>
</figure>
<t>The OSS/BSS may contains business related models. Those models, which
may or may not be written in YANG, are outside the scope of the IETF work</t>
</section> <!-- Typical Architecture -->
<section title="IETF, Other SDOs, and open source">
<t>IETF, as a standard defining organization (SDO), is well positioned to
standardize Network Element YANG models. With a wide range of expertise
found within its working groups focused on those technology definitions. As IETF
participants implement those protocols, they have deep expertise about
the implementation and finding a common base standard configuration model
between vendors should be a very viable goal.
</t>
<t>In some situation where the protocols are standardized by different SDOs,
those SDOs should be responsible for its YANG data modeling effort.
For example, the IETF has transferred the responsibility for some IEEE
technology-related MIB modules to the IEEE 802.1 and 802.3 Working Group
<xref target="RFC4663"/>, <xref target="RFC7448"/>. Similarly, the IEEE
should be responsible for similar YANG data modeling efforts.
</t>
<t>Developing Network Service YANG Models requires network operations expertise.
When those operators participate in IETF work, the right working group can be
formed, and those Service YANG Models can be developed within IETF.
However, some other groups, like Metro Ethernet Forum or CableLabs, could
be better positioned for service modeling related to their area of expertise.</t>
<t>Today there are many open source projects and some of them are becoming
de facto standards, like the Linux kernel. Many such open source projects,
like Open Daylight, OpenStack, etc, are doing very good work and their
work is being accepted and deployed in production environments. They bring
a lot of very valuable experience to other groups. From IETF perspective,
if there is such a work present, it can be used as a very good starting
point for modeling within IETF.
</t>
</section> <!-- IETF, Other SDOs, open source -->
<section anchor="security" title="Security Considerations">
<t>At this stage, authors of the draft didn't look into security considerations.</t>
</section> <!-- security -->
<section anchor="iana" title="IANA Considerations">
<t>This document requests no action by IANA. </t>
</section> <!-- iana -->
<section anchor="ack" title="Acknowledgements">
<t>Thanks to David Ball for his enlightenments on Metro Ethernet Forum service aspects.</t>
</section> <!-- ack -->
<section anchor ="changes" title="Change log [RFC Editor: Please remove]">
<t>version 1: restructure the document, add the two dimensions, add the interaction with the different
SDOs and opensource projects, add the definitions.</t>
<t>version 2: added definitions for config and service models
clarified second dimension of model classification.
fixed typos
</t>
</section> <!-- changes -->
</middle>
<back>
<references title="Normative References">
<?rfc include='reference.RFC.6020'?>
</references>
<references title="Informative References">
<?rfc include='reference.RFC.4663'?>
<?rfc include='reference.RFC.6241'?>
<?rfc include='reference.RFC.6728'?>
<?rfc include='reference.RFC.7223'?>
<?rfc include='reference.RFC.7448'?>
<?rfc include='reference.I-D.ietf-netmod-routing-cfg'?>
<?rfc include='reference.I-D.ietf-netmod-acl-model'?>
<?rfc include='reference.I-D.ietf-ospf-yang'?>
<reference anchor="Writable-MIB-Module-IESG-Statement" target="https://www.ietf.org/iesg/statement/writable-mib-module.html">
<front>
<title>Writable MIB Module IESG Statement</title>
<author/>
<date/>
</front>
</reference>
</references>
</back>
</rfc>
| PAFTECH AB 2003-2026 | 2026-04-23 10:50:29 |