System-defined Configuration

System-defined Configuration Huawei

101 Software Avenue, Yuhua District Nanjing Jiangsu 210012 China maqiufang1@huawei.com

Huawei

101 Software Avenue, Yuhua District Nanjing Jiangsu 210012 China bill.wu@huawei.com

fengchongllly@gmail.com

ops NETMOD system config The Network Management Datastore Architecture (NMDA) in RFC 8342 defines several configuration datastores holding configuration. The contents of these configuration datastores are controlled by clients. This document introduces the concept of system configuration datastore holding configuration controlled by the system on which a server is running. The system configuration can be referenced (e.g., leafref) by configuration explicitly created by clients. This document updates RFC 8342.

The Network Management Datastore Architecture (NMDA) defines system configuration as the configuration that is supplied by the device itself and appears in <operational> when it is in use (Figure 2 in ). However, there is a desire to enable a server to better expose the system configuration, regardless of whether it is in use. For example, some implementations defines the system configuration which must be referenced to be active. NETCONF/RESTCONF clients can benefit from a standard mechanism to retrieve what system configuration is available on a server. Some servers allow the descendant nodes of system-defined configuration to be configured or modified. For example, the system configuration may contain an almost empty physical interface, whose existence in the system configuration is tied to the presence of particular hardware, while the client needs to be able to add, modify, or remove a number of descendant nodes. Some descendant nodes may not be modifiable (e.g., the interface "type" set by the system). This document updates the NMDA defined in with a read-only conventional configuration datastore called "system" to expose system-defined configuration. The solution enables configuration explicitly created by the clients to reference nodes defined in <system>, override system-provided values, and configure descendant nodes of system-defined configuration. The solution defined in this document requires the use of NMDA for both clients and servers. Conformance to this document requires NMDA servers implement the "ietf-system-datastore" YANG module ().

This document assumes that the reader is familiar with the contents of , , , , and and uses terminologies from those documents. The following terms are defined in this document: RFC 8342 defines it as "Configuration that is supplied by the device itself." The definition herein refines that definition of system configuration to represent configuration present in the system configuration datastore (regardless of whether it is applied or referenced). It may also be referred to as "system-defined configuration" or "system-provided configuration" throughout this document. A configuration datastore holding configuration provided by the system itself. This datastore is referred to as "<system>". This document redefines the term "conventional configuration datastore" in to add "system" to the list of conventional configuration datastores: One of the following set of configuration datastores: <running>, <startup>, <candidate>, <system>, and <intended>. These datastores share a common datastore schema, and protocol operations allow copying data between these datastores. The term "conventional" is chosen as a generic umbrella term for these datastores. An instance in the data tree that is provided by the system itself. System node may also be called "system-defined node" or "system-provided node" throughout this document. A referenced node is one of: Targets of leafref values defined via the "path" statement. Targets of "instance-identifier" type values. Nodes present in an XPath expression of "when" constraints. Nodes present in an XPath expression of "must" constraints. Nodes defined to satisfy the "mandatory true" constraints. Nodes defined to satisfy the "min-elements" constraints.

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 when, and only when, they appear in all capitals, as shown here.

This document updates RFC 8342 to define a configuration datastore called "system" to hold system configuration (), it also redefines the term "conventional configuration datastore" from to add "system" to the list of conventional configuration datastores. Configuration in <running> is merged with <system> to create the contents of <intended> after the configuration transformations (e.g., template expansion, removal of inactive configuration defined in ) have been performed, as described in . This document also updates the definition of "intended" origin metadata annotation identity defined in . The "intended" identity of origin value defined in represents the origin of configuration provided by <intended>, this document updates its definition as the origin source of configuration explicitly provided by clients, and allows a subset of configuration in <intended> that flows from <system> yet is not configured or overridden explicitly in <running> to use "system" as its origin value. As per , all configuration with origin value being reported as "intended" MUST originate from <running>, which includes any configuration in <system> that has been copied into <running>. Configuration that is in <system> and not also present in <running> MUST be reported as origin "system" in <operational>.

This document defines two types of system configuration. Configuration that is always-present and configuration that is conditionally-present. These types of system configuration are described in and , respectively.

Always-present refers to system configuration which is generated in <system> when the device is powered on, irrespective of physical resource present or not, a special functionality enabled or not. An example of always-present system configuration is an always-existing loopback interface.

Conditionally-present refers to system configuration which is generated in <system> based on specific conditions being met in a system. For example, if a physical resource is present (e.g., an interface card is inserted), the system automatically detects it and loads the associated configuration; when the physical resource is not present (an interface card is removed), the system configuration will automatically be removed from <system>. Another example is when a special functionality is enabled, e.g., when a license or feature is enabled, specific configuration may be created by the system.

Following guidelines for defining datastores in the , this document introduces a new datastore resource named "system" that represents the system configuration. NMDA servers compliant with this document MUST implement a system configuration datastore, and they SHOULD also implement <intended>. Name: "system". YANG modules: all. YANG nodes: all "config true" data nodes up to the root of the tree, generated by the system. Management operations: The datastore can be read using network management protocols such as NETCONF and RESTCONF, but its contents cannot be changed by manage operations via NETCONF and RESTCONF protocols. Origin: This document does not define any new origin identity. The "system" identity of origin metadata annotation is used to indicate the origin of a data item provided by the system. Protocols: YANG-driven management protocols, such as NETCONF and RESTCONF. Defining YANG module: "ietf-system-datastore" (). The system configuration datastore doesn't persist across reboots.

Clients may provide configuration nodes that reference nodes defined in <system>, override system-provided values, and configure descendant nodes of system-defined configuration in <running>, as detailed in . To ensure the validity of <intended>, configuration in <running> is merged with <system> to become <intended>, in which process, configuration appearing in <running> takes precedence over the same node in <system>. Since it is unspecified how to merge configuration before transformations, if <system> or <running> includes configuration that requires further transformation (e.g., template expansion, removal of inactive configuration defined in ) before it can be applied, configuration transformations MUST be performed before <running> is merged with <system>. Whenever configuration in <system> changes, the server MUST also immediately update and validate <intended>. As a result, Figure 2 in is updated with the below conceptual model of datastores which incorporates the system configuration datastore.

The system datastore is read-only (i.e., edits towards <system> directly MUST be denied), though the client may be allowed to provide configuration that overrides the value of a system-initialized node (see ).

This work does not change the definition of <operational>; it clarifies origin reporting, i.e., the origin of nodes sourced from <system> is reported as "system" unless explicitly configured or overridden in <running>. <system> enables system-generated nodes to be defined like configuration, i.e., made visible to clients in order for being referenced or configurable prior to present in <operational>. "config false" nodes are out of scope, hence existing "config false" nodes are not impacted by this work.

The contents of <system> MAY change dynamically under various conditions, such as license change, software upgrade, and system-controlled resources change (see ). The updates of system configuration may be obtained through YANG notifications (e.g., on-change notification) . If system configuration changes, <running> SHOULD remain a valid configuration data tree. Any mechanisms to achieve this are outside the scope of this document.

Clients may create configuration data in <running> that references nodes in <system>. Some implementations may define system nodes solely as a convenience for clients to reference. It is also possible for the clients to define their customized nodes for reference. provides an example of a client referencing system-defined nodes.

In some cases, a server may allow some parts of system configuration (e.g., a leaf's value) to be modified. Modification of system configuration is achieved by the client writing configuration data in <running> that overrides the values of matched configuration nodes at the corresponding level in <system>. Configurations defined in <running> take precedence over system configuration nodes in <system> if the server allows the nodes to be modified (some implementations may have immutable system configuration as per ). provides an example of a client overriding a system-instantiated leaf's value.

A server may also allow a client to add nodes to a list entry in <system> by writing those additional nodes in <running>. Those additional data nodes may not exist in <system> (i.e., an addition rather than an override). provides an example of a client configuring descendant nodes of a system-defined node.

This YANG module defines a new YANG identity named "system" that uses the "ds:conventional" identity defined in as its base. A client can discover the system configuration datastore support on the server by reading the YANG library information from the operational state datastore. The system datastore is defined as a conventional configuration datastore and shares a common datastore schema with other conventional datastores. The following diagram illustrates the relationship amongst the "identity" statements defined in the "ietf-system-datastore" and "ietf-datastores" YANG modules:

<CODE BEGINS> file "ietf-system-datastore@2025-11-28.yang" WG List: Author: Qiufang Ma Author: Qin Wu Author: Chong Feng "; description "This module defines a new YANG identity that uses the ds:conventional identity defined in [RFC8342]. Copyright (c) 2025 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 XXXX (https://www.rfc-editor.org/info/rfcXXXX); see the RFC itself for full legal notices."; revision 2025-11-28 { description "Initial version."; reference "RFC XXXX: System-defined Configuration"; } identity system { base ds:conventional; description "This read-only datastore contains the configuration provided by the system itself."; } } ]]> <CODE ENDS>

This document registers one XML namespace URI in the 'IETF XML registry', following the format defined in .

URI: urn:ietf:params:xml:ns:yang:ietf-system-datastore Registrant Contact: The IESG. XML: N/A, the requested URIs are XML namespaces.

This document registers one YANG module in the 'YANG Module Names' registry, defined in .

name: ietf-system-datastore prefix: sysds namespace: urn:ietf:params:xml:ns:yang:ietf-system-datatstore maintained by IANA? N RFC: XXXX // RFC Ed.: replace XXXX and remove this comment

This section is modeled after the template described in . The "ietf-system-datastore" YANG module defines a data model that is designed to be accessed via YANG-based management protocols, such as NETCONF and RESTCONF . These protocols have to use a secure transport layer (e.g., SSH , TLS , and QUIC ) and have to use mutual authentication. The Network Configuration Access Control Model (NACM) 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. The YANG module only defines an identity that uses the "ds:conventional" identity as its base. The module by itself does not expose any data nodes that are writable, data nodes that contain read-only state, or RPCs. As such, there are no additional security issues related to the YANG module that need to be considered.

This section presents some sample data models and corresponding contents of various datastores with different dynamic behaviors described in . The XML snippets are used only for illustration purposes. Note this section does not show the contents of <intended> as they are related to the configuration in <operational> assuming the intended configuration is applied successfully.

In this subsection, the following fictional module is used:

A fictional ACL YANG module is used as follows, which defines a leafref for the leaf-list "application" data node to refer to an existing application name.

The server may predefine some applications as a convenience for clients, these applications are immediately-present system configuration. When the device is powered on, the system-instantiated application entries may be present in <system> as follows:

ftp

001

tcp

low

tftp

002

udp

low

smtp

003

tcp

low

]]> The client may also define its customized applications. Those applications may be present in <running> as follows:

my-smtp

101

tcp

2345

customized smtp application

high

my-foo

102

udp

1024

customized application ]]> If a client configures an ACL rule referencing some system-provided or customized applications, the configuration of ACL rule may be shown as follows:

allow-access-to-ftp-tftp

198.51.100.0/24

192.0.2.0/24

ftp tftp my-smtp

forward

]]> As different entries of application configuration in <system> and <running> are merged to create <intended> and there are no merging conflicts in the contents between <system> and <running>, <operational> might contain the configuration of applications with the values of origin reflecting the source of entries as follows:

my-smtp

101

tcp

2345

customized smtp application

high

my-foo

102

udp

1024

customized application ftp

001

tcp

low

tftp

002

udp

low

smtp

003

tcp

low

]]>

This subsection uses the following fictional interface YANG module:

Suppose the system provides an always-present loopback interface (named "lo0") with an MTU value "65536", a default IPv4 address of "127.0.0.1", and a default IPv6 address of "::1". The configuration of "lo0" interface is present in <system> as follows:

lo0 65536

127.0.0.1

::1

]]> A client modifies the value of MTU to 9216 by adding the following configuration into <running> using a "merge" operation:

lo0 9216 ]]> Since the MTU value provided by the client takes precedence over the system-provided value, and the "origin" value of configuration provided by the client is set to "intended", the configuration of interfaces that is present in <operational> is as follows:

lo0 9216

127.0.0.1

::1

]]>

Based on the example in , imagine the client further configures the description node of a "lo0" interface in <running> using a "merge" operation as follows:

lo0 loopback ]]> The configuration of interface "lo0" is present in <operational> as follows:

lo0 loopback 9216

127.0.0.1

::1

]]>

This section provides three use cases related to how <system> interacts with other datastores (e.g., <candidate>, <running>, <intended>, and <operational>). The following fictional interface data model is used:

For each use case, corresponding sample configuration in <running>, <system>, <intended> and <operational> are shown. The XML snippets are used only for illustration purposes.

When the device is powered on, suppose the system provides an always-present loopback interface (named "lo0") which is not explicitly configured in <running>. Thus, no configuration for interfaces appears in <running>; And the contents of <system> are:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> </interfaces> In this case, the configuration of loopback interface is only present in <system>, the configuration of interface in <intended> would be identical to the one in <system> shown above. And <operational> will show the system-provided loopback interface, note that <operational> also includes the default value specified in the YANG module:

<interfaces xmlns="urn:example:interfacemgmt" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:system"> <interface> <name>lo0</name> <type>loopback</type> <enabled or:origin="or:default">true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> </interfaces>

If a client creates an interface "et-0/0/0" but the interface does not physically exist at this point, what is in <running> appears as follows:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>et-0/0/0</name> <ip-address>192.168.10.10</ip-address> <description>pre-provisioned interface</description> </interface> </interfaces> And the contents of <system> remain unchanged, only containing the "lo" loopback interface, since the interface "et-0/0/0" is not physically present:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <ip-address>192.168.10.10</ip-address> <description>pre-provisioned interface</description> </interface> </interfaces> Since the interface named "et-0/0/0" does not exist, the associated configuration is not present in <operational>, which appears as follows:

When the interface is installed by the operator, the system will detect it and generate the associated conditionally-present interface configuration in <system>. The contents of <running> keep unchanged:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <description>system-defined interface</description> </interface> </interfaces> Then <intended> contains the merged configuration of <system> and <running>:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <ip-address>192.168.10.10</ip-address> <description>pre-provisioned interface</description> </interface> </interfaces> And the contents of <operational> appear as follows:

<interfaces xmlns="urn:example:interfacemgmt" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <interface or:origin="or:system"> <name>lo0</name> <type>loopback</type> <enabled or:origin="or:default">true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type or:origin="or:system">ethernet</type> <enabled or:origin="or:default">true</enabled> <ip-address>192.168.10.10</ip-address> <description>pre-provisioned interface</description> </interface> </interfaces>

If the client further sets the speed of interface "et-0/0/0" in <running> using a "merge" operation:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>et-0/0/0</name> <speed>10Mb</speed> </interface> </interfaces> The contents of <system> keep unchanged:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <description>system-defined interface</description> </interface> </interfaces> And the contents of <intended> which represents a merged results of <running> and <system> are as follows:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <ip-address>192.168.10.10</ip-address> <speed>10Mb</speed> <description>pre-provisioned interface</description> </interface> </interfaces> And <operational> would appear as follows:

<interfaces xmlns="urn:example:interfacemgmt" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <interface or:origin="or:system"> <name>lo0</name> <type>loopback</type> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>system-defined interface</description> </interface> <interface> <name>et-0/0/0</name> <type or:origin="or:system">ethernet</type> <enabled or:origin="or:default">true</enabled> <ip-address>192.168.10.10</ip-address> <speed>10Mb</speed> <description>pre-provisioned interface</description> </interface> </interfaces>

The authors would like to thank for following for discussions and providing input to this document: Balazs Lengyel, Robert Wilton, Juergen Schoenwaelder, Andy Bierman, Martin Bjorklund, Mohamed Boucadair, Michal Vaško, Alexander Clemm, and Timothy Carey.

Kent Watsen Watsen Networks Email: kent+ietf@watsen.net Jan Lindblad Cisco Systems Email: jlindbla@cisco.com Jason Sterne Nokia Email: jason.sterne@nokia.com Chongfeng Xie China Telecom Beijing China Email: xiechf@chinatelecom.cn