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 This document defines how a management client and management server handle YANG-modeled configuration data that is instantiated by the server itself. The system-defined configuration can be referenced (e.g., leafref) by configuration explicitly created by a client. The Network Management Datastore Architecture (NMDA) defined in RFC 8342 is updated with a read-only conventional configuration datastore called "system" to expose system-defined configuration. This document updates RFC 6241, RFC 8040, RFC 8342, and RFC 8526.

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 structure and expose the system configuration. NETCONF/RESTCONF clients can benefit from a standard mechanism to retrieve what system configuration is available on a server. Some servers allow clients to reference a system-defined node which is not present in the datastore. The absence of the system configuration in the datastore can render the datastore invalid from the perspective of a client or offline tools (e.g., missing leafref targets). This document describes several approaches to bring the datastore to a valid state and satisfy referential integrity constraints. 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, 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. As an alternative to clients explicitly copying referenced system-defined configuration so that the datastore is valid, a "resolve-system" parameter is defined to allow the server to copy referenced system nodes automatically. This solution enables clients to reference nodes defined in <system>, override system-provided values, and configure descendant nodes of system-defined configuration. If a system-defined node is referenced, it refers to one of the following cases throughout this document: It is present in a leafref "path" statement and referred as the leafref value. It is used as an "instance-identifier" type value. It is present in an XPath expression of "when" constraints. It is present in an XPath expression of "must" constraints. It is defined to satisfy the "mandatory true" constraints. It is defined to satisfy the "min-elements" constraints. Conformance to this document requires that 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: Configuration that is provided by the system itself. System configuration is present in the system configuration datastore (regardless of whether it is applied or referenced). It is a different and separate concept from factory default configuration defined in (which represents a preset initial configuration that is used to initialize the configuration of a server). 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.

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 to <running> (e.g., template expansion, removal of inactive configuration defined in ) have been performed (). The definition of "intended" origin metadata annotation identity is also updated ().

This document updates RFC 6241 to augment the NETCONF <edit-config>, <copy-config>, <validate>, and <commit> operations with an additional input parameter named "resolve-system", as specified in . This document also updates RFC 8526 to augment the NETCONF <edit-data> operation with the "resolve-system" parameter, as specified in .

This document extends Sections and in to add a new query parameter "resolve-system" and corresponding query parameter capability URI ().

This document defines two types of system configuration: configuration that is generated in <system> immediately when the device boots and configuration that is generated in <system> only when specific conditions being met on a device, they are described in and , respectively.

Immediately-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 immediately-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 associated configuration; when the physical resource is not present (an interface card is removed), the system configuration will be automatically cleared. 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" origin Metadata Annotation is used to indicate the origin of a data item in 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.

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

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) . In general, any update of <system> should not cause the automatic update of <running> to not surprise clients with unexpected changes. In particular, the behavior of system data migration during software upgrade is outside the scope of this document. That said, here are some examples of how a server might handle this scenario ensuring <running> remains accurate: Servers migrate system configuration update in <running>. Servers reject the operation to change system configuration (e.g., software upgrade fails) and needs the client to update the configuration in <running> as a prerequisite. Servers are recommended to include some hints in error responses to help clients understand how <running> should be updated.

This work intends to have no impact to <operational> and does not define any new origin identity beyond . The existence of <system> enables a subset of those 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.

Clients MAY reference nodes defined in <system>, override system-provided values, and configure descendant nodes of system-defined configuration in <running>, as detailed in , , and . 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>. If <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.

This document does not define any new origin identity when <system> interacts with <intended> and flows into <operational>. The "intended" identity of origin value defined in to represent the origin of configuration provided by <intended>, this document updates its definition as origin source of configuration explicitly provided by <running>, 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. Configuration copied from <system> into <running> has its origin value reported as "intended" when it flows into <operational>.

It is possible for a client to explicitly declare system configuration nodes with the same values as in <system>, by configuring a node (list/leaf-list entry, leaf, etc.) in the target datastore (e.g., <candidate> and <running>) that matches the same node and value in <system>. The explicit declaration of system-defined nodes that are referenced elsewhere can be useful, for example, when the client does not support the "resolve-system" parameter () but needs the datastore to be referentially complete. Clients MUST declare the system configuration that are required to make the datastore appear valid, which may include: any targets of leafrefs with "require-instance true". any targets of instance-identifiers with "require-instance true". any nodes referenced by any "when" expressions. any nodes referenced by any "must" expressions. any nodes needed to satisfy the "min-elements" statement with a value greater than zero. When declaring a node having descendants, clients MUST also declare all descendant nodes, including any leafs, leaf-lists, lists, presence containers, non-presence containers that have any child nodes.

This document defines a new parameter "resolve-system" to the input for some of the NETCONF and RESTCONF operations. Clients that are aware of the "resolve-system" parameter MAY use this parameter to avoid the requirement to provide a referentially complete configuration. The "resolve-system" parameter is optional and has no value. If it is present, and the server supports this capability, similar to , the server MUST copy the entire referenced system configuration, including all descendants into the target datastore (e.g., <candidate> and <running>) without the client doing the copy/paste explicitly, to resolve any references not resolved by the client. The copy operation MUST NOT override any explicit configuration in the target datastore. The server copies the referenced system-defined nodes only when triggered by the "resolve-system" parameter. Legacy clients don't see any changes in the server behaviors. There is no distinction between the configuration automatically configured by the server and the one explicitly declared by the client, e.g., a read back of the datastore (e.g., NETCONF <get>/<get-config>/<get-data> operation, or RESTCONF GET method) returns automatically configured nodes. Note that even though an auto-configured node is allowed to be deleted from the target datastore by the client, the system may automatically recreate the deleted node to make configuration valid, when a "resolve-system" parameter is carried. It is also possible that the operation request (e.g., <edit-config>) may not succeed due to incomplete referential integrity. Support for the "resolve-system" parameter is OPTIONAL. Servers not supporting NMDA MAY also implement this parameter without implementing the system configuration datastore, which would only eliminate the ability to retrieve the system configuration via protocol operations. If a server implements <system>, referenced system configuration is copied from <system> into the target datastore when the "resolve-system" parameter is used. If a server does not implement <system>, it is up to the implementation to determine how the "resolve-system" mechanism fills in the missing configuration items in the target datastore, e.g., <candidate> or <running>.

This document defines a NETCONF protocol capability to indicate support for this parameter. NETCONF server that supports "resolve-system" parameter MUST advertise the following capability identifier:

urn:ietf:params:netconf:capability:resolve-system:1.0

The "resolve-system" parameter may be present in the request URI of some RESTCONF operations as shown in . This parameter is only allowed with no values carried. If this parameter has any unexpected value, then a "400 Bad Request" status-line is returned.

To enable a RESTCONF client to discover if the "resolve-system" query parameter is supported by the server, the following capability URI is defined, which is advertised by the server if supported, using the "ietf-restconf-monitoring" module defined in :

urn:ietf:params:restconf:capability:resolve-system:1.0

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 to <running> that overrides the system configuration. Configurations defined in <running> take precedence over system configuration nodes in <system> if the server allows the nodes to be modified. For instance, descendant nodes in a system-defined list entry may be modifiable or not, even if some system configuration has been copied into <running> earlier. If a system node is non-modifiable, then writing a different value for that node MUST return an error during a <edit-config>, <validate> or <commit> operation, depending on the target datastore. The immutability of system configuration is defined in .

This section presents some sample data models and corresponding contents of various datastores with different dynamic behaviors described above. The XML snippets are used only for illustration purposes.

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. 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. Suppose the configuration of applications is present in <running> as follows:

my-app-1

101

tcp

2345

customized application

high

my-app-2

102

udp

customized application ]]> If a client configures an ACL rule referencing system-provided applications which are not present in <running>, it is possible for the client to explicitly declare the referenced system configuration. For instance, the client explicitly configuring the entire application entries named "ftp" and "tftp" are as follows:

ftp

001

tcp

low

tftp

002

udp

low

]]> And the configuration of ACL rules referencing application "ftp" and "tftp":

allow-access-to-ftp-tftp

198.51.100.0/24

192.0.2.0/24

ftp tftp my-app-1

forward

]]> And <operational> might contain the following:

my-app-1

101

tcp

2345

customized application

high

my-app-2

102

udp

customized application ftp

001

tcp

low

tftp

002

udp

low

smtp

003

tcp

low

]]>

In the above example, a client configures an ACL rule referencing system-provided applications which are not present in <running>, the client may also issue an <edit-config> operation with the parameter "resolve-system" to the NETCONF server as follows:

allow-access-to-ftp-tftp

198.51.100.0/24

192.0.2.0/24

ftp tftp my-app-1

forward

]]> The server receiving the "resolve-system" parameter copies the entire application list entries named "ftp" and "tftp" per . The following shows the configuration of applications in <running> which is returned in the response to a follow-up retrieval operation:

my-app-1

101

tcp

2345

customized application

high

my-app-2

102

udp

customized application ftp

001

tcp

low

tftp

002

udp

low

]]> Once the data is written into <running>, it makes no difference whether it is explicitly declared by the client or automatically copied by the server. The configuration for applications in <running> and <operational> would be identical to the ones in .

This subsection uses the following fictional interface YANG module:

Suppose the system provides a loopback interface (named "lo0") with a 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 and adds the following configuration into <running> using a "merge" operation:

lo0 9216 ]]> Then the configuration of interfaces is present in <operational> as follows:

lo0 9216

127.0.0.1

::1

]]>

In the above example, 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

]]>

<system> should not contain the configuration using the schema default value, either specified in the "default" statement or described in the "description" statement. Any value provided by the system that is not the schema default value MUST be contained in <system>. If system provides a value that is not the schema default value, and the node is not explicitly set by the client, it MUST be copied into the target datastore when its closest ancestor node needs to be copied to satisfy referential integrity constraints, when triggered by the "resolve-system" parameter.

Any deletable system-provided configuration that is populated as part of <running> by the system at boot up, without being part of the contents of a <startup> datastore, must be defined in <factory-default> , which is used to initialize <running> when the device is first-time powered on or reset to its factory default condition. Deletable system configuration must not be defined in <system>. The <factory-reset> RPC operation can reset <system> to its factory default contents.

If the device supports the :candidate or :private-candidate capability, a client may edit the candidate or private-candidate datastore without expecting it to be valid until a <commit> or <validate> operation takes place. The client may use the "resolve-system" parameter in one of the following situations: The client makes an edit (e.g., NETCONF <edit-config>/<edit-data>, or RESTCONF edit operation) to the candidate/private-candidate datastore. This is possible, though may not be required. The client issues a <validate> operation. The client issues a <commit> operation. In particular, defines the concept of conflict, the server's copy referenced system nodes triggered by "resolve-system" parameter might conflict with the contents of <running>, the conflict resolution is no different than the resolution of conflict caused by configuration explicitly provided by the client.

This YANG module defines a new YANG identity named "system" that uses the "ds:datastore" identity defined in [RFC8342]. 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:

This section gives an example of data retrieval from <system>. The fictional YANG module which imports type defined in is used as follows:

Suppose the following BGP peer configuration is added to <running> ( The message is presented in a protocol-independent manner. JSON is used to not imply a preferred encoding in this document):

Since both the "local-port" and "remote-port" nodes are not provided in <running>, and there is a default value specified for "remote-port", the system will select a value for "local-port". Note that per , the configuration using the schema default value described in the "description" statement will not be included in <system>. The following example shows the <get-data> RPC towards <system>:

ds:system

]]>

2001:db8::2:3

60794

]]>

<CODE BEGINS> file "ietf-system-datastore@2024-06-18.yang" Author: Qiufang Ma Author: Qin Wu Author: Chong Feng "; description "This module defines a new YANG identity that uses the ds:datastore identity defined in [RFC8342]. Copyright (c) 2024 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 2024-06-18 { 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 YANG module is optional to implement.

The following tree diagram illustrates the "ietf-netconf-resolve-system" module:

module: ietf-netconf-resolve-system augment /nc:edit-config/nc:input: +---w resolve-system? empty augment /nc:copy-config/nc:input: +---w resolve-system? empty augment /nc:validate/nc:input: +---w resolve-system? empty augment /nc:commit/nc:input: +---w resolve-system? empty augment /ncds:edit-data/ncds:input: +---w resolve-system? empty

Please refer to for example usage of the "resolve-system" parameter.

This module imports modules "ietf-netconf" and "ietf-netconf-nmda", defined in and , respectively.

<CODE BEGINS> file "ietf-netconf-resolve-system@2024-06-18.yang" WG List: Author: Qiufang Ma Author: Qin Wu Author: Chong Feng "; description "This module defines an extension to the NETCONF protocol that allows the NETCONF client to control whether the server is allowed to copy referenced system configuration automatically without the client doing so explicitly. Copyright (c) 2024 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. 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 (RFC 2119) (RFC 8174) when, and only when, they appear in all capitals, as shown here."; revision 2024-06-18 { description "Initial version."; reference "RFC XXXX: System-defined Configuration"; } grouping resolve-system-grouping { description "Define the resolve-system parameter grouping."; leaf resolve-system { type empty; description "When present, and the server supports this capability, the server MUST copy the entire referenced system configuration, including all descendants into the target datastore (e.g., and ) without the client doing the copy/paste explicitly, to resolve any references not resolved by the client. The copy operation MUST NOT override any explicit configuration in the target datastore."; } } augment "/nc:edit-config/nc:input" { description "Adds the 'resolve-system' parameter to the input of the NETCONF operation."; uses resolve-system-grouping; } augment "/nc:copy-config/nc:input" { description "Adds the 'resolve-system' parameter to the input of the NETCONF operation."; uses resolve-system-grouping; } augment "/nc:validate/nc:input" { description "Adds the 'resolve-system' parameter to the input of the NETCONF operation."; uses resolve-system-grouping; } augment "/nc:commit/nc:input" { description "Adds the 'resolve-system' parameter to the input of the NETCONF operation."; uses resolve-system-grouping; } augment "/ncds:edit-data/ncds:input" { description "Adds the 'resolve-system' parameter to the input of the NETCONF operation."; uses resolve-system-grouping; } } ]]> <CODE ENDS>

This document registers two XML namespace URNs 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. URI: urn:ietf:params:xml:ns:yang:ietf-netconf-resolve-system Registrant Contact: The IESG. XML: N/A, the requested URIs are XML namespaces.

This document registers two module names 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 name: ietf-netconf-resolve-system prefix: ncrs namespace: urn:ietf:params:xml:ns:yang:ietf-netconf-resolve-system maintained by IANA? N RFC: XXXX // RFC Ed.: replace XXXX and remove this comment

This document registers the following capability identifier URN in the 'Network Configuration Protocol (NETCONF) Capability URNs' registry:

urn:ietf:params:netconf:capability:resolve-system:1.0

This document registers a capability in the 'RESTCONF Capability URNs' registry :

Index Capability Identifier ----------------------------------------------------------------------- :resolve-system urn:ietf:params:restconf:capability:resolve-system:1.0

This section uses the template described in . The "ietf-system-datastore" YANG module defines a schema for data that is designed to be accessed via network management protocols such as NETCONF or RESTCONF . These network management protocols are required to use a secure transport layer and mutual authentication, e.g., SSH without the "none" authentication option, Transport Layer Security (TLS) with mutual X.509 authentication, and HTTPS with HTTP 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 a identity that uses the "ds:conventional" identity as its base. The module by itself does not expose any data nodes that are writable, date 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 uses the template described in . The "ietf-netconf-resolve-system" YANG module defines a schema for data that is designed to be accessed via network management protocols such as NETCONF or RESTCONF . These network management protocols are required to use a secure transport layer and mutual authentication, e.g., SSH without the "none" authentication option, Transport Layer Security (TLS) with mutual X.509 authentication, and HTTPS with HTTP 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 "ietf-netconf-resolve-system" YANG module extends the base operations of NETCONF protocol in and . The security considerations for the NETCONF protocol operations (see and ) apply to the extended RPC operations defined in this document. There is not any beyond the potential performance impacts of implementing the "resolve-system" parameter defined in the YANG module, which may mean employing some form of rate limiting or adapting the rate threshold for limiting might be a good idea to avoid DoS attacks.

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 a 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> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined 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:

<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>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined 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> <type>ethernet</type> <description>pre-provisioned interface</description> </interface> </interfaces> And the contents of <system> keep unchanged since the interface is not physically present:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <description>pre-provisioned interface</description> </interface> </interfaces> Since the interface named "eth-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 configuration in <system>. The contents of <running> keep unchanged:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <mtu>1500</mtu> <speed>100Mb</speed> </interface> </interfaces> Then <intended> contains the merged configuration of <system> and <running>:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <mtu>1500</mtu> <speed>100Mb</speed> <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>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <enabled or:origin="or:default">false</enabled> <mtu or:origin="or:system">1500</mtu> <speed or:origin="or:system">100Mb</speed> <description>pre-provisioned interface</description> </interface> </interfaces>

If the client further sets the speed of interface "eth-0/0/0" to a lower rate in <running> using a "merge" operation with the referenced node "type" being explicitly declared and enables this interface:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>et-0/0/0</name> <type>ethernet</type> <enabled>false</enabled> <mtu>1500</mtu> <speed>10Mb</speed> <description>pre-provisioned interface</description> </interface> </interfaces> The contents of <system> keep unchanged:

<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <type>loopback</type> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <mtu>1500</mtu> <speed>100Mb</speed> </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> <enabled>true</enabled> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> <description>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <enabled>true</enabled> <mtu>1500</mtu> <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>predefined interface</description> </interface> <interface> <name>et-0/0/0</name> <type>ethernet</type> <enabled>true</enabled> <mtu or:origin="or:system">1500</mtu> <speed>10Mb</speed> <description>pre-provisioned interface</description> </interface> </interfaces>

v05 - v06 remove inactive-until-referenced system config add a new section (sec.6) to clarify the interplay between system config and defaults add a new section (sec.7) to clarify relation to other datastores, which includes <factory-default> and <candidate>/<priv-candidate> leave the merge behavior of <system> and <running> unspecified augment <validate> and <commit> PRC operation to support "resolve-system" parameter editorial updates v04 - v05 Explicitly state that system configuration copied from <system> into <running> have its origin value being reported as "intended" and update the examples accordingly to reflect it Update the definition of "intended" origin identity in 8342 to allow a subset of configuration in <intended> to use "system" as origin value State server behaviors of migrating updated system data into <running> is beyond the scope of this document, and give a couple of implementation examples Remove the related statement which mandates referenced system configuration must be copied into <running> Refine usage examples (e.g., fix validation errors, remove redundancy) v03 - v04 Add some implementation consideration for "resolve-system" parameter Define a NETCONF capability identifier for "resolve-system" parameter so that the client can discover if it is supported by the server. state servers may upgrade copied system configuration in <running> as well during device upgrade or licensing change. v02 - v03 remove the merge mechanism related comments, as discussed in https://github.com/netconf-wg/netconf-next/issues/19 Editorial changes v01 - v02 Define referenced system configuration better clarify "resolve-system" parameter update Figure 2 in NMDA RFC Editorial changes v00 - v01 Clarify why client's explicit copy is not preferred but cannot be avoided if resolve-system parameter is not defined Clarify active system configuration Update the timing when the server's auto copy should be enforced if a resolve-system parameter is used Editorial changes

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 Chongfeng Xie China Telecom Beijing China Email: xiechf@chinatelecom.cn Jason Sterne Nokia Email: jason.sterne@nokia.com