Internet-Draft | System-defined Configuration | February 2022 |
Ma, et al. | Expires 15 August 2022 | [Page] |
This document updates NMDA [RFC8342] to define a read-only conventional configuration datastore called "system" to hold system-defined configurations. To avoid clients' explicit copy/paste of referenced system-defined configuration, a "resolve-system" parameter has been defined to allow the server acting as a "system client" to populate referenced system-defined nodes automatically. The solution enables clients to reference nodes defined in <system>, overwrite values of configurations defined in <system>, and configure descendant nodes of system-defined nodes.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 15 August 2022.¶
Copyright (c) 2022 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.¶
NMDA [RFC8342] defines system configuration as the configuration that is supplied by the device itself and should be present in <operational> when it is in use.¶
However, there is a desire to enable a server to better document the system configuration. Clients can benefit from a standard mechanism to see what system configuration is available in a server.¶
In some cases, the client references a system configuration which isn't returned when the datastore is read. Having to copy the entire contents of the system configuration into <running> should be avoided or reduced when possible while ensuring that all referential integrity constraints are satisfied.¶
In some other cases, configuration of descendant nodes of system defined configuration needs to be supported. For example, the system configuration may contain an almost empty physical interface, while the client needs to be able to add, modify, remove a number of descendant nodes. Some descendant nodes may not be modifiable (e.g., "name" and "type" set by the system).¶
This document updates NMDA [RFC8342] to define a read-only conventional configuration datastore called "system" to hold system-defined configurations. To avoid clients' explicit copy/paste of referenced system-defined configuration, a "resolve-system" parameter has been defined to allow the server acting as a "system client" to populate referenced system-defined nodes automatically. The solution enables clients to reference nodes defined in <system>, overwrite values of configurations defined in <system>, and configure descendant nodes of system-defined nodes.¶
Conformance to this document requires servers to implement the "ietf-system-datastore" YANG Module.¶
This document assumes that the reader is familiar with the contents of [RFC6241], [RFC7950], [RFC8342], [RFC8407], and [RFC8525] and uses terminologies from those documents.¶
The following terms are defined in this document as follows:¶
Configuration that is provided by the system itself [RFC8342].¶
This document redefines the term "conventional configuration datastore" from RFC 8342 to add "system" to the list 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 [RFC2119] [RFC8174] 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 RFC 8342 to add "system" to the list conventional configuration datastores. The contents of <system> datastore are read-only to clients but may change dynamically. The <system> aware client may retrieve all three types of system configuration defined in Section 2, reference nodes defined in <system>, overwrite values of configurations defined in <system>, and configure descendant nodes of system-defined nodes.¶
The server will merge <running> and <system> to create <intended>. As always, system configuration will appear in <operational> with origin="system".¶
The <system> datastore makes system configuration visible to clients in order for being referenced or configurable prior to present in <operational>.¶
This document augments <edit-config> and <edit-data> RPC operations defined in [RFC6241] and [RFC8526] respectively, with a new additional input parameter "resolve-system".¶
The "resolve-system" parameter is optional and has no value. When it is provided and the server detects that there is a reference to a system-defined node during the validation, the server will automatically populate the referenced system configuration into the validated datastore to make the configuration valid without the client doing so explicitly. Legacy Clients interacting with servers that support this parameter don't see any changes in <edit-config> and <edit-data> behaviors.¶
According to the NETCONF constraint enforcement model defined in the section 8.3 of [RFC7950], if the target datastore of the <edit-config> or <edit-data> is "running" or "startup", the server-populating of the target datastore MUST be enforced at the end of the <edit-config> or <edit-data> operations during the validation. If the target datastore of the <edit-config> or <edit-data> is "candidate", the server-populating of the target datastore is delayed until a <commit> or <validate> operation takes place.¶
Comment: How should we update RESTCONF protocol to support the "resolve-system" parameter ?¶
There are three types of system configurations: immediately-active system configuration, conditionally-active system configuration and inactive-until-referenced system configuration.¶
Immediately-active system configurations are those applied and active immediately (e.g., a loop-back interface) , irrespective of physical resource present or not, a special functionality enabled or not.¶
System configurations which are provided and activated based on specific conditions being met in a system, e.g., if a physical resource is present (e.g., insert interface card), the system will automatically detect it and load pre-provisioned configuration; when the physical resource is not present( remove interface card), the system configuration will be automatically cleared. Another example is when a special functionality is enabled, e.g., when QoS function is enabled, QoS policies are automatically created by the system.¶
There are some predefined objects(e.g., application ids, anti-x signatures, trust anchor certs, etc.) as a convenience for the clients. The clients can also define their own data objects for their unique requirements. Inactive-until-referenced system configurations are not applied and active immediately but only after they are referenced by client-defined configuration.¶
From the client's perspective, the contents of the <system> datastore are read-only. There is no way to delete system configuration from a server. Any deletable system-provided configuration must be defined in <factory-default> [RFC8808], which is used to initialize <running> when the device is first-time powered on or reset to its factory default condition.¶
System configuration MAY change dynamically, e.g., depending on factors like device upgrade or if system-controlled resources(e.g., HW available) change. In some implementations, when QoS function is enabled, QoS-related predefined policies are created by system. If the system configuration gets changed, YANG notification (e.g., "push-change-update" notification) [RFC8641][RFC8639][RFC6470] can be used to notify the client.¶
This work intends to have no impact to <operational>. As always, system configuration will appear in <operational> with "origin=system". This work 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>. The referenced system configuration in <running> automatically copied from <system> by the server MUST have its origin set to "system" when present in <operational>. "Config false" nodes are completely out of scope, hence existing "config false" nodes are not impacted by this work.¶
This document introduces a mandatory datastore named "system" which is used to hold all three types of system configurations defined in Section 2.¶
When the device is powered on, immediately-active system configuration will be provided and activated immediately but inactive-until-referenced system configuration only becomes active if it is referenced by client-defined configuration. While conditionally-active system configuration will be created and immediately activated if the condition on system resources is met when the device is powered on or running.¶
All above three types of system configurations will go into <system>. The client may reference nodes defined in <system>, overwrite values of configurations defined in <system>, and configure descendant nodes of system-defined nodes in <running>. Then the server will merge <running> and <system> to create <intended>, in which process, <running> MAY overwrite and/or extend <system>. If a server implements <intended>, <system> MUST be merged into <intended>.¶
Servers MUST enforce that configuration references in <running> are resolved within the <running> datastore and ensure that <running> contains any referenced system objects. Clients MUST either explicitly configure system-defined nodes in <running> or use the "resolve-system" parameter. The server MUST enforce that the referenced system nodes injected into <running> by the client is consistent with <system>. Note that only <system> aware clients copy referenced system nodes from <system>. How clients unaware of the <system> datastore can find appropriate configurations are beyond the scope of this document.¶
No matter how the referenced system objects are populated, the nodes copied into <running> would always be returned in a read of <running>, regardless if the client is <system> aware.¶
This document defines a new parameter "resolve-system" to the input for the <edit-config> and <edit-data> operations. Clients that are aware of the "resolve-system" parameter MAY use this parameter to avoid the requirement to provide a referentially complete configuration in <running>.¶
If the "resolve-system" is present, the server MUST populate relevant referenced system-defined nodes into the target datastore without the client doing the copy/paste explicitly, to resolve any references not resolved by the client. The server acting as a "system client" like any other remote clients populates the referenced system-defined nodes when triggered by the "resolve-system" parameter. If the "resolve-system" parameter is not given by the client, the server MUST NOT modify <running> in any way not specified by the client.¶
The server may automatically configure the list entries (with at least the keys) in the target datastore (e.g., <running>) for any system configuration list entries that are referenced elsewhere by the clients. Not all the contents of the list entry (i.e., the descendant nodes) are necessarily populated by the sever - only the parts that are required to make the <running> valid. A read back of <running> (i.e., <get>, <get-config> or <get-data> operation) returns those automatically populated nodes.¶
It is also possible for a client to explicitly declare system configuration nodes in the target datastore (e.g., <running>) with the same values as in <system>. When a client configures a node (list entry, leaf, etc) in <running> that matches the same node and value in <system>, then that node becomes part of <running>. A read back of <running> returns those explicitly configured nodes.¶
This explicit configuration of system configuration in <running> can be useful, for example, when the client doesn't want a "system client" to have a role or hasn't implemented the "resolve-system" parameter. The client can explicitly declare (i.e. configure in <running>) the list entries (with at least the keys) for any system configuration list entries that are referenced elsewhere in <running>. Similarly, The client does not necessarily need to declare all the contents of the list entry (i.e. the descendant nodes) - only the parts that are required to make the <running> appear valid.¶
In some cases, a server may allow some parts of system configuration to be modified. List keys in system configuration can't be changed by a client, but other descendant nodes in a list entry may be modifiable or non-modifiable. Leafs and leaf-lists outside of lists may also be modifiable or non-modifiable. Modification of system configuration is achieved by the client writing configuration to <running> that overrides the system configuration. Client configuration statements in <running> take precedence over system configuration nodes in <system> if the server allows the nodes to be modified. If a system configuration node is non-modifiable, then writing a different value for that node in <running> MUST return an error.¶
A server may also allow a client to add data 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).¶
While modifying (overriding) system configuration nodes may be supported by a server, there is no mechanism for deleting a system configuration node. A "mandatory true" leaf, for example, may have a value in <system> which can be modified (overridden) by a client setting that leaf to a value in <running>. But the leaf could not be deleted.¶
Comment 1: What if <system> contains a set of values for a leaf-list, and a client configures another set of values for that leaf-list in <running>, will the set of values in <running> completely replace the set of values in <system>? Or the two sets of values are merged together?¶
Comment 2: how "ordered-by user" lists and leaf-lists are merged? Do the <running> values go before or after, or is this a case where a full-replace is needed.¶
This section shows the examples of server-populating of <running> automatically, declaring a system-defined node in <running> explicitly, modifying a system-instantiated leaf's value and configuring descendant nodes of a system-defined node. For each example, the corresponding XML snippets are provided.¶
In this subsection, the following fictional module is used:¶
module example-application { yang-version 1.1; namespace "urn:example:application"; prefix "app"; import ietf-inet-types { prefix "inet"; } container applications { list application { key "name"; leaf name { type string; } leaf protocol { type enumeration { enum tcp; enum udp; } } leaf destination-port { type inet:port-number; } } } }¶
The server may predefine some applications as a convenience for the clients. These predefined objects are applied only after being referenced by other configurations, which fall into the "inactive-until-referenced" system configuration as defined in Section 2. The system-instantiated application entries may be present in <system> as follows:¶
<applications xmlns="urn:example:application"> <application> <name>ftp</name> <protocol>tcp</protocol> <destination-port>21</destination-port> </application> <application> <name>tftp</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> <application> <name>smtp</name> <protocol>tcp</protocol> <destination-port>25</destination-port> </application> ... </applications>¶
The client may also define its customized applications. Suppose the configuration of applications is present in <running> as follows:¶
<applications xmlns="urn:example:application"> <application> <name>my-app-1</name> <protocol>tcp</protocol> <destination-port>2345</destination-port> </application> <application> <name>my-app-2</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> </applications>¶
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.¶
module example-acl { yang-version 1.1; namespace "urn:example:acl"; prefix "acl"; import example-application { prefix "app"; } import ietf-inet-types { prefix "inet"; } container acl { list acl_rule { key "name"; leaf name { type string; } container matches { choice l3 { container ipv4 { leaf source_address { type inet:ipv4-prefix; } leaf destination_address { type inet:ipv4-prefix; } } } choice applications { leaf-list application { type leafref { path "/app:applications/app:application/app:name"; } } } } leaf packet_action { type enumeration { enum forward; enum drop; enum redirect; } } } } }¶
If a client configures an ACL rule referencing system predefined nodes which are not present in <running>, the client MAY issue an <edit-config> operation with the parameter "resolve-system" as follows:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <edit-config> <target> <running/> </target> <config> <acl xmlns="urn:example:acl"> <acl_rule> <name>allow_access_to_ftp_tftp</name> <matches> <ipv4> <source_address>198.51.100.0/24</source_address> <destination_address>192.0.2.0/24</destination_address> </ipv4> <application>ftp</application> <application>tftp</application> <application>my-app-1</application> </matches> <packet_action>forward</packet_action> </acl_rule> </acl> </config> <resolve-system/> </edit-config> </rpc>¶
Then following gives the configuration of applications in <running> which is returned in the response to a follow-up <get-config> operation:¶
<applications xmlns="urn:example:application"> <application> <name>my-app-1</name> <protocol>tcp</protocol> <destination-port>2345</destination-port> </application> <application> <name>my-app-2</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> <application> <name>ftp</name> </application> <application> <name>tftp</name> </application> </applications>¶
Then the configuration of applications is present in <operational> as follows:¶
<applications xmlns="urn:example:application" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <application> <name>my-app-1</name> <protocol>tcp</protocol> <destination-port>2345</destination-port> </application> <application> <name>my-app-2</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> <application or:origin="or:system"> <name>ftp</name> <protocol>tcp</protocol> <destination-port>21</destination-port> </application> <application or:origin="or:system"> <name>tftp</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> </applications>¶
Since the configuration of application "smtp" is not referenced by the client, it does not appear in <operational> but only in <system>.¶
It's also possible for a client to explicitly declare the system-defined configurations that are referenced. For instance, in the above example, the client MAY also explicitly configure the following system defined applications "ftp" and "tftp" only with the list key "name" before referencing:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <edit-config> <target> <running/> </target> <config> <applications xmlns="urn:example:application"> <application> <name>ftp</name> </application> <application> <name>tftp</name> </application> </applications> </config> </edit-config> </rpc>¶
Then the client issues an <edit-config> operation to configure an ACL rule referencing applications "ftp" and "tftp" without the parameter "resolve-system" as follows:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <edit-config> <target> <running/> </target> <config> <acl xmlns="urn:example:acl"> <acl_rule> <name>allow_access_to_ftp_tftp</name> <matches> <ipv4> <source_address>198.51.100.0/24</source_address> <destination_address>192.0.2.0/24</destination_address> </ipv4> <application>ftp</application> <application>tftp</application> <application>my-app-1</application> </matches> <packet_action>forward</packet_action> </acl_rule> </acl> </config> </edit-config> </rpc>¶
Then following gives the configuration of applications in <running> which is returned in the response to a follow-up <get-config> operation, all the configuration of applications are explicitly configured by the client:¶
<applications xmlns="urn:example:application"> <application> <name>my-app-1</name> <protocol>tcp</protocol> <destination-port>2345</destination-port> </application> <application> <name>my-app-2</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> <application> <name>ftp</name> </application> <application> <name>tftp</name> </application> </applications>¶
Then the configuration of applications is present in <operational> as follows:¶
<applications xmlns="urn:example:application" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <application> <name>my-app-1</name> <protocol>tcp</protocol> <destination-port>2345</destination-port> </application> <application> <name>my-app-2</name> <protocol>udp</protocol> <destination-port>69</destination-port> </application> <application> <name>ftp</name> <protocol or:origin="or:system">tcp</protocol> <destination-port or:origin="or:system">21</destination-port> </application> <application> <name>tftp</name> <protocol or:origin="or:system">udp</protocol> <destination-port or:origin="or:system">69</destination-port> </application> </applications>¶
Since the application names "ftp" and "tftp" are explicitly configured by the client, they take precedence as the value in <system>, the "origin" attribute will be set to "intended".¶
In this subsection, we will use this fictional QoS data model:¶
module example-qos-policy { yang-version 1.1; namespace "urn:example:qos"; prefix "qos"; container qos-policies { list policy { key "name"; leaf name { type string; } list queue { key "queue-id"; leaf queue-id { type int32 { range "1..32"; } } leaf maximum-burst-size { type int32 { range "0..100"; } } } } } }¶
Suppose a client creates a qos policy "my-policy" with 4 system instantiated queues(1~4). The Configuration of qos-policies is present in <system> as follows:¶
<qos-policies xmlns="urn:example:qos"> <name>my-policy</name> <queue> <queue-id>1</queue-id> <maximum-burst-size>50</maximum-burst-size> </queue> <queue> <queue-id>2</queue-id> <maximum-burst-size>60</maximum-burst-size> </queue> <queue> <queue-id>3</queue-id> <maximum-burst-size>70</maximum-burst-size> </queue> <queue> <queue-id>4</queue-id> <maximum-burst-size>80</maximum-burst-size> </queue> </qos-policies>¶
A client modifies the value of maximum-burst-size to 55 in queue-id 1:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <edit-config> <target> <running/> </target> <config> <qos-policies xmlns="urn:example:qos"> <name>my-policy</name> <queue> <queue-id>1</queue-id> <maximum-burst-size>55</maximum-burst-size> </queue> </qos-policies> </config> </edit-config> </rpc>¶
Then the configuration of qos-policies is present in <operational> as follows:¶
<qos-policies xmlns="urn:example:qos" xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <name>my-policy</name> <queue> <queue-id>1</queue-id> <maximum-burst-size>55</maximum-burst-size> </queue> <queue or:origin="or:system"> <queue-id>2</queue-id> <maximum-burst-size>60</maximum-burst-size> </queue> <queue or:origin="or:system"> <queue-id>3</queue-id> <maximum-burst-size>70</maximum-burst-size> </queue> <queue or:origin="or:system"> <queue-id>4</queue-id> <maximum-burst-size>80</maximum-burst-size> </queue> </qos-policies>¶
This subsection also uses the fictional interface YANG module defined in Appendix C.3 of [RFC8342]. Suppose the system provides a loopback interface (named "lo0") with 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:¶
<interfaces> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
The configuration of "lo0" interface is present in <operational> as follows:¶
<interfaces xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:system"> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
Later on, the client further configures the description node of a "lo0" interface as follows:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <edit-config> <target> <running/> </target> <config> <interfaces> <interface> <name>lo0</name> <description>loopback</description> </interface> </interfaces> </config> </edit-config> </rpc>¶
Then the configuration of interface "lo0" is present in <operational> as follows:¶
<interfaces xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <interface> <name>lo0</name> <description>loopback</description> <ip-address or:origin="or:system">127.0.0.1</ip-address> <ip-address or:origin="or:system">::1</ip-address> </interface> </interfaces>¶
NMDA servers claiming to support this document MUST implement a <system> configuration datastore, and they SHOULD also implement the <intended> datastore.¶
Following guidelines for defining datastores in the appendix A of [RFC8342], this document introduces a new datastore resource named 'system' that represents the system configuration. A device MAY implement the mechanism defined in this document without implementing the "system" datastore, which would only eliminate the ability to programmatically determine the system configuration.¶
Origin: This document does not define any new origin identity when it interacts with <intended> datastore and finally flows into <operational>. The "system" origin Metadata Annotation [RFC7952] is used to indicate the origin of a data item is system.¶
The datastore's content is populated by the server and read-only to clients. Upon the content is created or changed, it will be merged into <intended> datastore. Unlike <factory-default>[RFC8808], it MAY change dynamically, e.g., depending on factors like during device upgrade or system-controlled resources(e.g., HW available) and the <system> datastore does not have to persist across reboots. <factory-reset> RPC operation defined in [RFC8808] can reset it to its factory default configuration without including configuration generated due to the system update or client-enabled functionality.¶
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> datastore support on the server by reading the YANG library information from the operational state datastore. Note that no new origin identity is defined in this document, the "or:system" origin Metadata Annotation [RFC7952] is used to indicate the origin of a data item is system. Support for the "origin" annotation is identified with the feature "origin" defined in [RFC8526].¶
The following diagram illustrates the relationship amongst the "identity" statements defined in the "ietf-system-datastore" and "ietf-datastores" YANG modules:¶
Identities: +--- datastore | +--- conventional | | +--- running | | +--- candidate | | +--- startup | | +--- system | | +--- intended | +--- dynamic | +--- operational The diagram above uses syntax that is similar to but not defined in [RFC8340].¶
This section gives an example of data retrieval from <system>. The YANG module used are shown in Appendix C.2 of [RFC8342]. All the messages are presented in a protocol-independent manner. JSON is used only for its conciseness.¶
Suppose the following data is added to <running>:¶
{ "bgp": { "local-as": "64501", "peer-as": "64502", "peer": { "name": "2001:db8::2:3" } } }¶
REQUEST (a <get-data> or GET request sent from the NETCONF or RESTCONF client):¶
Datastore: <system> Target:/bgp¶
An example of RESTCONF request:¶
GET /restconf/ds/system/bgp HTTP/1.1 Host: example.com Accept: application/yang-data+xml¶
RESPONSE ("local-port" leaf value is supplied by the system):¶
{ "bgp": { "peer": { "name": "2001:db8::2:3", "local-port": "60794" } } }¶
<CODE BEGINS> file="ietf-system-datastore@2021-05-14.yang" module ietf-system-datastore { yang-version 1.1; namespace "urn:ietf:params:xml:ns:yang:ietf-system-datastore"; prefix sysds; import ietf-datastores { prefix ds; reference "RFC 8342: Network Management Datastore Architecture(NMDA)"; } organization "IETF NETMDOD (Network Modeling) Working Group"; contact "WG Web: <http://tools.ietf.org/wg/netmod/> WG List: <mailto:netmod@ietf.org> Author: Qiufang Ma <mailto:maqiufang1@huawei.com> Author: Chong Feng <mailto:frank.fengchong@huawei.com> Author: Qin Wu <mailto:bill.wu@huawei.com>"; description "This module defines a new YANG identity that uses the ds:datastore identity defined in [RFC8342]. 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 Simplified 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 HHHH (https://www.rfc-editor.org/info/rfcHHHH); 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 2021-05-14 { description "Initial version."; reference "RFC XXXX: System-defined Configuration"; } identity system { base ds:conventional; description "This read-only datastore contains the complete configuration provided by the system itself."; } } <CODE ENDS>¶
This YANG module is optional to implement.¶
This YANG module augments NETCONF <edit-config> and <edit-data> operations with a new parameter "resolve-system" in the input parameters. If the "resolve-system" parameter is present, the server will populate the referenced system configuration into target datastore automatically. A NETCONF client can discover the "resolve-system" parameter support on the server by checking the YANG library information with "ietf-netconf-resolve-system" included from the operational state datastore.¶
Comment: How does a RESTCONF client know if the RESTCONF server implements the "resolve-system" parameter?¶
The following tree diagram [RFC8340] illustrates the "ietf-netconf-resolve-system" module:¶
module: ietf-netconf-resolve-system augment /nc:edit-config/nc:input: +---w resolve-system? empty augment /ncds:edit-data/ncds:input: +---w resolve-system? empty¶
The following tree diagram [RFC8340] illustrates "edit-config" and "edit-data" rpcs defined in "ietf-netconf" and "ietf-netconf-nmda" respectively, augmented by "ietf-netconf-resolve-system" YANG module :¶
rpcs: +---x edit-config | +---w input | +---w target | | +---w (config-target) | | +--:(candidate) | | | +---w candidate? empty {candidate}? | | +--:(running) | | +---w running? empty {writable-running}? | +---w default-operation? enumeration | +---w test-option? enumeration {validate}? | +---w error-option? enumeration | +---w (edit-content) | | +--:(config) | | | +---w config? <anyxml> | | +--:(url) | | +---w url? inet:uri {url}? | +---w resolve-system? empty +---x edit-data +---w input +---w datastore ds:datastore-ref +---w default-operation? enumeration +---w (edit-content) | +--:(config) | | +---w config? <anydata> | +--:(url) | +---w url? inet:uri {nc:url}? +---w resolve-system? empty¶
This section gives an example of an <edit-config> request to reference system-defined data nodes which are not present in <running> with a "resolve-system" parameter. A retrieval of <running> to show the auto-populated referenced system objects after the <edit-config> request is also given. The YANG module used is shown as follows, leafrefs refer to an existing name and address of an interface:¶
module example-interface-management { yang-version 1.1; namespace "urn:example:interfacemgmt"; prefix "inm"; container interfaces { list interface { key name; leaf name { type string; } leaf description { type string; } leaf mtu { type uint16; } leaf ip-address { type inet:ip-address; } } } container default-address { leaf ifname { type leafref { path "../../interfaces/interface/name"; } } leaf address { type leafref { path "../../interfaces/interface[name = current()/../ifname]" + "/ip-address"; } } } }¶
Image that the system provides a loopback interface (named "lo0") with a predefined MTU value of "1500" and a predefined IP address of "127.0.0.1". The <system> datastore shows the following configuration of loopback interface:¶
<interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <mtu>1500</mtu> <ip-address>127.0.0.1</ip-address> </interface> </interfaces>¶
The client sends an <edit-config> operation to add the configuration of default-address with a "resolve-system" parameter:¶
<rpc xmlns="urn:ietf:params:xml:ns:netconf:base:1.0" message-id="101"> <edit-config> <target> <running/> </target> <config> <default-address xmlns="urn:example:interfacemgmt"> <if-name>lo0</if-name> <address>127.0.0.1</address> </default-address> </config> </edit-config> <resolve-system/> </rpc>¶
Since the "resolve-system" parameter is provided, the server will resolve any leafrefs to system configurations and copy the referenced system-defined nodes into <running> automatically with the same value (i.e., the name and ip-address data nodes of lo0 interface) in <system> at the end of <edit-config> operation constraint enforcement. After the processing, a positive resonse is returned:¶
<rpc-reply message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <ok/> </rpc-reply>¶
Then the client sends a <get-config> operation towards <running>:¶
<rpc message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <get-config> <source> <running/> </source> <filter type="subtree"> <interfaces xmlns="urn:example:interfacemgmt"/> </filter> </get-config> </rpc>¶
Given that the referenced interface "name" and "ip-address" of lo0 are populated by the server, the following response is returned:¶
<rpc-reply message-id="101" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0"> <data> <interfaces xmlns="urn:example:interfacemgmt"> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> </interface> </interfaces> </data> </rpc-reply>¶
<CODE BEGINS> file="ietf-netconf-resolve-system@2021-05-14.yang" module ietf-netconf-resolve-system { yang-version 1.1; namespace "urn:ietf:params:xml:ns:yang:ietf-netconf-resolve-system"; prefix ncrs; import ietf-netconf { prefix nc; reference "RFC 6241: Network Configuration Protocol (NETCONF)"; } import ietf-netconf-nmda { prefix ncds; reference "RFC 8526: NETCONF Extensions to Support the Network Management Datastore Architecture"; } organization "IETF NETMOD (Network Modeling) Working Group"; contact "WG Web: <http://tools.ietf.org/wg/netmod/> WG List: <mailto:netmod@ietf.org> Author: Qiufang Ma <mailto:maqiufang1@huawei.com> Author: Chong Feng <mailto:frank.fengchong@huawei.com> Author: Qin Wu <mailto:bill.wu@huawei.com>"; description "This module defines an extension to the NETCONF protocol that allows the NETCONF client to control whether the server is allowed to populate referenced system configuration automatically without the client doing so explicitly. 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 Simplified 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 HHHH (https://www.rfc-editor.org/info/rfcHHHH); 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 2021-05-14 { description "Initial version."; reference "RFC XXXX: System-defined Configuration"; } augment /nc:edit-config/nc:input { description "Allows the server to automatically populate referenced system configuration to make configuration valid."; leaf resolve-system { type empty ; description "When present, the server is allowed to automatically populate referenced system configuration into <running>."; } } augment /ncds:edit-data/ncds:input { description "Allows the server to automatically populate referenced system configuration to make configuration valid."; leaf resolve-system { type empty ; description "When present, the server is allowed to automatically populate referenced system configuration into <running>."; } } } <CODE ENDS>¶
This document registers two XML namespace URNs in the 'IETF XML registry', following the format defined in [RFC3688].¶
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 [RFC6020] .¶
name: ietf-system-datastore prefix: sys namespace: urn:ietf:params:xml:ns:yang:ietf-system-datatstore 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 RFC: XXXX // RFC Ed.: replace XXXX and remove this comment¶
The YANG module defined in this document extends the base operations for NETCONF [RFC6241] and 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 users to a preconfigured subset of all available NETCONF protocol operations and content.¶
The YANG module defined in this document extends the base operations for NETCONF [RFC6241] and [RFC8526]. 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 users to a preconfigured subset of all available NETCONF protocol operations and content.¶
The security considerations for the base NETCONF protocol operations (see Section 9 of [RFC6241] apply to the new extended RPC operations defined in this document.¶
Chongfeng Xie China Telecom Beijing China Email: xiechf@chinatelecom.cn Kent Watsen Watsen Networks Email: kent+ietf@watsen.net Jason Sterne Nokia Email: jason.sterne@nokia.com¶
Thanks to Robert Wilton, Balazs Lengyel, Andy Bierman, Juergen Schoenwaelder, Alex Clemm, Martin Bjorklund, Timothy Carey for reviewing, and providing important input to, this document.¶
Following provides three use cases related to system-defined configuration lifecycle management. The simple interface data model defined in Appendix C.3 of [RFC8342] is used. For each use case, snippets of <running>, <system>, <intended> and <operational> are shown.¶
<running>:¶
No configuration for “lo0” appears in <running>;¶
<system>:¶
<interfaces> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
<intended>:¶
<interfaces> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
<operational>:¶
<interfaces xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:system"> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
If a client creates an interface "et-0/0/0" but the interface does not physically exist at this point:¶
<running>:¶
<interfaces> <interface> <name>et-0/0/0</name> <description>Test interface</description> </interface> </interfaces>¶
<system>:¶
<interfaces> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
<intended>:¶
<interfaces> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> <interface> <name>et-0/0/0</name> <description>Test interface</description> </interface> <interface> </interfaces>¶
<operational>:¶
<interfaces xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <interface or:origin="or:system"> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> </interfaces>¶
<running>:¶
<interfaces> <interface> <name>et-0/0/0</name> <description>Test interface</description> </interface> </interfaces>¶
<system>:¶
<interfaces> <interface> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> <interface> <name>et-0/0/0</name> <mtu>1500</mtu> </interface> </interfaces>¶
<intended>:¶
<interfaces> <name>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> <interface> <name>et-0/0/0</name> <description>Test interface</description> <mtu>1500</mtu> </interface> <interface> </interfaces>¶
<operational>:¶
<interfaces xmlns:or="urn:ietf:params:xml:ns:yang:ietf-origin" or:origin="or:intended"> <interface or:origin="or:system"> <name or:origin>lo0</name> <ip-address>127.0.0.1</ip-address> <ip-address>::1</ip-address> </interface> <interface> <name>et-0/0/0</name> <description>Test interface</description> <mtu or:origin="or:system">1500</mtu> </interface> <interface> </interfaces>¶
v00 - v01¶
v02 - v00¶