Network Working Group R. Enns, Ed. Request for Comments: 4741 Juniper Networks Category: Standards Track December 2006 NETCONF Configuration Protocol Status of This Memo This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited. Copyright Notice Copyright (C) The IETF Trust (2006). Abstract The Network Configuration Protocol (NETCONF) defined in this document provides mechanisms to install, manipulate, and delete the configuration of network devices. It uses an Extensible Markup Language (XML)-based data encoding for the configuration data as well as the protocol messages. The NETCONF protocol operations are realized on top of a simple Remote Procedure Call (RPC) layer. Enns Standards Track [Page 1] RFC 4741 NETCONF Protocol December 2006 Table of Contents 1. Introduction ....................................................5 1.1. Protocol Overview ..........................................6 1.2. Capabilities ...............................................7 1.3. Separation of Configuration and State Data .................7 2. Transport Protocol Requirements .................................8 2.1. Connection-Oriented Operation ..............................9 2.2. Authentication, Integrity, and Confidentiality .............9 2.3. Authentication .............................................9 2.4. Mandatory Transport Protocol ..............................10 3. XML Considerations .............................................10 3.1. Namespace .................................................10 3.2. No Document Type Declarations .............................10 4. RPC Model ......................................................10 4.1. Element .............................................10 4.2. Element .......................................12 4.3. Element .......................................12 4.4. Element ..............................................16 4.5. Pipelining ................................................16 5. Configuration Model ............................................16 5.1. Configuration Datastores ..................................16 5.2. Data Modeling .............................................17 6. Subtree Filtering ..............................................17 6.1. Overview ..................................................17 6.2. Subtree Filter Components .................................18 6.2.1. Namespace Selection ................................18 6.2.2. Attribute Match Expressions ........................19 6.2.3. Containment Nodes ..................................19 6.2.4. Selection Nodes ....................................20 6.2.5. Content Match Nodes ................................20 6.3. Subtree Filter Processing .................................22 6.4. Subtree Filtering Examples ................................22 6.4.1. No Filter ..........................................22 6.4.2. Empty Filter .......................................23 6.4.3. Select the Entire Subtree ..................23 6.4.4. Select All Elements within the Subtree ....................................25 6.4.5. One Specific Entry ..........................26 6.4.6. Specific Elements from a Specific Entry .....27 6.4.7. Multiple Subtrees ..................................28 6.4.8. Elements with Attribute Naming .....................29 7. Protocol Operations ............................................31 7.1. ..............................................31 7.2. .............................................34 7.3. .............................................39 7.4. ...........................................41 7.5. ....................................................42 Enns Standards Track [Page 2] RFC 4741 NETCONF Protocol December 2006 7.6. ..................................................44 7.7. .....................................................45 7.8. ...........................................47 7.9. ............................................48 8. Capabilities ...................................................49 8.1. Capabilities Exchange .....................................49 8.2. Writable-Running Capability ...............................50 8.2.1. Description ........................................50 8.2.2. Dependencies .......................................50 8.2.3. Capability Identifier ..............................50 8.2.4. New Operations .....................................51 8.2.5. Modifications to Existing Operations ...............51 8.3. Candidate Configuration Capability ........................51 8.3.1. Description ........................................51 8.3.2. Dependencies .......................................52 8.3.3. Capability Identifier ..............................52 8.3.4. New Operations .....................................52 8.3.5. Modifications to Existing Operations ...............53 8.4. Confirmed Commit Capability ...............................55 8.4.1. Description ........................................55 8.4.2. Dependencies .......................................55 8.4.3. Capability Identifier ..............................56 8.4.4. New Operations .....................................56 8.4.5. Modifications to Existing Operations ...............56 8.5. Rollback on Error Capability ..............................57 8.5.1. Description ........................................57 8.5.2. Dependencies .......................................57 8.5.3. Capability Identifier ..............................57 8.5.4. New Operations .....................................57 8.5.5. Modifications to Existing Operations ...............57 8.6. Validate Capability .......................................58 8.6.1. Description ........................................58 8.6.2. Dependencies .......................................58 8.6.3. Capability Identifier ..............................58 8.6.4. New Operations .....................................58 8.7. Distinct Startup Capability ...............................60 8.7.1. Description ........................................60 8.7.2. Dependencies .......................................60 8.7.3. Capability Identifier ..............................60 8.7.4. New Operations .....................................60 8.7.5. Modifications to Existing Operations ...............60 8.8. URL Capability ............................................61 8.8.1. Description ........................................61 8.8.2. Dependencies .......................................61 8.8.3. Capability Identifier ..............................62 8.8.4. New Operations .....................................62 8.8.5. Modifications to Existing Operations ...............62 Enns Standards Track [Page 3] RFC 4741 NETCONF Protocol December 2006 8.9. XPath Capability ..........................................63 8.9.1. Description ........................................63 8.9.2. Dependencies .......................................63 8.9.3. Capability Identifier ..............................63 8.9.4. New Operations .....................................63 8.9.5. Modifications to Existing Operations ...............63 9. Security Considerations ........................................64 10. IANA Considerations ...........................................66 10.1. NETCONF XML Namespace ....................................66 10.2. NETCONF XML Schema .......................................66 10.3. NETCONF Capability URNs ..................................66 11. Authors and Acknowledgements ..................................68 12. References ....................................................68 12.1. Normative References .....................................68 12.2. Informative References ...................................69 Appendix A. NETCONF Error List ....................................70 Appendix B. XML Schema for NETCONF RPC and Protocol Operations ....74 Appendix C. Capability Template ...................................86 C.1. capability-name (template) ................................86 C.1.1. Overview ...........................................86 C.1.2. Dependencies .......................................86 C.1.3. Capability Identifier ..............................86 C.1.4. New Operations .....................................86 C.1.5. Modifications to Existing Operations ...............86 C.1.6. Interactions with Other Capabilities ...............86 Appendix D. Configuring Multiple Devices with NETCONF ............87 D.1. Operations on Individual Devices ..........................87 D.1.1. Acquiring the Configuration Lock ...................87 D.1.2. Loading the Update .................................88 D.1.3. Validating the Incoming Configuration ..............89 D.1.4. Checkpointing the Running Configuration ............89 D.1.5. Changing the Running Configuration .................90 D.1.6. Testing the New Configuration ......................91 D.1.7. Making the Change Permanent ........................91 D.1.8. Releasing the Configuration Lock ...................92 D.2. Operations on Multiple Devices ............................92 Appendix E. Deferred Features .....................................93 Enns Standards Track [Page 4] RFC 4741 NETCONF Protocol December 2006 1. Introduction The NETCONF protocol defines a simple mechanism through which a network device can be managed, configuration data information can be retrieved, and new configuration data can be uploaded and manipulated. The protocol allows the device to expose a full, formal application programming interface (API). Applications can use this straightforward API to send and receive full and partial configuration data sets. The NETCONF protocol uses a remote procedure call (RPC) paradigm. A client encodes an RPC in XML [1] and sends it to a server using a secure, connection-oriented session. The server responds with a reply encoded in XML. The contents of both the request and the response are fully described in XML DTDs or XML schemas, or both, allowing both parties to recognize the syntax constraints imposed on the exchange. A key aspect of NETCONF is that it allows the functionality of the management protocol to closely mirror the native functionality of the device. This reduces implementation costs and allows timely access to new features. In addition, applications can access both the syntactic and semantic content of the device's native user interface. NETCONF allows a client to discover the set of protocol extensions supported by a server. These "capabilities" permit the client to adjust its behavior to take advantage of the features exposed by the device. The capability definitions can be easily extended in a noncentralized manner. Standard and non-standard capabilities can be defined with semantic and syntactic rigor. Capabilities are discussed in Section 8. The NETCONF protocol is a building block in a system of automated configuration. XML is the lingua franca of interchange, providing a flexible but fully specified encoding mechanism for hierarchical content. NETCONF can be used in concert with XML-based transformation technologies, such as XSLT [8], to provide a system for automated generation of full and partial configurations. The system can query one or more databases for data about networking topologies, links, policies, customers, and services. This data can be transformed using one or more XSLT scripts from a task-oriented, vendor-independent data schema into a form that is specific to the vendor, product, operating system, and software release. The resulting data can be passed to the device using the NETCONF protocol. Enns Standards Track [Page 5] RFC 4741 NETCONF Protocol December 2006 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [3]. 1.1. Protocol Overview NETCONF uses a simple RPC-based mechanism to facilitate communication between a client and a server. The client can be a script or application typically running as part of a network manager. The server is typically a network device. The terms "device" and "server" are used interchangeably in this document, as are "client" and "application". A NETCONF session is the logical connection between a network administrator or network configuration application and a network device. A device MUST support at least one NETCONF session and SHOULD support multiple sessions. Global configuration attributes can be changed during any authorized session, and the effects are visible in all sessions. Session-specific attributes affect only the session in which they are changed. NETCONF can be conceptually partitioned into four layers: Layer Example +-------------+ +-----------------------------+ (4) | Content | | Configuration data | +-------------+ +-----------------------------+ | | +-------------+ +-----------------------------+ (3) | Operations | | , | +-------------+ +-----------------------------+ | | +-------------+ +-----------------------------+ (2) | RPC | | , | +-------------+ +-----------------------------+ | | +-------------+ +-----------------------------+ (1) | Transport | | BEEP, SSH, SSL, console | | Protocol | | | +-------------+ +-----------------------------+ 1. The transport protocol layer provides a communication path between the client and server. NETCONF can be layered over any transport protocol that provides a set of basic requirements. Section 2 discusses these requirements. 2. The RPC layer provides a simple, transport-independent framing mechanism for encoding RPCs. Section 4 documents this protocol. Enns Standards Track [Page 6] RFC 4741 NETCONF Protocol December 2006 3. The operations layer defines a set of base operations invoked as RPC methods with XML-encoded parameters. Section 7 details the list of base operations. 4. The content layer is outside the scope of this document. Given the current proprietary nature of the configuration data being manipulated, the specification of this content depends on the NETCONF implementation. It is expected that a separate effort to specify a standard data definition language and standard content will be undertaken. 1.2. Capabilities A NETCONF capability is a set of functionality that supplements the base NETCONF specification. The capability is identified by a uniform resource identifier (URI). These URIs should follow the guidelines as described in Section 8. Capabilities augment the base operations of the device, describing both additional operations and the content allowed inside operations. The client can discover the server's capabilities and use any additional operations, parameters, and content defined by those capabilities. The capability definition may name one or more dependent capabilities. To support a capability, the server MUST support any capabilities upon which it depends. Section 8 defines the capabilities exchange that allows the client to discover the server's capabilities. Section 8 also lists the set of capabilities defined in this document. Additional capabilities can be defined at any time in external documents, allowing the set of capabilities to expand over time. Standards bodies may define standardized capabilities, and implementations may define proprietary ones. A capability URI MUST sufficiently distinguish the naming authority to avoid naming collisions. 1.3. Separation of Configuration and State Data The information that can be retrieved from a running system is separated into two classes, configuration data and state data. Configuration data is the set of writable data that is required to transform a system from its initial default state into its current state. State data is the additional data on a system that is not Enns Standards Track [Page 7] RFC 4741 NETCONF Protocol December 2006 configuration data such as read-only status information and collected statistics. When a device is performing configuration operations, a number of problems would arise if state data were included: o Comparisons of configuration data sets would be dominated by irrelevant entries such as different statistics. o Incoming data could contain nonsensical requests, such as attempts to write read-only data. o The data sets would be large. o Archived data could contain values for read-only data items, complicating the processing required to restore archived data. To account for these issues, the NETCONF protocol recognizes the difference between configuration data and state data and provides operations for each. The operation retrieves configuration data only, while the operation retrieves configuration and state data. Note that the NETCONF protocol is focused on the information required to get the device into its desired running state. The inclusion of other important, persistent data is implementation specific. For example, user files and databases are not treated as configuration data by the NETCONF protocol. If a local database of user authentication data is stored on the device, whether it is included in configuration data is an implementation-dependent matter. 2. Transport Protocol Requirements NETCONF uses an RPC-based communication paradigm. A client sends a series of one or more RPC request operations, which cause the server to respond with a corresponding series of RPC replies. The NETCONF protocol can be layered on any transport protocol that provides the required set of functionality. It is not bound to any particular transport protocol, but allows a mapping to define how it can be implemented over any specific protocol. The transport protocol MUST provide a mechanism to indicate the session type (client or server) to the NETCONF protocol layer. This section details the characteristics that NETCONF requires from the underlying transport protocol. Enns Standards Track [Page 8] RFC 4741 NETCONF Protocol December 2006 2.1. Connection-Oriented Operation NETCONF is connection-oriented, requiring a persistent connection between peers. This connection must provide reliable, sequenced data delivery. NETCONF connections are long-lived, persisting between protocol operations. This allows the client to make changes to the state of the connection that will persist for the lifetime of the connection. For example, authentication information specified for a connection remains in effect until the connection is closed. In addition, resources requested from the server for a particular connection MUST be automatically released when the connection closes, making failure recovery simpler and more robust. For example, when a lock is acquired by a client, the lock persists until either it is explicitly released or the server determines that the connection has been terminated. If a connection is terminated while the client holds a lock, the server can perform any appropriate recovery. The lock operation is further discussed in Section 7.5. 2.2. Authentication, Integrity, and Confidentiality NETCONF connections must provide authentication, data integrity, and confidentiality. NETCONF depends on the transport protocol for this capability. A NETCONF peer assumes that appropriate levels of security and confidentiality are provided independently of this document. For example, connections may be encrypted in TLS [9] or SSH [10], depending on the underlying protocol. 2.3. Authentication NETCONF connections must be authenticated. The transport protocol is responsible for authentication. The peer assumes that the connection's authentication information has been validated by the underlying protocol using sufficiently trustworthy mechanisms and that the peer's identity has been sufficiently proven. One goal of NETCONF is to provide a programmatic interface to the device that closely follows the functionality of the device's native interface. Therefore, it is expected that the underlying protocol uses existing authentication mechanisms defined by the device. For example, a device that supports RADIUS [11] should allow the use of RADIUS to authenticate NETCONF sessions. The authentication process should result in an identity whose permissions are known to the device. These permissions MUST be enforced during the remainder of the NETCONF session. Enns Standards Track [Page 9] RFC 4741 NETCONF Protocol December 2006 2.4. Mandatory Transport Protocol A NETCONF implementation MUST support the SSH transport protocol mapping [4]. 3. XML Considerations XML serves as the encoding format for NETCONF, allowing complex hierarchical data to be expressed in a text format that can be read, saved, and manipulated with both traditional text tools and tools specific to XML. This section discusses a small number of XML-related considerations pertaining to NETCONF. 3.1. Namespace All NETCONF protocol elements are defined in the following namespace: urn:ietf:params:xml:ns:netconf:base:1.0 NETCONF capability names MUST be URIs [5]. NETCONF capabilities are discussed in Section 8. 3.2. No Document Type Declarations Document type declarations MUST NOT appear in NETCONF content. 4. RPC Model The NETCONF protocol uses an RPC-based communication model. NETCONF peers use and elements to provide transport protocol-independent framing of NETCONF requests and responses. 4.1. Element The element is used to enclose a NETCONF request sent from the client to the server. The element has a mandatory attribute "message-id", which is an arbitrary string chosen by the sender of the RPC that will commonly encode a monotonically increasing integer. The receiver of the RPC does not decode or interpret this string but simply saves it to be used as a "message-id" attribute in any resulting message. For example: Enns Standards Track [Page 10] RFC 4741 NETCONF Protocol December 2006 If additional attributes are present in an element, a NETCONF peer MUST return them unmodified in the element. The name and parameters of an RPC are encoded as the contents of the element. The name of the RPC is an element directly inside the element, and any parameters are encoded inside this element. The following example invokes a method called , which has two parameters, , with a value of "14", and , with a value of "fred": 14 fred The following example invokes a method with a parameter of "27606-0100": 27606-0100 The following example invokes the NETCONF method with no parameters: Enns Standards Track [Page 11] RFC 4741 NETCONF Protocol December 2006 4.2. Element The message is sent in response to an operation. The element has a mandatory attribute "message-id", which is equal to the "message-id" attribute of the for which this is a response. A NETCONF peer MUST also return any additional attributes included in the element unmodified in the element. The response name and response data are encoded as the contents of the element. The name of the reply is an element directly inside the element, and any data is encoded inside this element. For example: The following element invokes the NETCONF method and includes an additional attribute called "user-id". Note that the "user-id" attribute is not in the NETCONF namespace. The returned element returns the "user-id" attribute, as well as the requested content. 4.3. Element The element is sent in messages if an error occurs during the processing of an request. If a server encounters multiple errors during the processing of an request, the MAY contain multiple elements. However, a server is not required to detect or report more Enns Standards Track [Page 12] RFC 4741 NETCONF Protocol December 2006 than one element, if a request contains multiple errors. A server is not required to check for particular error conditions in a specific sequence. A server MUST return an element if any error conditions occur during processing and SHOULD return an element if any warning conditions occur during processing. A server MUST NOT return application-level- or data-model-specific error information in an element for which the client does not have sufficient access rights. The element includes the following information: error-type: Defines the conceptual layer that the error occurred. Enumeration. One of: * transport * rpc * protocol * application error-tag: Contains a string identifying the error condition. See Appendix A for allowed values. error-severity: Contains a string identifying the error severity, as determined by the device. One of: * error * warning error-app-tag: Contains a string identifying the data-model-specific or implementation-specific error condition, if one exists. This element will not be present if no appropriate application error tag can be associated with a particular error condition. error-path: Contains the absolute XPath [2] expression identifying the element path to the node that is associated with the error being reported in a particular rpc-error element. This element will not be present if no appropriate payload element can be associated with a particular error condition, or if the 'bad-element' QString returned in the 'error-info' container is sufficient to identify the node associated with the error. When Enns Standards Track [Page 13] RFC 4741 NETCONF Protocol December 2006 the XPath expression is interpreted, the set of namespace declarations are those in scope on the rpc-error element, including the default namespace. error-message: Contains a string suitable for human display that describes the error condition. This element will not be present if no appropriate message is provided for a particular error condition. This element SHOULD include an xml:lang attribute as defined in [1] and discussed in [12]. error-info: Contains protocol- or data-model-specific error content. This element will not be present if no such error content is provided for a particular error condition. The list in Appendix A defines any mandatory error-info content for each error. After any protocol-mandated content, a data model definition may mandate that certain application-layer error information be included in the error-info container. An implementation may include additional elements to provide extended and/or implementation- specific debugging information. Appendix A enumerates the standard NETCONF errors. Example: An error is returned if an element is received without a message-id attribute. Note that only in this case is it acceptable for the NETCONF peer to omit the message-id attribute in the element. rpc missing-attribute error message-id rpc Enns Standards Track [Page 14] RFC 4741 NETCONF Protocol December 2006 The following illustrates the case of returning multiple elements. Note that the data models used in the examples in this section use the element to distinguish between multiple instances of the element. application invalid-value error MTU value 25000 is not within range 256..9192 Ethernet0/0 25000 application invalid-value error Invalid IP address for interface Ethernet1/0 Ethernet1/0
1.4 24
Enns Standards Track [Page 15] RFC 4741 NETCONF Protocol December 2006 4.4. Element The element is sent in messages if no errors or warnings occurred during the processing of an request. For example: 4.5. Pipelining NETCONF requests MUST be processed serially by the managed device. Additional requests MAY be sent before previous ones have been completed. The managed device MUST send responses only in the order the requests were received. 5. Configuration Model NETCONF provides an initial set of operations and a number of capabilities that can be used to extend the base. NETCONF peers exchange device capabilities when the session is initiated as described in Section 8.1. 5.1. Configuration Datastores NETCONF defines the existence of one or more configuration datastores and allows configuration operations on them. A configuration datastore is defined as the complete set of configuration data that is required to get a device from its initial default state into a desired operational state. The configuration datastore does not include state data or executive commands. Only the configuration datastore is present in the base model. Additional configuration datastores may be defined by capabilities. Such configuration datastores are available only on devices that advertise the capabilities. o Running: The complete configuration currently active on the network device. Only one configuration datastore of this type exists on the device, and it is always present. NETCONF protocol operations refer to this datastore using the element. The capabilities in Sections 8.3 and 8.7 define the and configuration datastores, respectively. Enns Standards Track [Page 16] RFC 4741 NETCONF Protocol December 2006 5.2. Data Modeling Data modeling and content issues are outside the scope of the NETCONF protocol. An assumption is made that the device's data model is well-known to the application and that both parties are aware of issues such as the layout, containment, keying, lookup, replacement, and management of the data, as well as any other constraints imposed by the data model. NETCONF carries configuration data inside the element that is specific to device's data model. The protocol treats the contents of that element as opaque data. The device uses capabilities to announce the set of data models that the device implements. The capability definition details the operation and constraints imposed by data model. Devices and managers may support multiple data models, including both standard and proprietary data models. 6. Subtree Filtering 6.1. Overview XML subtree filtering is a mechanism that allows an application to select particular XML subtrees to include in the for a or operation. A small set of filters for inclusion, simple content exact-match, and selection is provided, which allows some useful, but also very limited, selection mechanisms. The agent does not need to utilize any data-model- specific semantics during processing, allowing for simple and centralized implementation strategies. Conceptually, a subtree filter is comprised of zero or more element subtrees, which represent the filter selection criteria. At each containment level within a subtree, the set of sibling nodes is logically processed by the server to determine if its subtree and path of elements to the root are included in the filter output. All elements present in a particular subtree within a filter must match associated nodes present in the server's conceptual data model. XML namespaces may be specified (via 'xmlns' declarations) within the filter data model. If they are, the declared namespace must first exactly match a namespace supported by the server. Note that prefix values for qualified namespaces are not relevant when comparing filter elements to elements in the underlying data model. Only data associated with a specified namespace will be included in the filter output. Enns Standards Track [Page 17] RFC 4741 NETCONF Protocol December 2006 Each node specified in a subtree filter represents an inclusive filter. Only associated nodes in underlying data model(s) within the specified configuration datastore on the server are selected by the filter. A node must exactly match the namespace and hierarchy of elements given in the filter data, except that the filter absolute path name is adjusted to start from the layer below . Response messages contain only the subtrees selected by the filter. Any selection criteria that were present in the request, within a particular selected subtree, are also included in the response. Note that some elements expressed in the filter as leaf nodes will be expanded (i.e., subtrees included) in the filter output. Specific data instances are not duplicated in the response in the event that the request contains multiple filter subtree expressions that select the same data. 6.2. Subtree Filter Components A subtree filter is comprised of XML elements and their XML attributes. There are five types of components that may be present in a subtree filter: o Namespace Selection o Attribute Match Expressions o Containment Nodes o Selection Nodes o Content Match Nodes 6.2.1. Namespace Selection If namespaces are used, then the filter output will only include elements from the specified namespace. A namespace is considered to match (for filter purposes) if the content of the 'xmlns' attributes are the same in the filter and the underlying data model. Note that namespace selection cannot be used by itself. At least one element must be specified in the filter any elements to be included in the filter output. Example: Enns Standards Track [Page 18] RFC 4741 NETCONF Protocol December 2006 In this example, the element is a selection node, and only this node and any child nodes (from the underlying data model) in the 'http://example.com/schema/1.2/config' namespace will be included in the filter output. 6.2.2. Attribute Match Expressions An attribute that appears in a subtree filter is part of an "attribute match expression". Any number of (unqualified or qualified) XML attributes may be present in any type of filter node. In addition to the selection criteria normally applicable to that node, the selected data must have matching values for every attribute specified in the node. If an element is not defined to include a specified attribute, then it is not selected in the filter output. Example: In this example, the , , and elements are containment nodes, and 'ifName' is an attribute match expression. Only 'interface' nodes in the 'http://example.com/schema/1.2/config' namespace that have an 'ifName' attribute with the value 'eth0' and occur within 'interfaces' nodes within 'top' nodes will be included in the filter output. 6.2.3. Containment Nodes Nodes that contain child elements within a subtree filter are called "containment nodes". Each child element can be any type of node, including another containment node. For each containment node specified in a subtree filter, all data model instances that exactly match the specified namespaces, element hierarchy, and any attribute match expressions are included in the filter output. Example: Enns Standards Track [Page 19] RFC 4741 NETCONF Protocol December 2006 In this example, the element is a containment node. 6.2.4. Selection Nodes An empty leaf node within a filter is called a "selection node", and it represents an "explicit selection" filter on the underlying data model. Presence of any selection nodes within a set of sibling nodes will cause the filter to select the specified subtree(s) and suppress automatic selection of the entire set of sibling nodes in the underlying data model. For filtering purposes, an empty leaf node can be declared either with an empty tag (e.g., ) or with explicit start and end tags (e.g., ). Any whitespace characters are ignored in this form. Example: In this example, the element is a containment node, and the element is a selection node. Only 'users' nodes in the 'http://example.com/schema/1.2/config' namespace that occur within a 'top' element that is the root of the configuration datastore will be included in the filter output. 6.2.5. Content Match Nodes A leaf node that contains simple content is called a "content match node". It is used to select some or all of its sibling nodes for filter output, and it represents an exact-match filter on the leaf node element content. The following constraints apply to content match nodes: o A content match node must not contain nested elements (i.e., must resolve to a simpleType in the XML Schema Definition (XSD)). o Multiple content match nodes (i.e., sibling nodes) are logically combined in an "AND" expression. o Filtering of mixed content is not supported. o Filtering of list content is not supported. o Filtering of whitespace-only content is not supported. Enns Standards Track [Page 20] RFC 4741 NETCONF Protocol December 2006 o A content match node must contain non-whitespace characters. An empty element (e.g., ) will be interpreted as a selection node (e.g., ). o Leading and trailing whitespace characters are ignored, but any whitespace characters within a block of text characters are not ignored or modified. If all specified sibling content match nodes in a subtree filter expression are 'true', then the filter output nodes are selected in the following manner: o Each content match node in the sibling set is included in the filter output. o If any containment nodes are present in the sibling set, then they are processed further and included if any nested filter criteria are also met. o If any selection nodes are present in the sibling set, then all of them are included in the filter output. o Otherwise (i.e., there are no selection or containment nodes in the filter sibling set), all the nodes defined at this level in the underlying data model (and their subtrees, if any) are returned in the filter output. If any of the sibling content match node tests are 'false', then no further filter processing is performed on that sibling set, and none of the sibling subtrees are selected by the filter, including the content match node(s). Example: fred In this example, the and nodes are both containment nodes, and is a content match node. Since no sibling nodes of are specified (and therefore no containment or selection nodes), all of the sibling nodes of are returned in the filter Enns Standards Track [Page 21] RFC 4741 NETCONF Protocol December 2006 output. Only 'user' nodes in the 'http://example.com/schema/1.2/config' namespace that match the element hierarchy and for which the element is equal to 'fred' will be included in the filter output. 6.3. Subtree Filter Processing The filter output (the set of selected nodes) is initially empty. Each subtree filter can contain one or more data model fragments, which represent portions of the data model that should be selected (with all child nodes) in the filter output. Each subtree data fragment is compared by the server to the internal data models supported by the server. If the entire subtree data- fragment filter (starting from the root to the innermost element specified in the filter) exactly matches a corresponding portion of the supported data model, then that node and all its children are included in the result data. The server processes all nodes with the same parent node (sibling set) together, starting from the root to the leaf nodes. The root elements in the filter are considered in the same sibling set (assuming they are in the same namespace), even though they do not have a common parent. For each sibling set, the server determines which nodes are included (or potentially included) in the filter output, and which sibling subtrees are excluded (pruned) from the filter output. The server first determines which types of nodes are present in the sibling set and processes the nodes according to the rules for their type. If any nodes in the sibling set are selected, then the process is recursively applied to the sibling sets of each selected node. The algorithm continues until all sibling sets in all subtrees specified in the filter have been processed. 6.4. Subtree Filtering Examples 6.4.1. No Filter Leaving out the filter on the get operation returns the entire data model. Enns Standards Track [Page 22] RFC 4741 NETCONF Protocol December 2006 6.4.2. Empty Filter An empty filter will select nothing because no content match or selection nodes are present. This is not an error. The filter type attribute used in these examples is discussed further in Section 7.1. 6.4.3. Select the Entire Subtree The filter in this example contains one selection node (), so just that subtree is selected by the filter. This example represents the fully-populated data model in most of the filter examples that follow. In a real data model, the would not likely be returned with the list of users for a particular host or network. NOTE: The filtering and configuration examples used in this document appear in the namespace "http://example.com/schema/1.2/config". The root element of this namespace is . The element and its descendents represent an example configuration data model only. Enns Standards Track [Page 23] RFC 4741 NETCONF Protocol December 2006 root superuser Charlie Root 1 1 fred admin Fred Flintstone 2 2 barney admin Barney Rubble 2 3 The following filter request would have produced the same result, but only because the container defines one child element (). Enns Standards Track [Page 24] RFC 4741 NETCONF Protocol December 2006 6.4.4. Select All Elements within the Subtree This filter contains two containment nodes (, ) and one selector node (). All instances of the element in the same sibling set are selected in the filter output. The manager may need to know that is used as an instance identifier in this particular data structure, but the server does not need to know that meta-data in order to process the request. Enns Standards Track [Page 25] RFC 4741 NETCONF Protocol December 2006 root fred barney 6.4.5. One Specific Entry This filter contains two containment nodes (, ) and one content match node (). All instances of the sibling set containing for which the value of equals "fred" are selected in the filter output. fred fred admin Fred Flintstone Enns Standards Track [Page 26] RFC 4741 NETCONF Protocol December 2006 2 2 6.4.6. Specific Elements from a Specific Entry This filter contains two containment nodes (, ), one content match node (), and two selector nodes (, ). All instances of the and elements in the same sibling set containing for which the value of equals "fred" are selected in the filter output. The element is not included because the sibling set contains selection nodes. fred fred admin Enns Standards Track [Page 27] RFC 4741 NETCONF Protocol December 2006 Fred Flintstone 6.4.7. Multiple Subtrees This filter contains three subtrees (name=root, fred, barney). The "root" subtree filter contains two containment nodes (, ), one content match node (), and one selector node (). The subtree selection criteria is met, and just the company-info subtree for "root" is selected in the filter output. The "fred" subtree filter contains three containment nodes (, , ), one content match node (), and one selector node (). The subtree selection criteria is met, and just the element within the company-info subtree for "fred" is selected in the filter output. The "barney" subtree filter contains three containment nodes (, , ), two content match nodes (, ), and one selector node (). The subtree selection criteria is not met because user "barney" is not a "superuser", and the entire subtree for "barney" (including its parent entry) is excluded from the filter output. root fred Enns Standards Track [Page 28] RFC 4741 NETCONF Protocol December 2006 barney superuser root 1 1 fred 2 6.4.8. Elements with Attribute Naming In this example, the filter contains one containment node (), one attribute match expression (ifName), and one selector node (). All instances of the subtree that have an ifName attribute equal to "eth0" are selected in the filter output. The filter data elements and attributes must be qualified because the ifName attribute will not be considered part of the 'schema/1.2' namespace if it is unqualified. Enns Standards Track [Page 29] RFC 4741 NETCONF Protocol December 2006 45621 774344 If ifName were a child node instead of an attribute, then the following request would produce similar results. eth0 Enns Standards Track [Page 30] RFC 4741 NETCONF Protocol December 2006 7. Protocol Operations The NETCONF protocol provides a small set of low-level operations to manage device configurations and retrieve device state information. The base protocol provides operations to retrieve, configure, copy, and delete configuration datastores. Additional operations are provided, based on the capabilities advertised by the device. The base protocol includes the following protocol operations: o get o get-config o edit-config o copy-config o delete-config o lock o unlock o close-session o kill-session A protocol operation may fail for various reasons, including "operation not supported". An initiator should not assume that any operation will always succeed. The return values in any RPC reply should be checked for error responses. The syntax and XML encoding of the protocol operations are formally defined in the XML schema in Appendix B. The following sections describe the semantics of each protocol operation. 7.1. Description: Retrieve all or part of a specified configuration. Enns Standards Track [Page 31] RFC 4741 NETCONF Protocol December 2006 Parameters: source: Name of the configuration datastore being queried, such as . filter: The filter element identifies the portions of the device configuration to retrieve. If this element is unspecified, the entire configuration is returned. The filter element may optionally contain a "type" attribute. This attribute indicates the type of filtering syntax used within the filter element. The default filtering mechanism in NETCONF is referred to as subtree filtering and is described in Section 6. The value "subtree" explicitly identifies this type of filtering. If the NETCONF peer supports the :xpath capability (Section 8.9), the value "xpath" may be used to indicate that the select attribute on the filter element contains an XPath expression. Positive Response: If the device can satisfy the request, the server sends an element containing a element with the results of the query. Negative Response: An element is included in the if the request cannot be completed for any reason. Example: To retrieve the entire subtree: Enns Standards Track [Page 32] RFC 4741 NETCONF Protocol December 2006 root superuser Charlie Root 1 1 If the configuration is available in multiple formats, such as XML and text, an XML namespace can be used to specify which format is desired. In the following example, the client uses a specific element () in a specific namespace to indicate to the server the desire to receive the configuration in an alternative format. The server may support any number of distinct formats or views into the configuration data, with the client using the parameter to select between them. Enns Standards Track [Page 33] RFC 4741 NETCONF Protocol December 2006 Section 6 contains additional examples of subtree filtering. 7.2. Description: The operation loads all or part of a specified configuration to the specified target configuration. This operation allows the new configuration to be expressed in several ways, such as using a local file, a remote file, or inline. If the target configuration does not exist, it will be created. If a NETCONF peer supports the :url capability (Section 8.8), the element can appear instead of the parameter and should identify a local configuration file. The device analyzes the source and target configurations and performs the requested changes. The target configuration is not necessarily replaced, as with the message. Instead, the target configuration is changed in accordance with the source's data and requested operations. Attributes: operation: Elements in the subtree may contain an "operation" attribute. The attribute identifies the point in the configuration to perform the operation and MAY appear on multiple elements throughout the subtree. If the operation attribute is not specified, the configuration is merged into the configuration datastore. The operation attribute has one of the following values: merge: The configuration data identified by the element containing this attribute is merged with the configuration at the corresponding level in the configuration datastore identified by the target parameter. This is the default behavior. Enns Standards Track [Page 34] RFC 4741 NETCONF Protocol December 2006 replace: The configuration data identified by the element containing this attribute replaces any related configuration in the configuration datastore identified by the target parameter. Unlike a operation, which replaces the entire target configuration, only the configuration actually present in the config parameter is affected. create: The configuration data identified by the element containing this attribute is added to the configuration if and only if the configuration data does not already exist on the device. If the configuration data exists, an element is returned with an value of data-exists. delete: The configuration data identified by the element containing this attribute is deleted in the configuration datastore identified by the target parameter. Parameters: target: Name of the configuration datastore being edited, such as or . default-operation: Selects the default operation (as described in the "operation" attribute) for this request. The default value for the default-operation parameter is "merge". The default-operation parameter is optional, but if provided, it must have one of the following values: merge: The configuration data in the parameter is merged with the configuration at the corresponding level in the target datastore. This is the default behavior. replace: The configuration data in the parameter completely replaces the configuration in the target datastore. This is useful for loading previously saved configuration data. none: The target datastore is unaffected by the configuration in the parameter, unless and until the incoming configuration data uses the "operation" attribute to request a different operation. If the configuration in the parameter contains data for which there is not a Enns Standards Track [Page 35] RFC 4741 NETCONF Protocol December 2006 corresponding level in the target datastore, an is returned with an value of data-missing. Using "none" allows operations like "delete" to avoid unintentionally creating the parent hierarchy of the element to be deleted. test-option: The test-option element may be specified only if the device advertises the :validate capability (Section 8.6). The test-option element has one of the following values: test-then-set: Perform a validation test before attempting to set. If validation errors occur, do not perform the operation. This is the default test-option. set: Perform a set without a validation test first. error-option: The error-option element has one of the following values: stop-on-error: Abort the edit-config operation on first error. This is the default error-option. continue-on-error: Continue to process configuration data on error; error is recorded, and negative response is generated if any errors occur. rollback-on-error: If an error condition occurs such that an error severity element is generated, the server will stop processing the edit-config operation and restore the specified configuration to its complete state at the start of this edit-config operation. This option requires the server to support the :rollback-on-error capability described in Section 8.5. config: A hierarchy of configuration data as defined by one of the device's data models. The contents MUST be placed in an appropriate namespace, to allow the device to detect the appropriate data model, and the contents MUST follow the constraints of that data model, as defined by its capability definition. Capabilities are discussed in Section 8. Enns Standards Track [Page 36] RFC 4741 NETCONF Protocol December 2006 Positive Response: If the device was able to satisfy the request, an is sent containing an element. Negative Response: An response is sent if the request cannot be completed for any reason. Example: The examples in this section utilize a simple data model, in which multiple instances of the 'interface' element may be present, and an instance is distinguished by the 'name' element within each 'interface' element. Set the MTU to 1500 on an interface named "Ethernet0/0" in the running configuration: Ethernet0/0 1500 Add an interface named "Ethernet0/0" to the running configuration, replacing any previous interface with that name: Enns Standards Track [Page 37] RFC 4741 NETCONF Protocol December 2006 Ethernet0/0 1500
192.0.2.4 24
Delete the configuration for an interface named "Ethernet0/0" from the running configuration: none Ethernet0/0 Enns Standards Track [Page 38] RFC 4741 NETCONF Protocol December 2006 Delete interface 192.0.2.4 from an OSPF area (other interfaces configured in the same area are unaffected): none 0.0.0.0 192.0.2.4 7.3. Description: Create or replace an entire configuration datastore with the contents of another complete configuration datastore. If the target datastore exists, it is overwritten. Otherwise, a new one is created, if allowed. If a NETCONF peer supports the :url capability (Section 8.8), the element can appear as the or parameter. Even if it advertises the :writable-running capability, a device may choose not to support the configuration datastore Enns Standards Track [Page 39] RFC 4741 NETCONF Protocol December 2006 as the parameter of a operation. A device may choose not to support remote-to-remote copy operations, where both the and parameters use the element. If the source and target parameters identify the same URL or configuration datastore, an error MUST be returned with an error- tag containing invalid-value. Parameters: target: Name of the configuration datastore to use as the destination of the copy operation. source: Name of the configuration datastore to use as the source of the copy operation or the element containing the configuration subtree to copy. Positive Response: If the device was able to satisfy the request, an is sent that includes an element. Negative Response: An element is included within the if the request cannot be completed for any reason. Example: https://user@example.com:passphrase/cfg/new.txt Enns Standards Track [Page 40] RFC 4741 NETCONF Protocol December 2006 7.4. Description: Delete a configuration datastore. The configuration datastore cannot be deleted. If a NETCONF peer supports the :url capability (Section 8.8), the element can appear as the parameter. Parameters: target: Name of the configuration datastore to delete. Positive Response: If the device was able to satisfy the request, an is sent that includes an element. Negative Response: An element is included within the if the request cannot be completed for any reason. Example: Enns Standards Track [Page 41] RFC 4741 NETCONF Protocol December 2006 7.5. Description: The lock operation allows the client to lock the configuration system of a device. Such locks are intended to be short-lived and allow a client to make a change without fear of interaction with other NETCONF clients, non-NETCONF clients (e.g., SNMP and command line interface (CLI) scripts), and human users. An attempt to lock the configuration MUST fail if an existing session or other entity holds a lock on any portion of the lock target. When the lock is acquired, the server MUST prevent any changes to the locked resource other than those requested by this session. SNMP and CLI requests to modify the resource MUST fail with an appropriate error. The duration of the lock is defined as beginning when the lock is acquired and lasting until either the lock is released or the NETCONF session closes. The session closure may be explicitly performed by the client, or implicitly performed by the server based on criteria such as failure of the underlying transport, or simple inactivity timeout. This criteria is dependent on the implementation and the underlying transport. The lock operation takes a mandatory parameter, target. The target parameter names the configuration that will be locked. When a lock is active, using the operation on the locked configuration and using the locked configuration as a target of the operation will be disallowed by any other NETCONF session. Additionally, the system will ensure that these locked configuration resources will not be modified by other non-NETCONF management operations such as SNMP and CLI. The message (at the RPC layer) can be used to force the release of a lock owned by another NETCONF session. It is beyond the scope of this document to define how to break locks held by other entities. A lock MUST not be granted if either of the following conditions is true: * A lock is already held by any NETCONF session or another entity. Enns Standards Track [Page 42] RFC 4741 NETCONF Protocol December 2006 * The target configuration is , it has already been modified, and these changes have not been committed or rolled back. The server MUST respond with either an element or an . A lock will be released by the system if the session holding the lock is terminated for any reason. Parameters: target: Name of the configuration datastore to lock. Positive Response: If the device was able to satisfy the request, an is sent that contains an element. Negative Response: An element is included in the if the request cannot be completed for any reason. If the lock is already held, the element will be 'lock-denied' and the element will include the of the lock owner. If the lock is held by a non- NETCONF entity, a of 0 (zero) is included. Note that any other entity performing a lock on even a partial piece of a target will prevent a NETCONF lock (which is global) from being obtained on that target. Example: The following example shows a successful acquisition of a lock. Enns Standards Track [Page 43] RFC 4741 NETCONF Protocol December 2006 Example: The following example shows a failed attempt to acquire a lock when the lock is already in use. protocol lock-denied error Lock failed, lock is already held 454 7.6. Description: The unlock operation is used to release a configuration lock, previously obtained with the operation. An unlock operation will not succeed if any of the following conditions are true: * the specified lock is not currently active Enns Standards Track [Page 44] RFC 4741 NETCONF Protocol December 2006 * the session issuing the operation is not the same session that obtained the lock The server MUST respond with either an element or an . Parameters: target: Name of the configuration datastore to unlock. A NETCONF client is not permitted to unlock a configuration datastore that it did not lock. Positive Response: If the device was able to satisfy the request, an is sent that contains an element. Negative Response: An element is included in the if the request cannot be completed for any reason. Example: 7.7. Description: Retrieve running configuration and device state information. Enns Standards Track [Page 45] RFC 4741 NETCONF Protocol December 2006 Parameters: filter: This parameter specifies the portion of the system configuration and state data to retrieve. If this parameter is empty, all the device configuration and state information is returned. The filter element may optionally contain a 'type' attribute. This attribute indicates the type of filtering syntax used within the filter element. The default filtering mechanism in NETCONF is referred to as subtree filtering and is described in Section 6. The value 'subtree' explicitly identifies this type of filtering. If the NETCONF peer supports the :xpath capability (Section 8.9), the value "xpath" may be used to indicate that the select attribute of the filter element contains an XPath expression. Positive Response: If the device was able to satisfy the request, an is sent. The section contains the appropriate subset. Negative Response: An element is included in the if the request cannot be completed for any reason. Example: eth0 Enns Standards Track [Page 46] RFC 4741 NETCONF Protocol December 2006 eth0 45621 774344 7.8. Description: Request graceful termination of a NETCONF session. When a NETCONF server receives a request, it will gracefully close the session. The server will release any locks and resources associated with the session and gracefully close any associated connections. Any NETCONF requests received after a request will be ignored. Positive Response: If the device was able to satisfy the request, an is sent that includes an element. Negative Response: An element is included in the if the request cannot be completed for any reason. Example: Enns Standards Track [Page 47] RFC 4741 NETCONF Protocol December 2006 7.9. Description: Force the termination of a NETCONF session. When a NETCONF entity receives a request for an open session, it will abort any operations currently in process, release any locks and resources associated with the session, and close any associated connections. If a NETCONF server receives a request while processing a confirmed commit (Section 8.4), it must restore the configuration to its state before the confirmed commit was issued. Otherwise, the operation does not roll back configuration or other device state modifications made by the entity holding the lock. Parameters: session-id: Session identifier of the NETCONF session to be terminated. If this value is equal to the current session ID, an 'invalid-value' error is returned. Positive Response: If the device was able to satisfy the request, an is sent that includes an element. Negative Response: An element is included in the if the request cannot be completed for any reason. Example: 4 Enns Standards Track [Page 48] RFC 4741 NETCONF Protocol December 2006 8. Capabilities This section defines a set of capabilities that a client or a server MAY implement. Each peer advertises its capabilities by sending them during an initial capabilities exchange. Each peer needs to understand only those capabilities that it might use and MUST ignore any capability received from the other peer that it does not require or does not understand. Additional capabilities can be defined using the template in Appendix C. Future capability definitions may be published as standards by standards bodies or published as proprietary extensions. A NETCONF capability is identified with a URI. The base capabilities are defined using URNs following the method described in RFC 3553 [6]. Capabilities defined in this document have the following format: urn:ietf:params:netconf:capability:{name}:1.0 where {name} is the name of the capability. Capabilities are often referenced in discussions and email using the shorthand :{name}. For example, the foo capability would have the formal name "urn:ietf:params:netconf:capability:foo:1.0" and be called ":foo". The shorthand form MUST NOT be used inside the protocol. 8.1. Capabilities Exchange Capabilities are advertised in messages sent by each peer during session establishment. When the NETCONF session is opened, each peer (both client and server) MUST send a element containing a list of that peer's capabilities. Each peer MUST send at least the base NETCONF capability, "urn:ietf:params:netconf:base:1.0". A server sending the element MUST include a element containing the session ID for this NETCONF session. A client sending the element MUST NOT include a element. A server receiving a element MUST NOT continue the NETCONF session. Similarly, a client that does not receive a element in the server's message MUST NOT continue the NETCONF session. In both cases, the underlying transport should be closed. Enns Standards Track [Page 49] RFC 4741 NETCONF Protocol December 2006 In the following example, a server advertises the base NETCONF capability, one NETCONF capability defined in the base NETCONF document, and one implementation-specific capability. urn:ietf:params:netconf:base:1.0 urn:ietf:params:netconf:capability:startup:1.0 http://example.net/router/2.3/myfeature 4 Each peer sends its element simultaneously as soon as the connection is open. A peer MUST NOT wait to receive the capability set from the other side before sending its own set. 8.2. Writable-Running Capability 8.2.1. Description The :writable-running capability indicates that the device supports direct writes to the configuration datastore. In other words, the device supports edit-config and copy-config operations where the configuration is the target. 8.2.2. Dependencies None. 8.2.3. Capability Identifier The :writable-running capability is identified by the following capability string: urn:ietf:params:netconf:capability:writable-running:1.0 Enns Standards Track [Page 50] RFC 4741 NETCONF Protocol December 2006 8.2.4. New Operations None. 8.2.5. Modifications to Existing Operations 8.2.5.1. The :writable-running capability modifies the operation to accept the element as a . 8.2.5.2. The :writable-running capability modifies the operation to accept the element as a . 8.3. Candidate Configuration Capability 8.3.1. Description The candidate configuration capability, :candidate, indicates that the device supports a candidate configuration datastore, which is used to hold configuration data that can be manipulated without impacting the device's current configuration. The candidate configuration is a full configuration data set that serves as a work place for creating and manipulating configuration data. Additions, deletions, and changes may be made to this data to construct the desired configuration data. A operation may be performed at any time that causes the device's running configuration to be set to the value of the candidate configuration. The operation effectively sets the running configuration to the current contents of the candidate configuration. While it could be modeled as a simple copy, it is done as a distinct operation for a number of reasons. In keeping high-level concepts as first class operations, we allow developers to see more clearly both what the client is requesting and what the server must perform. This keeps the intentions more obvious, the special cases less complex, and the interactions between operations more straightforward. For example, the :confirmed-commit capability (Section 8.4) would make no sense as a "copy confirmed" operation. The candidate configuration may be shared among multiple sessions. Unless a client has specific information that the candidate configuration is not shared, it must assume that other sessions may be able to modify the candidate configuration at the same time. It is therefore prudent for a client to lock the candidate configuration before modifying it. Enns Standards Track [Page 51] RFC 4741 NETCONF Protocol December 2006 The client can discard any uncommitted changes to the candidate configuration by executing the operation. This operation reverts the contents of the candidate configuration to the contents of the running configuration. 8.3.2. Dependencies None. 8.3.3. Capability Identifier The :candidate capability is identified by the following capability string: urn:ietf:params:netconf:capability:candidate:1.0 8.3.4. New Operations 8.3.4.1. Description: When a candidate configuration's content is complete, the configuration data can be committed, publishing the data set to the rest of the device and requesting the device to conform to the behavior described in the new configuration. To commit the candidate configuration as the device's new current configuration, use the operation. The operation instructs the device to implement the configuration data contained in the candidate configuration. If the device is unable to commit all of the changes in the candidate configuration datastore, then the running configuration MUST remain unchanged. If the device does succeed in committing, the running configuration MUST be updated with the contents of the candidate configuration. If the system does not have the :candidate capability, the operation is not available. Positive Response: If the device was able to satisfy the request, an is sent that contains an element. Enns Standards Track [Page 52] RFC 4741 NETCONF Protocol December 2006 Negative Response: An element is included in the if the request cannot be completed for any reason. Example: 8.3.4.2. If the client decides that the candidate configuration should not be committed, the operation can be used to revert the candidate configuration to the current running configuration. This operation discards any uncommitted changes by resetting the candidate configuration with the content of the running configuration. 8.3.5. Modifications to Existing Operations 8.3.5.1. , , , and The candidate configuration can be used as a source or target of any , , , or operation as a or parameter. The element is used to indicate the candidate configuration: Enns Standards Track [Page 53] RFC 4741 NETCONF Protocol December 2006 8.3.5.2. and The candidate configuration can be locked using the operation with the element as the parameter: Similarly, the candidate configuration is unlocked using the element as the parameter: When a client fails with outstanding changes to the candidate configuration, recovery can be difficult. To facilitate easy recovery, any outstanding changes are discarded when the lock is released, whether explicitly with the operation or implicitly from session failure. Enns Standards Track [Page 54] RFC 4741 NETCONF Protocol December 2006 8.4. Confirmed Commit Capability 8.4.1. Description The :confirmed-commit capability indicates that the server will support the and parameters for the protocol operation. See Section 8.3 for further details on the operation. A confirmed commit operation MUST be reverted if a follow-up commit (called the "confirming commit") is not issued within 600 seconds (10 minutes). The timeout period can be adjusted with the element. The confirming commit can itself include a parameter. If the session issuing the confirmed commit is terminated for any reason before the confirm timeout expires, the server MUST restore the configuration to its state before the confirmed commit was issued. If the device reboots for any reason before the confirm timeout expires, the server MUST restore the configuration to its state before the confirmed commit was issued. If a confirming commit is not issued, the device will revert its configuration to the state prior to the issuance of the confirmed commit. Note that any commit operation, including a commit which introduces additional changes to the configuration, will serve as a confirming commit. Thus to cancel a confirmed commit and revert changes without waiting for the confirm timeout to expire, the manager can explicitly restore the configuration to its state before the confirmed commit was issued. For shared configurations, this feature can cause other configuration changes (for example, via other NETCONF sessions) to be inadvertently altered or removed, unless the configuration locking feature is used (in other words, the lock is obtained before the edit-config operation is started). Therefore, it is strongly suggested that in order to use this feature with shared configuration databases, configuration locking should also be used. 8.4.2. Dependencies The :confirmed-commit capability is only relevant if the :candidate capability is also supported. Enns Standards Track [Page 55] RFC 4741 NETCONF Protocol December 2006 8.4.3. Capability Identifier The :confirmed-commit capability is identified by the following capability string: urn:ietf:params:netconf:capability:confirmed-commit:1.0 8.4.4. New Operations None. 8.4.5. Modifications to Existing Operations 8.4.5.1. The :confirmed-commit capability allows 2 additional parameters to the operation. Parameters: confirmed: Perform a confirmed commit operation. confirm-timeout: Timeout period for confirmed commit, in seconds. If unspecified, the confirm timeout defaults to 600 seconds. Example: 120 Enns Standards Track [Page 56] RFC 4741 NETCONF Protocol December 2006 8.5. Rollback on Error Capability 8.5.1. Description This capability indicates that the server will support the 'rollback-on-error' value in the parameter to the operation. For shared configurations, this feature can cause other configuration changes (for example, via other NETCONF sessions) to be inadvertently altered or removed, unless the configuration locking feature is used (in other words, the lock is obtained before the edit-config operation is started). Therefore, it is strongly suggested that in order to use this feature with shared configuration databases, configuration locking also be used. 8.5.2. Dependencies None 8.5.3. Capability Identifier The :rollback-on-error capability is identified by the following capability string: urn:ietf:params:netconf:capability:rollback-on-error:1.0 8.5.4. New Operations None. 8.5.5. Modifications to Existing Operations 8.5.5.1. The :rollback-on-error capability allows the 'rollback-on-error' value to the parameter on the operation. rollback-on-error Enns Standards Track [Page 57] RFC 4741 NETCONF Protocol December 2006 Ethernet0/0 100000 8.6. Validate Capability 8.6.1. Description Validation consists of checking a candidate configuration for syntactical and semantic errors before applying the configuration to the device. If this capability is advertised, the device supports the protocol operation and checks at least for syntax errors. In addition, this capability supports the test-option parameter to the operation and, when it is provided, checks at least for syntax errors. 8.6.2. Dependencies None. 8.6.3. Capability Identifier The :validate capability is identified by the following capability string: urn:ietf:params:netconf:capability:validate:1.0 8.6.4. New Operations 8.6.4.1. Description: This protocol operation validates the contents of the specified configuration. Enns Standards Track [Page 58] RFC 4741 NETCONF Protocol December 2006 Parameters: source: Name of the configuration datastore being validated, such as or the element containing the configuration subtree to validate. Positive Response: If the device was able to satisfy the request, an is sent that contains an element. Negative Response: An element is included in the if the request cannot be completed for any reason. A validate operation can fail for any of the following reasons: + Syntax errors + Missing parameters + References to undefined configuration data Example: Enns Standards Track [Page 59] RFC 4741 NETCONF Protocol December 2006 8.7. Distinct Startup Capability 8.7.1. Description The device supports separate running and startup configuration datastores. Operations that affect the running configuration will not be automatically copied to the startup configuration. An explicit operation from the to the must be invoked to update the startup configuration to the current contents of the running configuration. NETCONF protocol operations refer to the startup datastore using the element. 8.7.2. Dependencies None. 8.7.3. Capability Identifier The :startup capability is identified by the following capability string: urn:ietf:params:netconf:capability:startup:1.0 8.7.4. New Operations None. 8.7.5. Modifications to Existing Operations 8.7.5.1. General The :startup capability adds the configuration datastore to arguments of several NETCONF operations. The server MUST support the following additional values: Enns Standards Track [Page 60] RFC 4741 NETCONF Protocol December 2006 +--------------------+--------------------------+-------------------+ | Operation | Parameters | Notes | +--------------------+--------------------------+-------------------+ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If :validate is | | | | advertised | +--------------------+--------------------------+-------------------+ To save the startup configuration, use the copy-config operation to copy the configuration datastore to the configuration datastore. 8.8. URL Capability 8.8.1. Description The NETCONF peer has the ability to accept the element in and parameters. The capability is further identified by URL arguments indicating the URL schemes supported. 8.8.2. Dependencies None. Enns Standards Track [Page 61] RFC 4741 NETCONF Protocol December 2006 8.8.3. Capability Identifier The :url capability is identified by the following capability string: urn:ietf:params:netconf:capability:url:1.0?scheme={name,...} The :url capability URI MUST contain a "scheme" argument assigned a comma-separated list of scheme names indicating which schemes the NETCONF peer supports. For example: urn:ietf:params:netconf:capability:url:1.0?scheme=http,ftp,file 8.8.4. New Operations None. 8.8.5. Modifications to Existing Operations 8.8.5.1. The :url capability modifies the operation to accept the element as an alternative to the parameter. If the element is specified, then it should identify a local configuration file. 8.8.5.2. The :url capability modifies the operation to accept the element as the value of the and the parameters. 8.8.5.3. The :url capability modifies the operation to accept the element as the value of the parameters. If this parameter contains a URL, then it should identify a local configuration file. 8.8.5.4. The :url capability modifies the operation to accept the element as the value of the parameter. Enns Standards Track [Page 62] RFC 4741 NETCONF Protocol December 2006 8.9. XPath Capability 8.9.1. Description The XPath capability indicates that the NETCONF peer supports the use of XPath expressions in the element. XPath is described in [2]. The XPath expression must return a node-set. The XPath expression is evaluated in a context where the context node is the root node, and the set of namespace declarations are those in scope on the filter element, including the default namespace. 8.9.2. Dependencies None. 8.9.3. Capability Identifier The :xpath capability is identified by the following capability string: urn:ietf:params:netconf:capability:xpath:1.0 8.9.4. New Operations None. 8.9.5. Modifications to Existing Operations 8.9.5.1. and The :xpath capability modifies the and operations to accept the value "xpath" in the type attribute of the filter element. When the type attribute is set to "xpath", a select attribute MUST be present on the filter element. The select attribute will be treated as an XPath expression and used to filter the returned data. The filter element itself MUST be empty in this case. For example: Enns Standards Track [Page 63] RFC 4741 NETCONF Protocol December 2006 9. Security Considerations This document does not specify an authorization scheme, as such a scheme should be tied to a meta-data model or a data model. Implementors SHOULD provide a comprehensive authorization scheme with NETCONF. Authorization of individual users via the NETCONF server may or may not map 1:1 to other interfaces. First, the data models may be incompatible. Second, it may be desirable to authorize based on mechanisms available in the transport protocol layer (TELNET, SSH, etc). In addition, operations on configurations may have unintended consequences if those operations are also not guarded by the global lock on the files or objects being operated upon. For instance, a partially complete access list could be committed from a candidate configuration unbeknownst to the owner of the lock of the candidate configuration, leading to either an insecure or inaccessible device if the lock on the candidate configuration does not also apply to the operation when applied to it. Configuration information is by its very nature sensitive. Its transmission in the clear and without integrity checking leaves devices open to classic eavesdropping attacks. Configuration information often contains passwords, user names, service descriptions, and topological information, all of which are sensitive. Because of this, this protocol should be implemented carefully with adequate attention to all manner of attack one might expect to experience with other management interfaces. The protocol, therefore, must minimally support options for both confidentiality and authentication. It is anticipated that the underlying protocol (SSH, BEEP, etc) will provide for both confidentiality and authentication, as is required. It is further expected that the identity of each end of a NETCONF session will be available to the other in order to determine authorization for any given request. One could also easily envision additional information, such as transport and encryption methods, being made available for purposes of authorization. NETCONF itself provide no means to re-authenticate, much less authenticate. All such actions occur at lower layers. Enns Standards Track [Page 64] RFC 4741 NETCONF Protocol December 2006 Different environments may well allow different rights prior to and then after authentication. Thus, an authorization model is not specified in this document. When an operation is not properly authorized, a simple "access denied" is sufficient. Note that authorization information may be exchanged in the form of configuration information, which is all the more reason to ensure the security of the connection. That having been said, it is important to recognize that some operations are clearly more sensitive by nature than others. For instance, to the startup or running configurations is clearly not a normal provisioning operation, whereas is. Such global operations MUST disallow the changing of information that an individual does not have authorization to perform. For example, if a user A is not allowed to configure an IP address on an interface but user B has configured an IP address on an interface in the configuration, user A must not be allowed to commit the configuration. Similarly, just because someone says "go write a configuration through the URL capability at a particular place", this does not mean that an element should do it without proper authorization. The operation will demonstrate that NETCONF is intended for use by systems that have at least some trust of the administrator. As specified in this document, it is possible to lock portions of a configuration that a principal might not otherwise have access to. After all, the entire configuration is locked. To mitigate this problem, there are two approaches. It is possible to kill another NETCONF session programmatically from within NETCONF if one knows the session identifier of the offending session. The other possible way to break a lock is to provide an function within the device's native user interface. These two mechanisms suffer from a race condition that may be ameliorated by removing the offending user from an AAA server. However, such a solution is not useful in all deployment scenarios, such as those where SSH public/private key pairs are used. Enns Standards Track [Page 65] RFC 4741 NETCONF Protocol December 2006 10. IANA Considerations 10.1. NETCONF XML Namespace This document registers a URI for the NETCONF XML namespace in the IETF XML registry [7]. Following the format in RFC 3688, IANA has made the following registration. URI: urn:ietf:params:xml:ns:netconf:base:1.0 Registrant Contact: The IESG. XML: N/A, the requested URI is an XML namespace. 10.2. NETCONF XML Schema This document registers a URI for the NETCONF XML schema in the IETF XML registry [7]. Following the format in RFC 3688, IANA has made the following registration. URI: urn:ietf:params:xml:schema:netconf Registrant Contact: The IESG. XML: Appendix B of this document. 10.3. NETCONF Capability URNs This document creates a registry that allocates NETCONF capability identifiers. Additions to the registry require IETF Standards Action. The initial content of the registry contains the capability URNs defined in Section 8. Following the guidelines in RFC 3553 [6], IANA assigned a NETCONF sub-namespace as follows: Registry name: netconf Specification: Section 8 of this document. Repository: The following table. Enns Standards Track [Page 66] RFC 4741 NETCONF Protocol December 2006 +--------------------+----------------------------------------------+ | Index | Capability Identifier | +--------------------+----------------------------------------------+ | :writable-running | urn:ietf:params:netconf:capability:writable- | | | running:1.0 | | | | | :candidate | urn:ietf:params:netconf:capability:candidate | | | :1.0 | | | | | :confirmed-commit | urn:ietf:params:netconf:capability:confirmed | | | -commit:1.0 | | | | | :rollback-on-error | urn:ietf:params:netconf:capability:rollback- | | | on-error:1.0 | | | | | :validate | urn:ietf:params:netconf:capability:validate: | | | 1.0 | | | | | :startup | urn:ietf:params:netconf:capability:startup:1 | | | .0 | | | | | :url | urn:ietf:params:netconf:capability:url:1.0 | | | | | :xpath | urn:ietf:params:netconf:capability:xpath:1.0 | +--------------------+----------------------------------------------+ Index value: The capability name. Enns Standards Track [Page 67] RFC 4741 NETCONF Protocol December 2006 11. Authors and Acknowledgements This document was written by: Andy Bierman Ken Crozier, Cisco Systems Rob Enns, Juniper Networks Ted Goddard, IceSoft Eliot Lear, Cisco Systems Phil Shafer, Juniper Networks Steve Waldbusser Margaret Wasserman, ThingMagic The authors would like to acknowledge the members of the NETCONF working group. In particular, we would like to thank Wes Hardaker for his persistance and patience in assisting us with security considerations. We would also like to thank Randy Presuhn, Sharon Chisholm, Juergen Schoenwalder, Glenn Waters, David Perkins, Weijing Chen, Simon Leinen, Keith Allen, and Dave Harrington for all of their valuable advice. 12. References 12.1. Normative References [1] Sperberg-McQueen, C., Paoli, J., Maler, E., and T. Bray, "Extensible Markup Language (XML) 1.0 (Second Edition)", World Wide Web Consortium, http://www.w3.org/TR/2000/REC-xml-20001006, October 2000. [2] Clark, J. and S. DeRose, "XML Path Language (XPath) Version 1.0", World Wide Web Consortium Recommendation, http://www.w3.org/TR/1999/REC-xpath-19991116, November 1999. [3] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [4] Wasserman, M. and T. Goddard, "Using the NETCONF Configuration Protocol over Secure SHell (SSH)", RFC 4742, December 2006. Enns Standards Track [Page 68] RFC 4741 NETCONF Protocol December 2006 [5] Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform Resource Identifier (URI): Generic Syntax", STD 66, RFC 3986, January 2005. [6] Mealling, M., Masinter, L., Hardie, T., and G. Klyne, "An IETF URN Sub-namespace for Registered Protocol Parameters", BCP 73, RFC 3553, June 2003. [7] Mealling, M., "The IETF XML Registry", BCP 81, RFC 3688, January 2004. 12.2. Informative References [8] Clark, J., "XSL Transformations (XSLT) Version 1.0", World Wide Web Consortium Recommendation, http://www.w3.org/TR/1999/REC- xslt-19991116, November 1999. [9] Dierks, T. and E. Rescorla, "The Transport Layer Security (TLS) Protocol Version 1.1", RFC 4346, April 2006. [10] Ylonen, T. and C. Lonvick, "The Secure Shell (SSH) Protocol Architecture", RFC 4251, January 2006. [11] Rigney, C., Willens, S., Rubens, A., and W. Simpson, "Remote Authentication Dial In User Service (RADIUS)", RFC 2865, June 2000. [12] Hollenbeck, S., Rose, M., and L. Masinter, "Guidelines for the Use of Extensible Markup Language (XML) within IETF Protocols", BCP 70, RFC 3470, January 2003. Enns Standards Track [Page 69] RFC 4741 NETCONF Protocol December 2006 Appendix A. NETCONF Error List Tag: in-use Error-type: protocol, application Severity: error Error-info: none Description: The request requires a resource that already in use. Tag: invalid-value Error-type: protocol, application Severity: error Error-info: none Description: The request specifies an unacceptable value for one or more parameters. Tag: too-big Error-type: transport, rpc, protocol, application Severity: error Error-info: none Description: The request or response (that would be generated) is too large for the implementation to handle. Tag: missing-attribute Error-type: rpc, protocol, application Severity: error Error-info: : name of the missing attribute : name of the element that should contain the missing attribute Description: An expected attribute is missing. Tag: bad-attribute Error-type: rpc, protocol, application Severity: error Error-info: : name of the attribute w/ bad value : name of the element that contains the attribute with the bad value Description: An attribute value is not correct; e.g., wrong type, out of range, pattern mismatch. Tag: unknown-attribute Error-type: rpc, protocol, application Severity: error Error-info: : name of the unexpected attribute : name of the element that contains the unexpected attribute Description: An unexpected attribute is present. Enns Standards Track [Page 70] RFC 4741 NETCONF Protocol December 2006 Tag: missing-element Error-type: rpc, protocol, application Severity: error Error-info: : name of the missing element Description: An expected element is missing. Tag: bad-element Error-type: rpc, protocol, application Severity: error Error-info: : name of the element w/ bad value Description: An element value is not correct; e.g., wrong type, out of range, pattern mismatch. Tag: unknown-element Error-type: rpc, protocol, application Severity: error Error-info: : name of the unexpected element Description: An unexpected element is present. Tag: unknown-namespace Error-type: rpc, protocol, application Severity: error Error-info: : name of the element that contains the unexpected namespace : name of the unexpected namespace Description: An unexpected namespace is present. Tag: access-denied Error-type: rpc, protocol, application Severity: error Error-info: none Description: Access to the requested RPC, protocol operation, or data model is denied because authorization failed. Tag: lock-denied Error-type: protocol Severity: error Error-info: : session ID of session holding the requested lock, or zero to indicate a non-NETCONF entity holds the lock Description: Access to the requested lock is denied because the lock is currently held by another entity. Enns Standards Track [Page 71] RFC 4741 NETCONF Protocol December 2006 Tag: resource-denied Error-type: transport, rpc, protocol, application Severity: error Error-info: none Description: Request could not be completed because of insufficient resources. Tag: rollback-failed Error-type: protocol, application Severity: error Error-info: none Description: Request to rollback some configuration change (via rollback-on-error or discard-changes operations) was not completed for some reason. Tag: data-exists Error-type: application Severity: error Error-info: none Description: Request could not be completed because the relevant data model content already exists. For example, a 'create' operation was attempted on data that already exists. Tag: data-missing Error-type: application Severity: error Error-info: none Description: Request could not be completed because the relevant data model content does not exist. For example, a 'replace' or 'delete' operation was attempted on data that does not exist. Tag: operation-not-supported Error-type: rpc, protocol, application Severity: error Error-info: none Description: Request could not be completed because the requested operation is not supported by this implementation. Tag: operation-failed Error-type: rpc, protocol, application Severity: error Error-info: none Description: Request could not be completed because the requested operation failed for some reason not covered by any other error condition. Enns Standards Track [Page 72] RFC 4741 NETCONF Protocol December 2006 Tag: partial-operation Error-type: application Severity: error Error-info: : identifies an element in the data model for which the requested operation has been completed for that node and all its child nodes. This element can appear zero or more times in the container. : identifies an element in the data model for which the requested operation has failed for that node and all its child nodes. This element can appear zero or more times in the container. : identifies an element in the data model for which the requested operation was not attempted for that node and all its child nodes. This element can appear zero or more times in the container. Description: Some part of the requested operation failed or was not attempted for some reason. Full cleanup has not been performed (e.g., rollback not supported) by the server. The error-info container is used to identify which portions of the application data model content for which the requested operation has succeeded (), failed (), or not been attempted (). Enns Standards Track [Page 73] RFC 4741 NETCONF Protocol December 2006 Appendix B. XML Schema for NETCONF RPC and Protocol Operations BEGIN This import accesses the xml: attribute groups for the xml:lang as declared on the error-message element. Enns Standards Track [Page 74] RFC 4741 NETCONF Protocol December 2006 Enns Standards Track [Page 75] RFC 4741 NETCONF Protocol December 2006 Enns Standards Track [Page 76] RFC 4741 NETCONF Protocol December 2006 Use of the rollback-on-error value requires the :rollback-on-error capability. Enns Standards Track [Page 77] RFC 4741 NETCONF Protocol December 2006 Use of the xpath value requires the :xpath capability. Enns Standards Track [Page 78] RFC 4741 NETCONF Protocol December 2006 The startup datastore can be used only if the :startup capability is advertised. The candidate datastore can be used only if the :candidate datastore is advertised. Enns Standards Track [Page 79] RFC 4741 NETCONF Protocol December 2006 Use of the url element requires the :url capability. Enns Standards Track [Page 80] RFC 4741 NETCONF Protocol December 2006 Use of the test-option element requires the :validate capability. Use of the url element requires the :url capability. Enns Standards Track [Page 81] RFC 4741 NETCONF Protocol December 2006 Enns Standards Track [Page 82] RFC 4741 NETCONF Protocol December 2006 The validate operation requires the :validate capability. Enns Standards Track [Page 83] RFC 4741 NETCONF Protocol December 2006 The commit operation requires the :candidate capability. Use of the confirmed and confirm-timeout elements requires the :confirmed-commit capability. The discard-changes operation requires the :candidate capability. Enns Standards Track [Page 84] RFC 4741 NETCONF Protocol December 2006 END Enns Standards Track [Page 85] RFC 4741 NETCONF Protocol December 2006 Appendix C. Capability Template C.1. capability-name (template) C.1.1. Overview C.1.2. Dependencies C.1.3. Capability Identifier The {name} capability is identified by the following capability string: {capability uri} C.1.4. New Operations C.1.4.1. C.1.5. Modifications to Existing Operations C.1.5.1. If existing operations are not modified by this capability, this section may be omitted. C.1.6. Interactions with Other Capabilities If this capability does not interact with other capabilities, this section may be omitted. Enns Standards Track [Page 86] RFC 4741 NETCONF Protocol December 2006 Appendix D. Configuring Multiple Devices with NETCONF D.1. Operations on Individual Devices Consider the work involved in performing a configuration update against a single individual device. In making a change to the configuration, the application needs to build trust that its change has been made correctly and that it has not impacted the operation of the device. The application (and the application user) should feel confident that their change has not damaged the network. Protecting each individual device consists of a number of steps: o Acquiring the configuration lock. o Loading the update. o Validating the incoming configuration. o Checkpointing the running configuration. o Changing the running configuration. o Testing the new configuration. o Making the change permanent (if desired). o Releasing the configuration lock. Let's look at the details of each step. D.1.1. Acquiring the Configuration Lock A lock should be acquired to prevent simultaneous updates from multiple sources. If multiple sources are affecting the device, the application is hampered in both testing of its change to the configuration and in recovery should the update fail. Acquiring a short-lived lock is a simple defense to prevent other parties from introducing unrelated changes. The lock can be acquired using the operation. Enns Standards Track [Page 87] RFC 4741 NETCONF Protocol December 2006 D.1.2. Loading the Update The configuration can be loaded onto the device without impacting the running system. If the :url capability is supported and lists "file" as a supported scheme, incoming changes can be placed in a local file. file://incoming.conf If the :candidate capability is supported, the candidate configuration can be used. If the update fails, the user file can be deleted using the operation, or the candidate configuration can be reverted using the operation. Enns Standards Track [Page 88] RFC 4741 NETCONF Protocol December 2006 D.1.3. Validating the Incoming Configuration Before the incoming configuration is applied, validating it is often useful. Validation allows the application to gain confidence that the change will succeed and simplifies recovery if it does not. If the device supports the :url capability and lists "file" as a supported scheme, use the operation with the parameter set to the proper user file: file://incoming.conf If the device supports the :candidate capability, some validation will be performed as part of loading the incoming configuration into the candidate. For full validation, either pass the parameter during the step given above, or use the operation with the parameter set to . D.1.4. Checkpointing the Running Configuration The running configuration can be saved into a local file as a checkpoint before loading the new configuration. If the update fails, the configuration can be restored by reloading the checkpoint file. The checkpoint file can be created using the operation. file://checkpoint.conf Enns Standards Track [Page 89] RFC 4741 NETCONF Protocol December 2006 To restore the checkpoint file, reverse the source and target parameters. D.1.5. Changing the Running Configuration When the incoming configuration has been safely loaded onto the device and validated, it is ready to impact the running system. If the device supports the :url capability and lists "file" as a supported scheme, use the operation to merge the incoming configuration into the running configuration. file://incoming.conf If the device supports the :candidate capability, use the operation to set the running configuration to the candidate configuration. Use the parameter to allow automatic reversion to the original configuration if connectivity to the device fails. 120 Enns Standards Track [Page 90] RFC 4741 NETCONF Protocol December 2006 D.1.6. Testing the New Configuration Now that the incoming configuration has been integrated into the running configuration, the application needs to gain trust that the change has affected the device in the way intended without affecting it negatively. To gain this confidence, the application can run tests of the operational state of the device. The nature of the test is dependent on the nature of the change and is outside the scope of this document. Such tests may include reachability from the system running the application (using ping), changes in reachability to the rest of the network (by comparing the device's routing table), or inspection of the particular change (looking for operational evidence of the BGP peer that was just added). D.1.7. Making the Change Permanent When the configuration change is in place and the application has sufficient faith in the proper function of this change, the application should make the change permanent. If the device supports the :startup capability, the current configuration can be saved to the startup configuration by using the startup configuration as the target of the operation. If the device supports the :candidate capability and a confirmed commit was requested, the confirming commit must be sent before the timeout expires. Enns Standards Track [Page 91] RFC 4741 NETCONF Protocol December 2006 D.1.8. Releasing the Configuration Lock When the configuration update is complete, the lock must be released, allowing other applications access to the configuration. Use the operation to release the configuration lock. D.2. Operations on Multiple Devices When a configuration change requires updates across a number of devices, care should be taken to provide the required transaction semantics. The NETCONF protocol contains sufficient primitives upon which transaction-oriented operations can be built. Providing complete transactional semantics across multiple devices is prohibitively expensive, but the size and number of windows for failure scenarios can be reduced. There are two classes of multi-device operations. The first class allows the operation to fail on individual devices without requiring all devices to revert to their original state. The operation can be retried at a later time, or its failure simply reported to the user. An example of this class might be adding an NTP server. For this class of operations, failure avoidance and recovery are focused on the individual device. This means recovery of the device, reporting the failure, and perhaps scheduling another attempt. The second class is more interesting, requiring that the operation should complete on all devices or be fully reversed. The network should either be transformed into a new state or be reset to its original state. For example, a change to a VPN may require updates to a number of devices. Another example of this might be adding a class-of-service definition. Leaving the network in a state where only a portion of the devices have been updated with the new definition will lead to future failures when the definition is referenced. To give transactional semantics, the same steps used in single device operations listed above are used, but are performed in parallel across all devices. Configuration locks should be acquired on all Enns Standards Track [Page 92] RFC 4741 NETCONF Protocol December 2006 target devices and kept until all devices are updated and the changes made permanent. Configuration changes should be uploaded and validation performed across all devices. Checkpoints should be made on each device. Then the running configuration can be changed, tested, and made permanent. If any of these steps fail, the previous configurations can be restored on any devices upon which they were changed. After the changes have been completely implemented or completely discarded, the locks on each device can be released. Appendix E. Deferred Features The following features have been deferred until a future revision of this document. o Granular locking of configuration objects. o Named configuration files/datastores. o Support for multiple NETCONF channels. o Asynchronous notifications. o Explicit protocol support for rollback of configuration changes to prior versions. Enns Standards Track [Page 93] RFC 4741 NETCONF Protocol December 2006 Editor's Address Rob Enns Juniper Networks 1194 North Mathilda Ave Sunnyvale, CA 94089 US EMail: rpe@juniper.net Enns Standards Track [Page 94] RFC 4741 NETCONF Protocol December 2006 Full Copyright Statement Copyright (C) The IETF Trust (2006). This document is subject to the rights, licenses and restrictions contained in BCP 78, and except as set forth therein, the authors retain all their rights. This document and the information contained herein are provided on an "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST, AND THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Intellectual Property The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; nor does it represent that it has made any independent effort to identify any such rights. Information on the procedures with respect to rights in RFC documents can be found in BCP 78 and BCP 79. Copies of IPR disclosures made to the IETF Secretariat and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this specification can be obtained from the IETF on-line IPR repository at http://www.ietf.org/ipr. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights that may cover technology that may be required to implement this standard. Please address the information to the IETF at ietf-ipr@ietf.org. Acknowledgement Funding for the RFC Editor function is currently provided by the Internet Society. Enns Standards Track [Page 95]