Internet-Draft | ACaaS | April 2024 |
Boucadair, et al. | Expires 13 October 2024 | [Page] |
This document specifies a YANG service data model for Attachment Circuits (ACs). This model can be used for the provisioning of ACs before or during service provisioning (e.g., Network Slice Service). The document also specifies a service model for managing bearers over which ACs are established.¶
Also, the document specifies a set of reusable groupings. Whether other service models reuse structures defined in the AC models or simply include an AC reference is a design choice of these service models. Utilizing the AC service model to manage ACs over which a service is delivered has the advantage of decoupling service management from upgrading AC components to incorporate recent AC technologies or features.¶
This note is to be removed before publishing as an RFC.¶
Discussion of this document takes place on the Operations and Management Area Working Group Working Group mailing list (opsawg@ietf.org), which is archived at https://mailarchive.ietf.org/arch/browse/opsawg/.¶
Source for this draft and an issue tracker can be found at https://github.com/boucadair/attachment-circuit-model.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 13 October 2024.¶
Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
Connectivity services are provided by networks to customers via dedicated terminating points, such as Service Functions (SFs) [RFC7665], Customer Edges (CEs), peer Autonomous System Border Routers (ASBRs), data centers gateways, or Internet Exchange Points. A connectivity service is basically about ensuring data transfer received from or destined to a given terminating point to or from other terminating points within the same customer/service, an interconnection node, or an ancillary node. The objectives for the connectivity service can be negotiated and agreed upon between the customer and the network provider. To facilitate data transfer within the provider network, it is assumed that the appropriate setup is provisioned over the links that connect customer terminating points and a provider network (usually via a Provider Edge (PE)), allowing successfully data exchanged over these links. The required setup is referred to in this document as Attachment Circuits (ACs), while the underlying link is referred to as "bearers".¶
This document adheres to the definition of an Attachment Circuit as provided in Section 1.2 of [RFC4364], especially:¶
Routers can be attached to each other, or to end systems, in a variety of different ways: PPP connections, ATM Virtual Circuits (VCs), Frame Relay VCs, ethernet interfaces, Virtual Local Area Networks (VLANs) on ethernet interfaces, GRE tunnels, Layer 2 Tunneling Protocol (L2TP) tunnels, IPsec tunnels, etc. We will use the term "attachment circuit" to refer generally to some such means of attaching to a router. An attachment circuit may be the sort of connection that is usually thought of as a "data link", or it may be a tunnel of some sort; what matters is that it be possible for two devices to be network layer peers over the attachment circuit.¶
When a customer requests a new value-added service, the service can be bound to existing attachment circuits or trigger the instantiation of new attachment circuits. The provisioning of a value-added service should, thus, accommodate both deployments.¶
Also, because the instantiation of an attachment circuit requires coordinating the provisioning of endpoints that might not belong to the same administrative entity (customer vs. provider or distinct operational teams within the same provider, etc.), providing programmatic means to expose 'attachment circuits'-as-a-service greatly simplifies the provisioning of value-added services delivered over an attachment circuit. For example, management systems of adjacent domains that need to connect via an AC will use such means to agree upon the resources that are required for the activation of both sides of an AC (e.g., Layer 2 tags, IP address family, or IP subnets).¶
This document specifies a YANG service data model ("ietf-ac-svc") for managing attachment circuits that are exposed by a network to its customers, such as an enterprise site, an SF, a hosting infrastructure, or a peer network provider. The model can be used for the provisioning of ACs prior or during advanced service provisioning (e.g., Network Slice Service [RFC9543]).¶
The "ietf-ac-svc" module (Section 6.2) includes a set of reusable groupings. Whether a service model reuses structures defined in the "ietf-ac-svc" or simply includes an AC reference (that was communicated during AC service instantiation) is a design choice of these service models. Relying upon the AC service model to manage ACs over which services are delivered has the merit of decorrelating the management of the (core) service vs. upgrade the AC components to reflect recent AC technologies or new features (e.g., new encryption scheme, additional routing protocol). This document favors the approach of completely relying upon the AC service model instead of duplicating data nodes into specific modules of advanced services that are delivered over an Attachment Circuit.¶
Since the provisioning of an AC requires a bearer to be in place, this document introduces a new module called "ietf-bearer-svc" that enables customers to manage their bearer requests (Section 6.1). The customers can then retrieve a provider-assigned bearer reference that they will include in their AC service requests. Likewise, a customer may retrieve whether their bearers support a synchronization mechanism such as Sync Ethernet (SyncE) [ITU-T-G.781]. An example of retrieving a bearer reference is provided in Appendix A.1.¶
An AC service request can provide a reference to a bearer or a set of peer Service Attachment Points (SAPs) [RFC9408]. Both schemes are supported in the AC service model. When several bearers are available, the AC service request may filter them based on the bearer type, synchronization support, etc.¶
Each AC is identified with a unique identifier within a (provider) domain. From a network provider standpoint, an AC can be bound to a single or multiple SAPs [RFC9408]. Likewise, the same SAP can be bound to one or multiple ACs. However, the mapping between an AC and a PE in the provider network that terminates that AC is hidden to the application that makes use of the AC service model. Such mapping information is internal to the network controllers. As such, the details about the (node-specific) attachment interfaces are not exposed in the AC service model. However, these details are exposed at the network model per [I-D.ietf-opsawg-ntw-attachment-circuit]. [I-D.ietf-opsawg-ac-lxsm-lxnm-glue] specifies augmentations to the L2VPN Service Model (L2SM) [RFC8466] and the L3VPN Service Model (L3SM) [RFC8299] to bind LxVPN services to ACs.¶
The AC service model does not make any assumptions about the internal structure or even the nature or the services that will be delivered over an attachment circuit or a set of attachment circuits. Customers do not have access to that network view other than the ACs that they ordered. For example, the AC service model can be used to provision a set of ACs to connect multiple sites (Site1, Site2, ..., SiteX) for customer who also requested VPN services. If the provisioning of these services requires specific configuration on ASBR nodes, such configuration is handled at the network level and is not exposed to the customer at the service level. However, the network controller will have access to such a view as the service points in these ASBRs will be exposed as SAPs with "role" set to "ietf-sap-ntw:nni" [RFC9408].¶
The AC service model can be used in a variety of contexts, such as (but not limited to) those provided in Appendix A:¶
Create an AC over an existing bearer Appendix A.2.¶
Request an attachment circuit for a known peer SAP (Appendix A.3).¶
Instantiate multiple attachment circuits over the same bearer (Appendix A.4).¶
Control the precedence over multiple attachment circuits (Appendix A.5).¶
Create Multiple ACs bound to Multiple CEs (Appendix A.6).¶
Bind a slice service to a set of pre-provisioned attachment circuits (Appendix A.7).¶
Connect a Cloud Infrastructure to a service provider network (Appendix A.8).¶
Interconnect provider networks (e.g., [RFC8921] or [I-D.ramseyer-grow-peering-api]). Such ACs are identified with a "role" set to "ac-common:nni" or "ac-common:public-nni". See Appendix A.10 to illustrate the use of the AC model for peering.¶
Manage connectivity for complex containerized or virtualized functions in the cloud (Appendix A.11).¶
The examples provided in Appendix A use the IPv4 address blocks reserved for documentation [RFC5737], the IPv6 prefix reserved for documentation [RFC3849], and the Autonomous System (AS) numbers reserved for documentation [RFC5398].¶
The YANG data models in this document conform to the Network Management Datastore Architecture (NMDA) defined in [RFC8342].¶
The AC model specified in this document is not a network model [RFC8969]. As such, the model does not expose details related to specific nodes in the provider's network that terminate an AC (e.g., network node identifiers). The mapping between an AC as seen by a customer and the network implementation of an AC is maintained by the network controllers and is not exposed to the customer. This mapping can be maintained using a variety of network models, such as augmented SAP AC network model [I-D.ietf-opsawg-ntw-attachment-circuit].¶
The AC service model is not a device model. A network provider may use a variety of device models (e.g., Routing management [RFC8349] or BGP [I-D.ietf-idr-bgp-model]) to provision an AC service in relevant network nodes.¶
The L2VPN Service Model (L2SM) [RFC8466] covers some AC-related considerations. Nevertheless, the L2SM structure is primarily focused on Layer 2 aspects. For example, the L2SM does not cover Layer 3 provisioning, which is required for the typical AC instantiation.¶
Like the L2SM, the L3VPN Service Model (L3SM) [RFC8299] addresses certain AC-related aspects. However, the L3SM structure does not sufficiently address Layer 2 provisioning requirements. Additionally, the L3SM is primarily designed for conventional L3VPN deployments and, as such, has some limitations for instantiating ACs in other deployment contexts (e.g., cloud environments). For example, the L3SM does not provide the capability to provision multiple BGP peer groups over the same AC.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
The meanings of the symbols in the YANG tree diagrams are defined in [RFC8340].¶
LxSM refers to both the L2SM and the L3SM.¶
LxNM refers to both the L2NM and the L3NM.¶
This document uses the following terms:¶
A physical or logical link that connects a customer node (or site) to a provider network. A bearer can be a wireless or wired link. One or multiple technologies can be used to build a bearer. The bearer type can be specified by a customer.¶
The operator allocates a unique bearer reference to identify a bearer within its network (e.g., customer line identifier). Such a reference can be retrieved by a customer and used in subsequent service placement requests to unambiguously identify where a service is to be bound.¶
The concept of bearer can be generalized to refer to the required underlying connection for the provisioning of an attachment circuit. One or multiple attachment circuits may be hosted over the same bearer (e.g., multiple VLANs on the same bearer that is provided by a physical link).¶
Denotes a functional entity responsible for the management of the service provider network.¶
Refers to a functional entity that interacts with the customer of a network service. The service orchestrator is typically responsible for the attachment circuits, the PE selection, and requesting the activation of the requested service to a network controller.¶
A network that is able to provide network services (e.g., Layer 2 VPN, Layer 3 VPN, or Network Slice Services).¶
A service provider that offers network services (e.g., Layer 2 VPN, Layer 3 VPN, or Network Slice Services).¶
Figure 1 depicts the relationship between the various AC data models:¶
"ietf-ac-common" ([I-D.ietf-opsawg-teas-common-ac])¶
"ietf-bearer-svc" (Section 6.2)¶
"ietf-ac-svc" (Section 6.1)¶
"ietf-ac-ntw" ([I-D.ietf-opsawg-ntw-attachment-circuit])¶
"ietf-ac-glue" ([I-D.ietf-opsawg-ac-lxsm-lxnm-glue])¶
"ietf-ac-common" is imported by "ietf-bearer-svc", "ietf-ac-svc", and "ietf-ac-ntw". Bearers managed using "ietf-bearer-svc" may be referenced in the service ACs managed using "ietf-ac-svc". Similarly, a bearer managed using "ietf-bearer-svc" may list the set of ACs that use that bearer. In order to ease correlation between an AC service requests and the actual AC provisioned in the network, "ietf-ac-ntw" uses the AC references exposed by "ietf-ac-svc". To bind Layer 2 VPN or Layer 3 VPN services with ACs, "ietf-ac-glue" augments the LxSM and LxNM with AC service references exposed by "ietf-ac-svc" and AC network references exposed bt "ietf-ac-ntw".¶
Figure 2 depicts two target topology flavors that involve ACs. These topologies have the following characteristics:¶
A CE can be either a physical device or a logical entity. Such logical entity is typically a software component (e.g., a virtual service function that is hosted within the provider's network or a third-party infrastructure). A CE is seen by the network as a peer SAP.¶
An AC service request may include one or multiple ACs, which may be associated to a single CE or multiple CEs.¶
CEs may be either dedicated to one single connectivity service or host multiple connectivity services (e.g., CEs with roles of SFs [RFC7665]).¶
A network provider may bind a single AC to one or multiple peer SAPs (e.g., CE#1 and CE#2 are tagged as peer SAPs for the same AC). For example, and as discussed in [RFC4364], multiple CEs can be attached to a PE over the same attachment circuit. This scenario is typically implemented when the Layer 2 infrastructure between the CE and the network is a multipoint service.¶
A single CE may terminate multiple ACs, which can be associated with the same bearer or distinct bearers.¶
Customers may request protection schemes in which the ACs associated with their endpoints are terminated by the same PE (e.g., CE#3), distinct PEs (e.g., CE#34), etc. The network provider uses this request to decide where to terminate the AC in the network provider network and also whether to enable specific capabilities (e.g., Virtual Router Redundancy Protocol (VRRP) [RFC5798]). Note that placement constraints may also be requested during the instantiation of the underlying bearers (Section 5.1).¶
The procedure to provision a service in a service provider network may depend on the practices adopted by a service provider. This includes the flow put in place for the provisioning of network services and how they are bound to an attachment circuit. For example, a single attachment circuit may be used to host multiple connectivity services. In order to avoid service interference and redundant information in various locations, a service provider may expose an interface to manage ACs network-wide. Customers can then request a bearer or an attachment circuit to be put in place, and then refer to that bearer or AC when requesting services that are bound to the bearer or AC. [I-D.ietf-opsawg-ac-lxsm-lxnm-glue] specifies augmentations to the L2SM and the L3SM to bind LxVPN services to ACs.¶
Figure 3 shows the positioning of the AC service model is the overall service delivery process.¶
In order to ease the mapping between the service model and underlying network models (e.g., the L3VPN Network Model (L3NM), SAP), the name conventions used in existing network data models are reused as much as possible. For example, "local-address" is used rather than "provider-address" (or similar) to refer to an IP address used in the provider network. This approach is consistent with the automation framework defined in [RFC8969].¶
Figure 4 shows the tree for managing the bearers (that is, the properties of an attachment that are below Layer 3). A bearer can be a physical or logical link (e.g., Link Aggregation Group (LAG) [IEEE802.1AX]). Also, a bearer can be a wireless or wired link. A reference to a bearer is generated by the operator. Such a reference can be used, e.g., in a subsequent service request to create an AC. The anchoring of the AC can also be achieved by indicating (with or without a bearer reference), a peer SAP identifier (e.g., an identifier of an SF).¶
In some deployments, a customer may first retrieve a list of available presence locations before actually placing an order for a bearer creation. The request may be filtered based upon a customer name, role of the bearer, etc. The retrieved location name may be then referenced in the bearer creation request ("provider-location-reference").¶
The same customer site (CE, SF, etc.) can terminate one or multiple bearers; each of them uniquely identified by a reference that is assigned by the network provider. These bearers can terminate on the same or distinct network nodes. CEs that terminate multiple bearers are called multi-homed CEs.¶
A bearer can be created, modified, or discovered from the network. For example, the following deployment options can be considered:¶
In this scenario, bearers are created from scratch using specific requests made to a network controller. This method allows providers to tailor bearer creation to meet customer-specific needs. For example, a bearer request may indicate some hints about the placement constraints ('placement-constraints'). These constraints are used by a provider to determine how/where to terminate a bearer in the network side (e.g., Point of Presence (PoP) or PE selection).¶
Devices can use specific protocols (e.g., Link Layer Discovery Protocol (LLDP) [IEEE802.1AB]) to automatically discover and connect to available network resources. A network controller can use such reported information to expose discovered bearers from the network using the same bearer data model structure.¶
A request to create a bearer may include a set of constraints ("placement-constraints") that are used by a controller to decide the network terminating side of a bearer (e.g., PE selection, PE redundancy, or PoP selection). Future placement criteria ("constraint-type") may be defined in the future to accommodate specific deployment contexts.¶
The descriptions of the bearer data nodes are as follows:¶
Used to uniquely identify a bearer. This name is typically selected by the client when requesting a bearer.¶
Indicates the name of the customer who ordered the bearer.¶
Includes a textual description of the bearer.¶
Tags a bearer with one ore more identifiers that are used to group a set of bearers.¶
Includes operational comments that may be useful for managing the bearer (building, level, etc.). No structure is associated with this data node to accommodate all deployments.¶
Specifies the parent bearer. This data node can be used, e.g., if a bearer is a member of a LAG.¶
Lists the bearers that are members of a LAG. Members can be declared as part of a LAG using 'bearer-parent-ref'.¶
Reports whether a synchronization physical (Sync PHY) mechanism is supported for this bearer.¶
Indicates whether a Sync PHY mechanism is enabled for a bearer. Only applies when 'sync-phy-capable' is set to 'true'.¶
Specifies the Sync PHY mechanism (e.g., SynchE [ITU-T-G.781]) enabled for the bearer.¶
Indicates a location identified by a provider-assigned reference.¶
Specifies the customer terminating point for the bearer. A bearer request can indicate a device, a site, a combination thereof, or a custom information when requesting a bearer. All these schemes are supported in the model.¶
Specifies the bearer type (Ethernet, wireless, LAG, etc.).¶
Indicates that a request is only for test and not for setting, even if there are no errors. This is used for feasibility checks. This data node is applicable only when the data model is used with protocols which do not natively support such option. For example, this data node is redundant with the "test-only" value of the <test-option>
parameter in the NETCONF <edit-config>
operation (Section 7.2 of [RFC6241]).¶
Returns an internal reference for the service provider to uniquely identify the bearer. This reference can be used when requesting services. Appendix A.1 provides an example about how this reference can be retrieved by a customer.¶
Whether the 'bearer-reference' mirrors the content of the 'name' is deployment-specific. The module does not assume nor preclude such schemes.¶
Specifies the set of attachment circuits that are bound to the bearer.¶
Specifies the requested date and time when the bearer is expected to be active.¶
Specifies the requested date and time when the bearer is expected to be disabled.¶
Reports the actual date and time when the bearer actually was enabled.¶
Reports the actual date and time when the bearer actually was disabled.¶
Used to track the overall status of a given bearer. Both operational and administrative status are maintained together with a timestamp.¶
The "admin-status" attribute is typically configured by a network operator to indicate whether the service is enabled, disabled, or subjected to additional testing or pre-deployment checks. These additional options, such as 'admin-testing' and 'admin-pre-deployment', provide the operators the flexibility to conduct additional validations on the bearer before deploying services over that connection.¶
The "oper-status" of a bearer reflects its operational state as observed. As a bearer can contain multiple services, the operational status should only reflect the status of the bearer connection. To obtain network-level service status, specific network models such as those in Section 7.3 of [RFC9182] or Section 7.3 of [RFC9291] should be consulted.¶
It is important to note that the "admin-status" attribute should remain independent of the "oper-status". In other words, the setting of the intended administrative state (e.g., whether "admin-up" or "admin-testing") MUST NOT be influenced by the current operational state. If the bearer is administratively set to 'admin-down', it is expected that the bearer will also be operationally 'op-down' as a result of this administrative decision.¶
To assess the service delivery status for a given bearer comprehensively, it is recommended to consider both administrative and operational service status values in conjunction. This holistic approach allows a network controller or operator to identify anomalies effectively.¶
For instance, when a bearer is administratively enabled but the "operational-status" of that bearer is reported as "op-down", it should be expected that the "oper-status" of services transported over that bearer is also down. These status values differing should trigger the detection of an anomaly condition.¶
The full tree diagram of the module can be generated using, e.g., the "pyang" tool [PYANG]. That tree is not included here because it is too long (Section 3.4 of [I-D.ietf-netmod-rfc8407bis]). Instead, subtrees are provided for the reader's convenience. The full tree of the 'ac-svc' is provided in [AC-svc-Tree].¶
The overall tree structure of the AC service module is shown in Figure 5.¶
The rationale for deciding whether a reusable grouping should be maintained in this document or be moved into the AC common module [I-D.ietf-opsawg-teas-common-ac] is as follows:¶
Groupings that are reusable among the AC service module, AC network module, other service models, and network models are included in the AC common module.¶
Groupings that are reusable only by other service models are maintained in the "ietf-ac-svc" module.¶
Each AC is identified with a unique name ('../ac/name') within a domain. The mapping between this AC and a local PE that terminates the AC is hidden to the application that makes use of the AC service model. This information is internal to the Network controller. As such, the details about the (node-specific) attachment interfaces are not exposed in this service model.¶
The AC service model uses groupings and types defined in the AC common model [I-D.ietf-opsawg-teas-common-ac]. Therefore, the description of these nodes are not reiterated in the following subsections.¶
Features are used to tag conditional protions of the model in order to accomodate various deployments (support of layer 2 ACs, Layer 3 ACs, IPv4, IPv6, routing protocols, Bidirectional Forwarding Detection (BFD), etc.).¶
The 'specific-provisioning-profiles' container (Figure 6) can be used by a service provider to maintain a set of reusable profiles. The profiles definitions are similar to those defined in [RFC9181], including: Quality of Service (QoS), BFD, forwarding, and routing profiles. The exact definition of the profiles is local to each service provider. The model only includes an identifier for these profiles in order to facilitate identifying and binding local policies when building an AC.¶
As shown in Figure 6, two profile types can be defined: 'specific-provisioning-profiles' and 'service-provisioning-profiles'. Whether only specific profiles, service profiles, or a combination thereof are used is local to each service provider.¶
The following specific provisioning profiles can be defined:¶
Refers to a set of policies related to the encryption setup that can be applied when provisioning an AC.¶
Refers to a set of policies, such as classification, marking, and actions (e.g., [RFC3644]).¶
Refers to a set of Bidirectional Forwarding Detection (BFD) policies [RFC5880] that can be invoked when building an AC.¶
Refers to the policies that apply to the forwarding of packets conveyed within an AC. Such policies may consist, for example, of applying Access Control Lists (ACLs).¶
Refers to a set of routing policies that will be invoked (e.g., BGP policies) when building an AC.¶
All the abovementioned profiles are uniquely identified by the NETCONF/RESTCONF server by an identifier. To ease referencing these profiles by other data models, specific typedefs are defined for each of these profiles. Likewise, an attachment circuit reference typedef is defined when referencing a (global) attachment circuit by its name is required. These typedefs SHOULD be used when other modules need a reference to one of these profiles or attachment circuits.¶
The 'ac-group-profile' defines reusable parameters for a set of ACs. Each profile is identified by 'name'. Some of the data nodes can be adjusted at the 'ac'. These adjusted values take precedence over the global values. The structure of 'ac-group-profile' is similar to the one used to model each 'ac' (Figure 8).¶
The 'placement-constraints' specifies the placement constraints of an AC. For example, this container can be used to request avoidance of connecting two ACs to the same PE. The full set of supported constraints is defined in [RFC9181] (see 'placement-diversity', in particular).¶
The structure of 'placement-constraints' is shown in Figure 7.¶
The structure of 'attachment-circuits' is shown in Figure 8.¶
The description of the data nodes is as follows:¶
Indicates the name of the customer who ordered the AC or a set of ACs.¶
Includes a textual description of the AC.¶
Indicates that a request is only for test and not for setting, even if there are no errors. This is used for feasibility checks. This data node is applicable only when the data model is used with protocols which do not natively support such option.¶
Specifies the requested date and time when the attachment circuit is expected to be active.¶
Specifies the requested date and time when the attachment circuit is expected to be disabled.¶
Reports the actual date and time when the attachment circuit actually was enabled.¶
Reports the actual date and time when the attachment circuit actually was disabled.¶
Specifies whether an AC is used, e.g., as User-to-Network Interface (UNI) or Network-to-Network Interface (NNI).¶
Includes references to the remote endpoints of an attachment circuit [RFC9408].¶
Indicates references to one or more profiles that are defined in Section 5.2.3.¶
Specifies an AC that is inherited by an attachment circuit.¶
In contexts where dynamic terminating points are managed for a given AC, a parent AC can be defined with a set of stable and common information, while "child" ACs are defined to track dynamic information. These "child" ACs are bound to the parent AC, which is exposed to services (as a stable reference).¶
Whenever a parent AC is deleted, all its "child" ACs MUST be deleted.¶
Lists one or more references of child ACs that rely upon this attachment circuit as a parent AC.¶
Lists the groups to which an AC belongs [RFC9181]. For example, the 'group-id' is used to associate redundancy or protection constraints of ACs. An example is provided in Appendix A.5.¶
Reports the set of services that are bound to the attachment circuit. The services are indexed by their type.¶
Reports the internal reference that is assigned by the provider for this AC. This reference is used to accomodate deployment contexts (e.g., Section 9.1.2 of [RFC8921]) where an identifier is generated by the provider to identify a service order locally.¶
Associates a name that uniquely identifies an AC within a service provider network.¶
References a set of service-specific profiles.¶
See Section 5.2.5.1.¶
See Section 5.2.5.2.¶
See Section 5.2.5.3.¶
See Section 5.2.5.4.¶
See Section 5.2.5.5.¶
See Section 5.2.5.6.¶
The 'l2-connection' container (Figure 9) is used to configure the relevant Layer 2 properties of an AC including: encapsulation details and tunnel terminations. For the encapsulation details, the model supports the definition of the type as well as the Identifiers (e.g., VLAN-IDs) of each of the encapsulation-type defined. For the second case, attributes for pseudowire, Virtual Private LAN Service (VPLS), and Virtual eXtensible Local Area Network (VXLAN) tunnel terminations are included.¶
'bearer-reference' is used to link an AC with a bearer over which the AC is instantiated.¶
This structure relies upon the common groupings defined in [I-D.ietf-opsawg-teas-common-ac].¶
The 'ip-connection' container is used to configure the relevant IP properties of an AC. The model supports the usage of dynamic and static addressing. This structure relies upon the common groupings defined in [I-D.ietf-opsawg-teas-common-ac]. Both IPv4 and IPv6 parameters are supported.¶
Figure 10 shows the structure of the IPv4 connection.¶
Figure 11 shows the structure of the IPv6 connection.¶
As shown in the tree depicted in Figure 12, the 'routing-protocols' container defines the required parameters to enable the desired routing features for an AC. One or more routing protocols can be associated with an AC. Such routing protocols will be then enabled between a PE and the customer terminating points. Each routing instance is uniquely identified by the combination of the 'id' and 'type' to accommodate scenarios where multiple instances of the same routing protocol have to be configured on the same link.¶
In addition to static routing (Section 5.2.5.3.1), the module supports BGP (Section 5.2.5.3.2), OSPF (Section 5.2.5.3.3), IS-IS (Section 5.2.5.3.4), and RIP (Section 5.2.5.3.5). It also includes a reference to the 'routing-profile-identifier' defined in Section 5.2.2, so that additional constraints can be applied to a specific instance of each routing protocol. Moreover, the module supports VRRP (Section 5.2.5.3.6).¶
The static tree structure is shown in Figure 13.¶
As depicted in Figure 13, the following data nodes can be defined for a given IP prefix:¶
Indicates a local tag (e.g., "myfavorite-lan") that is used to enforce local policies.¶
Indicates the next hop to be used for the static route.¶
It can be identified by an IP address, a predefined next-hop type (e.g., 'discard' or 'local-link'), etc.¶
Indicates the metric associated with the static route entry. This metric is used when the route is exported into an IGP.¶
Indicates a BFD profile that applies for this entry.¶
Used to convey the status of a static route entry. This data node can also be used to control the (de)activation of individual static route entries.¶
The BGP tree structure is shown in Figure 14.¶
The following data nodes are supported for each BGP 'peer-group':¶
Defines a name for the peer group.¶
Indicates the provider's AS Number (ASN).¶
Indicates the customer's ASN.¶
Indicates the address family of the peer. It can be set to 'ipv4', 'ipv6', or 'dual-stack'.¶
This address family might be used together with the service type that uses an AC (e.g., 'vpn-type' [RFC9182]) to derive the appropriate Address Family Identifiers (AFIs) / Subsequent Address Family Identifiers (SAFIs) that will be part of the derived device configurations (e.g., unicast IPv4 MPLS L3VPN (AFI,SAFI = 1,128) as defined in Section 4.3.4 of [RFC4364]).¶
Specifies a provider's IP address to use when establishing the BGP transport session.¶
Indicates the maximum number of BGP prefixes allowed in a session for this group.¶
The module adheres to the recommendations in Section 13.2 of [RFC4364], as it allows enabling the TCP Authentication Option (TCP-AO) [RFC5925] and accommodates the installed base that makes use of MD5. In addition, the module includes a provision for using IPsec.¶
Similar to [RFC9182], this version of the ACaaS assumes that parameters specific to the TCP-AO are preconfigured as part of the key chain that is referenced in the ACaaS. No assumption is made about how such a key chain is preconfigured. However, the structure of the key chain should cover data nodes beyond those in [RFC8177], mainly SendID and RecvID (Section 3.1 of [RFC5925]).¶
For each neighbor, the following data nodes are supported in addition to similar parameters that are provided for a peer group:¶
Reports the internal reference that is assigned by the provider for this BGP session.¶
Specifies the customer's IP address used to establishing this BGP session.¶
Specifies the requested date and time when the BGP session is expected to be active.¶
Specifies the requested date and time when the BGP session is expected to be disabled.¶
Reports the actual date and time when the BGP session actually was enabled.¶
Reports the actual date and time when the BGP session actually was disabled.¶
Indicates the status of the BGP routing instance.¶
Specifies a name of a peer group.¶
Parameters that are provided at the 'neighbor' level takes precedence over the ones provided in the peer group.¶
Indicates a BFD profile that applies for a BGP neighbor.¶
The OSPF tree structure is shown in Figure 15.¶
The following OSPF data nodes are supported:¶
Indicates whether IPv4, IPv6, or both address families are to be activated.¶
Indicates the OSPF Area ID.¶
Associates a metric with OSPF routes.¶
Used to create OSPF sham links between two ACs sharing the same area and having a backdoor link (Section 4.2.7 of [RFC4577] and Section 5 of [RFC6565]).¶
Controls the authentication schemes to be enabled for the OSPF instance. The following options are supported: IPsec for OSPFv3 authentication [RFC4552], and the Authentication Trailer for OSPFv2 [RFC5709][RFC7474] and OSPFv3 [RFC7166].¶
Indicates the status of the OSPF routing instance.¶
The IS-IS tree structure is shown in Figure 16.¶
The following IS-IS data nodes are supported:¶
Indicates whether IPv4, IPv6, or both address families are to be activated.¶
Indicates the IS-IS area address.¶
Controls the authentication schemes to be enabled for the IS-IS instance. Both the specification of a key chain [RFC8177] and the direct specification of key and authentication algorithms are supported.¶
Indicates the status of the IS-IS routing instance.¶
The RIP tree structure is shown in Figure 17.¶
'address-family' indicates whether IPv4, IPv6, or both address families are to be activated. For example, this parameter is used to determine whether RIPv2 [RFC2453], RIP Next Generation (RIPng), or both are to be enabled [RFC2080].¶
The model supports the Virtual Router Redundancy Protocol (VRRP) [RFC5798] on an AC (Figure 18).¶
The following data nodes are supported:¶
Indicates whether IPv4, IPv6, or both address families are to be activated. Note that VRRP version 3 [RFC5798] supports both IPv4 and IPv6.¶
Indicates the status of the VRRP instance.¶
Note that no authentication data node is included for VRRP, as there isn't any type of VRRP authentication at this time (see Section 9 of [RFC5798]).¶
As shown in the tree depicted in Figure 19, the 'oam' container defines OAM-related parameters of an AC.¶
This version of the module supports BFD. The following BFD data nodes can be specified:¶
As shown in the tree depicted in Figure 20, the 'security' container defines a set of AC security parameters.¶
The 'security' container specifies the authentication and the encryption to be applied to traffic for a given AC. Tthe model can be used to directly control the encryption to be applied (e.g., Layer 2 or Layer 3 encryption) or invoke a local encryption profile.¶
The structure of the 'service' container is depicted in Figure 21.¶
The 'service' container defines the following data nodes:¶
Specifies the Layer 2 MTU, in bytes, for the AC.¶
Indicates the inbound bandwidth of the AC (i.e., download bandwidth from the service provider to the customer site).¶
Indicates the outbound bandwidth of the AC (i.e., upload bandwidth from the customer site to the service provider).¶
Both 'svc-pe-to-ce-bandwidth' and 'svc-ce-to-pe-bandwidth' can be represented using the Committed Information Rate (CIR), the Excess Information Rate (EIR), or the Peak Information Rate (PIR). Both reuse the 'bandwidth-per-type' grouping defined in [I-D.ietf-opsawg-teas-common-ac].¶
Specifies a list of QoS profiles to apply for this AC.¶
Specifies a list of ACL profiles to apply for this AC.¶
This module uses types defined in [RFC6991] and [RFC9181].¶
<CODE BEGINS> file "ietf-bearer-svc@2023-11-13.yang" module ietf-bearer-svc { yang-version 1.1; namespace "urn:ietf:params:xml:ns:yang:ietf-bearer-svc"; prefix bearer-svc; import ietf-inet-types { prefix inet; reference "RFC 6991: Common YANG Data Types, Section 4"; } import ietf-vpn-common { prefix vpn-common; reference "RFC 9181: A Common YANG Data Model for Layer 2 and Layer 3 VPNs"; } import ietf-ac-common { prefix ac-common; reference "RFC CCCC: A Common YANG Data Model for Attachment Circuits"; } import ietf-ac-svc { prefix ac-svc; reference "RFC XXXX: YANG Data Models for Bearers and 'Attachment Circuits'-as-a-Service (ACaaS)"; } organization "IETF OPSAWG (Operations and Management Area Working Group)"; contact "WG Web: <https://datatracker.ietf.org/wg/opsawg/> WG List: <mailto:opsawg@ietf.org> Editor: Mohamed Boucadair <mailto:mohamed.boucadair@orange.com> Author: Richard Roberts <mailto:rroberts@juniper.net> Author: Oscar Gonzalez de Dios <mailto:oscar.gonzalezdedios@telefonica.com> Author: Samier Barguil <mailto:ssamier.barguil_giraldo@nokia.com> Author: Bo Wu <mailto:lana.wubo@huawei.com>"; description "This YANG module defines a generic YANG model for exposing network bearers as a service. Copyright (c) 2024 IETF Trust and the persons identified as authors of the code. All rights reserved. Redistribution and use in source and binary forms, with or without modification, is permitted pursuant to, and subject to the license terms contained in, the Revised BSD License set forth in Section 4.c of the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info). This version of this YANG module is part of RFC xxx; see the RFC itself for full legal notices."; revision 2023-11-13 { description "Initial revision."; reference "RFC XXXX: YANG Data Models for Bearers and 'Attachment Circuits'-as-a-Service (ACaaS)"; } // Typedef to ease referencing cross-modules typedef bearer-ref { type leafref { path "/bearer-svc:bearers/bearer-svc:bearer/bearer-svc:name"; } description "Defines a type to reference a bearer."; } // Identities identity identification-type { description "Base identity for identification of bearers."; } identity device-id { base identification-type; description "Identification of bearers based on device."; } identity site-id { base identification-type; description "Identification of bearers based on site."; } identity site-and-device-id { base identification-type; description "Identification of bearers based on site and device."; } identity custom { base identification-type; description "Identification of bearers based on other custom criteria."; } identity bearer-type { description "Base identity for bearers type."; } identity ethernet { base bearer-type; description "Ethernet."; } identity wireless { base bearer-type; description "Wireless."; } identity lag { base bearer-type; description "Link Aggregation Group (LAG)."; } identity network-termination-hint { base vpn-common:placement-diversity; description "A hint about the termination at the network side is provided (e.g., geoproximity)."; } identity syncPHY-type { description "Base identity for physical layer synchronization."; } identity syncE { base syncPHY-type; description "Sync Ethernet (SyncE)."; reference "ITU-T G.781: Synchronization layer functions for frequency synchronization based on the physical layer"; } // Reusabel groupings grouping location-information { description "Basic location information"; leaf location-name { type string; description "Provides a location name. This data node can be mapped, e.g., to the 3GPP NRM IOC ManagedElement."; } leaf address { type string; description "Address (number and street) of the device/site."; } leaf postal-code { type string; description "Postal code of the device/site."; } leaf state { type string; description "State of the device/site. This leaf can also be used to describe a region for a country that does not have states."; } leaf city { type string; description "City of the device/site."; } leaf country-code { type string { pattern '[A-Z]{2}'; } description "Country of the device/site. Expressed as ISO ALPHA-2 code."; } } grouping placement-constraints { description "Constraints related to placement of a bearer."; list constraint { if-feature vpn-common:placement-diversity; key "constraint-type"; description "List of constraints."; leaf constraint-type { type identityref { base vpn-common:placement-diversity; } must "not(derived-from-or-self(current(), " + "'vpn-common:bearer-diverse') or " + "derived-from-or-self(current(), " + "'vpn-common:same-bearer'))" { error-message "Only bearer-specific diversity" + "constraints must be provided."; } description "Diversity constraint type for bearers."; } container target { description "The constraint will apply against this list of groups."; choice target-flavor { description "Choice for the group definition."; case id { list group { key "group-id"; description "List of groups."; leaf group-id { type string; description "The constraint will apply against this particular group ID."; } } } case all-bearers { leaf all-other-bearers { type empty; description "The constraint will apply against all other bearers of a site."; } } case all-groups { leaf all-other-groups { type empty; description "The constraint will apply against all other groups managed by the customer."; } } } } } } container locations { description "Retrieves the list of available provider locations for terminating bearers."; leaf customer-name { type string; description "Indicates the name of the customer that requested these bearers."; } leaf role { type identityref { base ac-common:role; } description "Indicates whether this bearer is used as UNI, NNI, etc."; } leaf local-as { type inet:as-number; description "Indicates a provider AS Number (ASN)."; } leaf peer-as { type inet:as-number; description "Indicates the customer's ASN."; } list location { key "location-name"; config false; description "Reports the list of available locations."; uses location-information; } } container bearers { description "Main container for the bearers."; leaf customer-name { type string; description "Indicates the name of the customer that requested these bearers."; } uses ac-common:op-instructions; container placement-constraints { description "Diversity constraint type."; uses placement-constraints; } list bearer { key "name"; description "Maintains a list of bearers."; leaf name { type string; description "A name that uniquely identifies a bearer for a given customer."; } leaf description { type string; description "A description of this bearer."; } leaf customer-name { type string; description "Indicates the name of the customer that requested this bearer."; } uses vpn-common:vpn-components-group; leaf op-comment { type string; description "Includes comments that can be shared with operational teams and which may be useful for the activation of a bearer. This may include, for example, information about the building, level, etc."; } leaf bearer-parent-ref { type bearer-svc:bearer-ref; description "Specifies the parent bearer. This can be used, e.g., for a Link Aggregation Group (LAG)."; } leaf-list bearer-lag-member { type bearer-svc:bearer-ref; config false; description "Reports LAG members."; } leaf sync-phy-capable { type boolean; config false; description "Indicates when set to true that a mechanism for physical layer synchronization is supported for this bearer. No such mechanism is supported if set to false."; } leaf sync-phy-enabled { type boolean; description "Indicates when set to true that a mechanism for physical layer synchronization is enabled for this bearer. No such mechanism is enabled if set to false."; } leaf sync-phy-type { when "../sync-phy-enabled='true'"; type identityref { base syncPHY-type; } description "Type of the physical layer synchronization."; } leaf provider-location-reference { type string; description "Specifies the provider's location reference."; } container customer-point { description "Base container to link the Bearer existence"; leaf identified-by { type identityref { base identification-type; } description "Attribute used to identify the bearer"; } container device { when "derived-from-or-self(../identified-by, " + "'bearer-svc:device-id') or " + "derived-from-or-self(../identified-by, " + "'bearer-svc:site-and-device-id')" { description "Only applicable if identified-by is device."; } description "Bearer is linked to device."; leaf device-id { type string; description "Identifier for the device where that bearer belongs."; } container location { description "Location of the node."; uses location-information; } } container site { when "derived-from-or-self(../identified-by, " + "'bearer-svc:site-id') or " + "derived-from-or-self(../identified-by, " + "'bearer-svc:site-and-device-id')" { description "Only applicable if identified-by is site."; } description "Bearer is linked to a site."; leaf site-id { type string; description "Identifier for the site or sites where that bearer belongs."; } container location { description "Location of the node."; uses location-information; } } leaf custom-id { when "derived-from-or-self(../identified-by, " + "'bearer-svc:custom')" { description "Only enabled id identified-by is custom."; } type string; description "The semantic of this identifier is shared between the customer/provider using out-of-band means."; } } leaf type { type identityref { base bearer-type; } description "Type of the bearer (e.g., Ethernet or wireless)."; } leaf test-only { type empty; description "When present, this indicates that this is a feasibility check request. No resources are commited for such bearer requests."; } leaf bearer-reference { if-feature "ac-common:server-assigned-reference"; type string; config false; description "This is an internal reference for the service provider to identify the bearers."; } leaf-list ac-svc-ref { type ac-svc:attachment-circuit-reference; config false; description "Specifies the set of ACes that are bound to the bearer."; } uses ac-common:op-instructions; uses ac-common:service-status; } } } <CODE ENDS>¶
This module uses types defined in [RFC6991], [RFC9181], [RFC8177], and [I-D.ietf-opsawg-teas-common-ac].¶
<CODE BEGINS> file "ietf-ac-svc@2023-11-13.yang" module ietf-ac-svc { yang-version 1.1; namespace "urn:ietf:params:xml:ns:yang:ietf-ac-svc"; prefix ac-svc; import ietf-ac-common { prefix ac-common; reference "RFC CCCC: A Common YANG Data Model for Attachment Circuits"; } import ietf-vpn-common { prefix vpn-common; reference "RFC 9181: A Common YANG Data Model for Layer 2 and Layer 3 VPNs"; } import ietf-netconf-acm { prefix nacm; reference "RFC 8341: Network Configuration Access Control Model"; } import ietf-inet-types { prefix inet; reference "RFC 6991: Common YANG Data Types, Section 4"; } import ietf-key-chain { prefix key-chain; reference "RFC 8177: YANG Data Model for Key Chains"; } organization "IETF OPSAWG (Operations and Management Area Working Group)"; contact "WG Web: <https://datatracker.ietf.org/wg/opsawg/> WG List: <mailto:opsawg@ietf.org> Editor: Mohamed Boucadair <mailto:mohamed.boucadair@orange.com> Author: Richard Roberts <mailto:rroberts@juniper.net> Author: Oscar Gonzalez de Dios <mailto:oscar.gonzalezdedios@telefonica.com> Author: Samier Barguil <mailto:ssamier.barguil_giraldo@nokia.com> Author: Bo Wu <mailto:lana.wubo@huawei.com>"; description "This YANG module defines a YANG model for exposing attachment circuits as a service (ACaaS). Copyright (c) 2024 IETF Trust and the persons identified as authors of the code. All rights reserved. Redistribution and use in source and binary forms, with or without modification, is permitted pursuant to, and subject to the license terms contained in, the Revised BSD License set forth in Section 4.c of the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info). This version of this YANG module is part of RFC XXXX; see the RFC itself for full legal notices."; revision 2023-11-13 { description "Initial revision."; reference "RFC XXXX: YANG Data Models for Bearers and 'Attachment Circuits'-as-a-Service (ACaaS)"; } /* A set of typedefs to ease referencing cross-modules */ typedef attachment-circuit-reference { type leafref { path "/ac-svc:attachment-circuits/ac-svc:ac/ac-svc:name"; } description "Defines a reference to an attachment circuit that can be used by other modules."; } typedef ac-group-reference { type leafref { path "/ac-svc:attachment-circuits/ac-svc:ac-group-profile" + "/ac-svc:name"; } description "Defines a reference to an attachment circuit profile."; } typedef encryption-profile-reference { type leafref { path "/ac-svc:specific-provisioning-profiles" + "/ac-svc:valid-provider-identifiers" + "/ac-svc:encryption-profile-identifier/ac-svc:id"; } description "Defines a reference to an encryption profile."; } typedef qos-profile-reference { type leafref { path "/ac-svc:specific-provisioning-profiles" + "/ac-svc:valid-provider-identifiers" + "/ac-svc:qos-profile-identifier/ac-svc:id"; } description "Defines a reference to a QoS profile."; } typedef bfd-profile-reference { type leafref { path "/ac-svc:specific-provisioning-profiles" + "/ac-svc:valid-provider-identifiers" + "/ac-svc:bfd-profile-identifier/ac-svc:id"; } description "Defines a reference to a BFD profile."; } typedef forwarding-profile-reference { type leafref { path "/ac-svc:specific-provisioning-profiles" + "/ac-svc:valid-provider-identifiers" + "/ac-svc:forwarding-profile-identifier/ac-svc:id"; } description "Defines a reference to a forwarding profile."; } typedef routing-profile-reference { type leafref { path "/ac-svc:specific-provisioning-profiles" + "/ac-svc:valid-provider-identifiers" + "/ac-svc:routing-profile-identifier/ac-svc:id"; } description "Defines a reference to a routing profile."; } typedef service-profile-reference { type leafref { path "/ac-svc:service-provisioning-profiles" + "/ac-svc:service-profile-identifier" + "/ac-svc:id"; } description "Defines a reference to a service profile."; } /******************** Reusable groupings ********************/ // Basic Layer 2 connection grouping l2-connection-basic { description "Defines Layer 2 protocols and parameters that can be factorized when provisioning Layer 2 connectivity among multiple ACs."; container encapsulation { description "Container for Layer 2 encapsulation."; leaf type { type identityref { base vpn-common:encapsulation-type; } description "Encapsulation type."; } container dot1q { when "derived-from-or-self(../type, 'vpn-common:dot1q')" { description "Only applies when the type of the tagged interface is 'dot1q'."; } description "Tagged interface."; uses ac-common:dot1q; } container qinq { when "derived-from-or-self(../type, 'vpn-common:qinq')" { description "Only applies when the type of the tagged interface is 'qinq'."; } description "Includes QinQ parameters."; uses ac-common:qinq; } } } // Full Layer 2 connection grouping l2-connection { description "Defines Layer 2 protocols and parameters that are used to enable AC connectivity."; container encapsulation { description "Container for Layer 2 encapsulation."; leaf type { type identityref { base vpn-common:encapsulation-type; } description "Indicates the encapsulation type."; } container dot1q { when "derived-from-or-self(../type, 'vpn-common:dot1q')" { description "Only applies when the type of the tagged interface is 'dot1q'."; } description "Tagged interface."; uses ac-common:dot1q; } container priority-tagged { when "derived-from-or-self(../type, " + "'vpn-common:priority-tagged')" { description "Only applies when the type of the tagged interface is 'priority-tagged'."; } description "Priority-tagged interface."; uses ac-common:priority-tagged; } container qinq { when "derived-from-or-self(../type, 'vpn-common:qinq')" { description "Only applies when the type of the tagged interface is 'qinq'."; } description "Includes QinQ parameters."; uses ac-common:qinq; } } choice l2-service { description "The Layer 2 connectivity service can be provided by indicating a pointer to an L2VPN or by specifying a Layer 2 tunnel service."; container l2-tunnel-service { description "Defines a Layer 2 tunnel termination. It is only applicable when a tunnel is required."; uses ac-common:l2-tunnel-service; } case l2vpn { leaf l2vpn-id { type vpn-common:vpn-id; description "Indicates the L2VPN service associated with an Integrated Routing and Bridging (IRB) interface."; } } } leaf bearer-reference { if-feature "ac-common:server-assigned-reference"; type string; description "This is an internal reference for the service provider to identify the bearer associated with this AC."; } } // Basic IP connection grouping ip-connection-basic { description "Defines basic IP connection parameters."; container ipv4 { if-feature "vpn-common:ipv4"; description "IPv4-specific parameters."; uses ac-common:ipv4-connection-basic; } container ipv6 { if-feature "vpn-common:ipv6"; description "IPv6-specific parameters."; uses ac-common:ipv6-connection-basic; } } // Full IP connection grouping ip-connection { description "Defines IP connection parameters."; container ipv4 { if-feature "vpn-common:ipv4"; description "IPv4-specific parameters."; uses ac-common:ipv4-connection; } container ipv6 { if-feature "vpn-common:ipv6"; description "IPv6-specific parameters."; uses ac-common:ipv6-connection; } } // Routing protocol list grouping routing-protocol-list { description "List of routing protocols used on the AC."; leaf type { type identityref { base vpn-common:routing-protocol-type; } description "Type of routing protocol."; } list routing-profiles { key "id"; description "Routing profiles."; leaf id { type routing-profile-reference; description "Reference to the routing profile to be used."; } leaf type { type identityref { base vpn-common:ie-type; } description "Import, export, or both."; } } } // Static routing with BFD grouping ipv4-static-rtg-with-bfd { description "Configuration specific to IPv4 static routing with BFD."; list ipv4-lan-prefixes { if-feature "vpn-common:ipv4"; key "lan next-hop"; description "List of LAN prefixes for the site."; uses ac-common:ipv4-static-rtg-entry; leaf bfd-profile { if-feature "vpn-common:bfd"; type bfd-profile-reference; description "Points to a BFD profile."; } uses ac-common:service-status; } } grouping ipv6-static-rtg-with-bfd { description "Configuration specific to IPv6 static routing with BFD."; list ipv6-lan-prefixes { if-feature "vpn-common:ipv6"; key "lan next-hop"; description "List of LAN prefixes for the site."; uses ac-common:ipv4-static-rtg-entry; leaf bfd-profile { if-feature "vpn-common:bfd"; type bfd-profile-reference; description "Points to a BFD profile."; } uses ac-common:service-status; } } // BGP Service grouping bgp-neighbor-without-name { description "A grouping with generic parameters for configuring a BGP neighbor."; leaf remote-address { type inet:ip-address; description "The remote IP address of this entry's BGP peer. This is a customer IP address. If this leaf is not present, this means that the primary customer IP address is used as remote IP address."; } leaf local-address { type inet:ip-address; description "The provider's IP address that will be used to establish the BGP session."; } uses ac-common:bgp-peer-group-without-name; container bgp-max-prefix { description "A container for the maximum number of BGP prefixes allowed in the BGP session."; leaf max-prefix { type uint32; description "Indicates the maximum number of BGP prefixes allowed in the BGP session. It allows control of how many prefixes can be received from a neighbor."; reference "RFC 4271: A Border Gateway Protocol 4 (BGP-4), Section 8.2.2"; } } uses ac-common:bgp-authentication; uses ac-common:op-instructions; uses ac-common:service-status; } grouping bgp-neighbor-with-name { description "A grouping with generic parameters for configuring a BGP neighbor with an identifier."; leaf id { type string; description "A neighbor identifier."; } uses ac-svc:bgp-neighbor-without-name; } grouping bgp-neighbor-with-server-reference { description "A grouping with generic parameters for configuring a BGP neighbor with a reference generated by the provider."; leaf server-reference { if-feature "ac-common:server-assigned-reference"; type string; config false; description "This is an internal reference for the service provider to identify the BGP session."; } uses ac-svc:bgp-neighbor-without-name; } grouping bgp-neighbor-with-name-server-reference { description "A grouping with generic parameters for configuring a BGP neighbor with an identifier and a reference generated by the provider."; leaf id { type string; description "A neighbor identifier."; } uses ac-svc:bgp-neighbor-with-server-reference; } grouping bgp-svc { description "Configuration specific to BGP."; container peer-groups { description "Configuration for BGP peer-groups"; list peer-group { key "name"; description "List of BGP peer-groups configured on the local system - uniquely identified by peer-group name."; uses ac-common:bgp-peer-group-with-name; leaf local-address { type inet:ip-address; description "The provider's local IP address that will be used to establish the BGP session."; } container bgp-max-prefix { description "A container for the maximum number of BGP prefixes allowed in the BGP session."; leaf max-prefix { type uint32; description "Indicates the maximum number of BGP prefixes allowed in the BGP session. It allows control of how many prefixes can be received from a neighbor."; reference "RFC 4271: A Border Gateway Protocol 4 (BGP-4), Section 8.2.2"; } } uses ac-common:bgp-authentication; } } list neighbor { key "id"; description "List of BGP neighbors."; uses ac-svc:bgp-neighbor-with-name-server-reference; leaf peer-group { type leafref { path "../../peer-groups/peer-group/name"; } description "The peer-group with which this neighbor is associated."; } leaf bfd-profile { if-feature "vpn-common:bfd"; type bfd-profile-reference; description "Points to a BFD profile."; } } } // OSPF Service grouping ospf-svc { description "Service configuration specific to OSPF."; uses ac-common:ospf-basic; uses ac-common:ospf-authentication; uses ac-common:service-status; } // IS-IS Service grouping isis-svc { description "Service configuration specific to IS-IS."; uses ac-common:isis-basic; uses ac-common:isis-authentication; uses ac-common:service-status; } // RIP Service grouping rip-svc { description "Service configuration specific to RIP routing."; leaf address-family { type identityref { base vpn-common:address-family; } description "Indicates whether IPv4, IPv6, or both address families are to be activated."; } uses ac-common:rip-authentication; uses ac-common:service-status; } // VRRP Service grouping vrrp-svc { description "Service configuration specific to VRRP."; reference "RFC 5798: Virtual Router Redundancy Protocol (VRRP) Version 3 for IPv4 and IPv6"; leaf address-family { type identityref { base vpn-common:address-family; } description "Indicates whether IPv4, IPv6, or both address families are to be enabled."; } uses ac-common:service-status; } // Basic routing parameters grouping routing-basic { description "Defines basic parameters for routing protocols."; list routing-protocol { key "id"; description "List of routing protocols used on the AC."; leaf id { type string; description "Unique identifier for the routing protocol."; } uses routing-protocol-list; container bgp { when "derived-from-or-self(../type, 'vpn-common:bgp-routing')" { description "Only applies when the protocol is BGP."; } if-feature "vpn-common:rtg-bgp"; description "Configuration specific to BGP."; container peer-groups { description "Configuration for BGP peer-groups"; list peer-group { key "name"; description "List of BGP peer-groups configured on the local system - uniquely identified by peer-group name."; uses ac-common:bgp-peer-group-with-name; } } } container ospf { when "derived-from-or-self(../type, " + "'vpn-common:ospf-routing')" { description "Only applies when the protocol is OSPF."; } if-feature "vpn-common:rtg-ospf"; description "Configuration specific to OSPF."; uses ac-common:ospf-basic; } container isis { when "derived-from-or-self(../type, " + "'vpn-common:isis-routing')" { description "Only applies when the protocol is IS-IS."; } if-feature "vpn-common:rtg-isis"; description "Configuration specific to IS-IS."; uses ac-common:isis-basic; } container rip { when "derived-from-or-self(../type, " + "'vpn-common:rip-routing')" { description "Only applies when the protocol is RIP. For IPv4, the model assumes that RIP version 2 is used."; } if-feature "vpn-common:rtg-rip"; description "Configuration specific to RIP routing."; leaf address-family { type identityref { base vpn-common:address-family; } description "Indicates whether IPv4, IPv6, or both address families are to be activated."; } } container vrrp { when "derived-from-or-self(../type, " + "'vpn-common:vrrp-routing')" { description "Only applies when the protocol is the Virtual Router Redundancy Protocol (VRRP)."; } if-feature "vpn-common:rtg-vrrp"; description "Configuration specific to VRRP."; leaf address-family { type identityref { base vpn-common:address-family; } description "Indicates whether IPv4, IPv6, or both address families are to be enabled."; } } } } // Full routing parameters grouping routing { description "Defines routing protocols."; list routing-protocol { key "id"; description "List of routing protocols used on the AC."; leaf id { type string; description "Unique identifier for the routing protocol."; } uses routing-protocol-list; container static { when "derived-from-or-self(../type, " + "'vpn-common:static-routing')" { description "Only applies when the protocol is static routing protocol."; } description "Configuration specific to static routing."; container cascaded-lan-prefixes { description "LAN prefixes from the customer."; uses ipv4-static-rtg-with-bfd; uses ipv6-static-rtg-with-bfd; } } container bgp { when "derived-from-or-self(../type, " + "'vpn-common:bgp-routing')" { description "Only applies when the protocol is BGP."; } if-feature "vpn-common:rtg-bgp"; description "Configuration specific to BGP."; uses bgp-svc; } container ospf { when "derived-from-or-self(../type, " + "'vpn-common:ospf-routing')" { description "Only applies when the protocol is OSPF."; } if-feature "vpn-common:rtg-ospf"; description "Configuration specific to OSPF."; uses ospf-svc; } container isis { when "derived-from-or-self(../type, " + "'vpn-common:isis-routing')" { description "Only applies when the protocol is IS-IS."; } if-feature "vpn-common:rtg-isis"; description "Configuration specific to IS-IS."; uses isis-svc; } container rip { when "derived-from-or-self(../type, " + "'vpn-common:rip-routing')" { description "Only applies when the protocol is RIP. For IPv4, the model assumes that RIP version 2 is used."; } if-feature "vpn-common:rtg-rip"; description "Configuration specific to RIP routing."; uses rip-svc; } container vrrp { when "derived-from-or-self(../type, " + "'vpn-common:vrrp-routing')" { description "Only applies when the protocol is the Virtual Router Redundancy Protocol (VRRP)."; } if-feature "vpn-common:rtg-vrrp"; description "Configuration specific to VRRP."; uses vrrp-svc; } } } // Encryption choice grouping encryption-choice { description "Container for the encryption profile."; choice profile { description "Choice for the encryption profile."; case provider-profile { leaf provider-profile { type encryption-profile-reference; description "Reference to a provider encryption profile."; } } case customer-profile { leaf customer-key-chain { type key-chain:key-chain-ref; description "Customer-supplied key chain."; } } } } // Basic security parameters grouping ac-security-basic { description "AC-specific security parameters."; container encryption { if-feature "vpn-common:encryption"; description "Container for AC security encryption."; leaf enabled { type boolean; description "If set to 'true', traffic encryption on the connection is required. Otherwise, it is disabled."; } leaf layer { when "../enabled = 'true'" { description "Included only when encryption is enabled."; } type enumeration { enum layer2 { description "Encryption occurs at Layer 2."; } enum layer3 { description "Encryption occurs at Layer 3. For example, IPsec may be used when a customer requests Layer 3 encryption."; } } description "Indicates the layer on which encryption is applied."; } } container encryption-profile { when "../encryption/enabled = 'true'" { description "Indicates the layer on which encryption is enabled."; } description "Container for the encryption profile."; uses encryption-choice; } } // Bandwith parameters grouping bandwidth { description "Container for bandwidth."; container svc-pe-to-ce-bandwidth { if-feature "vpn-common:inbound-bw"; description "From the customer site's perspective, the inbound bandwidth of the AC or download bandwidth from the service provider to the site."; uses ac-common:bandwidth-per-type; } container svc-ce-to-pe-bandwidth { if-feature "vpn-common:outbound-bw"; description "From the customer site's perspective, the outbound bandwidth of the AC or upload bandwidth from the CE to the PE."; uses ac-common:bandwidth-per-type; } } // Basic AC parameters grouping ac-basic { description "Grouping for basic parameters for an attachment circuit."; leaf name { type string; description "A name that uniquely identifies the AC."; } container l2-connection { if-feature "ac-common:layer2-ac"; description "Defines Layer 2 protocols and parameters that are required to enable AC connectivity."; uses l2-connection-basic; } container ip-connection { if-feature "ac-common:layer3-ac"; description "Defines IP connection parameters."; uses ip-connection-basic; } container routing-protocols { description "Defines routing protocols."; uses routing-basic; } container oam { description "Defines the Operations, Administration, and Maintenance (OAM) mechanisms used."; container bfd { if-feature "vpn-common:bfd"; description "Container for BFD."; uses ac-common:bfd; } } container security { description "AC-specific security parameters."; uses ac-security-basic; } container service { description "AC-specific bandwith parameters."; leaf mtu { type uint32; units "bytes"; description "Layer 2 MTU."; } uses bandwidth; } } // Full AC parameters grouping ac { description "Grouping for an attachment circuit."; leaf name { type string; description "A name of the AC. Data models that need to reference an attachment circuit should use attachment-circuit-reference."; } leaf-list service-profile { type service-profile-reference; description "A reference to a service profile."; } container l2-connection { if-feature "ac-common:layer2-ac"; description "Defines Layer 2 protocols and parameters that are required to enable AC connectivity."; uses l2-connection; } container ip-connection { if-feature "ac-common:layer3-ac"; description "Defines IP connection parameters."; uses ip-connection; } container routing-protocols { description "Defines routing protocols."; uses routing; } container oam { description "Defines the OAM mechanisms used."; container bfd { if-feature "vpn-common:bfd"; description "Container for BFD."; leaf profile { type bfd-profile-reference; description "Points to a BFD profile."; } uses ac-common:bfd; uses ac-common:service-status; } } container security { description "AC-specific security parameters."; uses ac-security-basic; } container service { description "AC-specific bandwith parameters."; leaf mtu { type uint32; units "bytes"; description "Layer 2 MTU."; } uses bandwidth; container qos { if-feature "vpn-common:qos"; description "QoS configuration."; container qos-profiles { description "QoS profile configuration."; list qos-profile { key "profile"; description "Points to a QoS profile."; leaf profile { type qos-profile-reference; description "QoS profile to be used."; } leaf direction { type identityref { base vpn-common:qos-profile-direction; } description "The direction to which the QoS profile is applied."; } } } } container access-control-list { description "Container for the Access Control List (ACL)."; container acl-profiles { description "ACL profile configuration."; list acl-profile { key "profile"; description "Points to an ACL profile."; leaf profile { type forwarding-profile-reference; description "Forwarding profile to be used."; } } } } } } // Parent and Child ACs grouping ac-hierarchy { description "Container for parent and child AC references."; leaf ac-parent-ref { type ac-svc:attachment-circuit-reference; description "Specifies the parent AC that is inherited by an AC. In contexts where dynamic terminating points are bound to the same AC, a parent AC with stable information is created with a set of child ACs to track dynamic AC information."; } leaf-list child-ac-ref { type ac-svc:attachment-circuit-reference; config false; description "Specifies a child AC that relies upon a parent AC."; } } /******************** Main AC containers ********************/ container specific-provisioning-profiles { description "Contains a set of valid profiles to reference for an AC."; uses ac-common:ac-profile-cfg; } container service-provisioning-profiles { description "Contains a set of valid profiles to reference for an AC."; list service-profile-identifier { key "id"; description "List of generic service profile identifiers."; leaf id { type string; description "Identification of the service profile to be used. The profile only has significance within the service provider's administrative domain."; } } nacm:default-deny-write; } container attachment-circuits { description "Main container for the attachment circuits."; list ac-group-profile { key "name"; description "Maintains a list of profiles that are shared among a set of ACs."; uses ac; } container placement-constraints { description "Diversity constraint type."; uses vpn-common:placement-constraints; } leaf customer-name { type string; description "Indicates the name of the customer that requested these ACs."; } uses ac-common:op-instructions; list ac { key "name"; description "Global provisioning of attachment circuits."; leaf customer-name { type string; description "Indicates the name of the customer that requested this AC."; } leaf description { type string; description "Associates a description with an AC."; } leaf test-only { type empty; description "When present, this indicates that this is a feasibility check request. No resources are commited for such AC requests."; } uses ac-common:op-instructions; leaf role { type identityref { base ac-common:role; } description "Indicates whether this AC is used as UNI, NNI, etc."; } leaf-list peer-sap-id { type string; description "One or more peer SAPs can be indicated."; } leaf-list ac-group-profile { type ac-group-reference; description "A reference to an AC profile."; } uses ac-hierarchy; uses ac-common:redundancy-group; list service-ref { key "service-type service-id"; config false; description "Reports the set of services that are bound to the AC."; leaf service-type { type identityref { base vpn-common:service-type; } description "Indicates the service type (e.g., L3VPN or Network Slice Service)."; reference "RFC 9408: A YANG Network Data Model for Service Attachment Points (SAPs), Section 5"; } leaf service-id { type string; description "Indicates an identifier of a service instance of a given type that uses the AC."; } } leaf server-reference { if-feature "ac-common:server-assigned-reference"; type string; config false; description "Reports an internal reference for the service provider to identify the AC."; } uses ac; } } } <CODE ENDS>¶
This section uses the template described in Section 3.7 of [I-D.ietf-netmod-rfc8407bis].¶
The YANG modules specified in this document define schema for data that is designed to be accessed via network management protocols such as NETCONF [RFC6241] or RESTCONF [RFC8040]. The lowest NETCONF layer is the secure transport layer, and the mandatory-to-implement secure transport is Secure Shell (SSH) [RFC6242]. The lowest RESTCONF layer is HTTPS, and the mandatory-to-implement secure transport is TLS [RFC8446].¶
The Network Configuration Access Control Model (NACM) [RFC8341] provides the means to restrict access for particular NETCONF or RESTCONF users to a preconfigured subset of all available NETCONF or RESTCONF protocol operations and content.¶
There are a number of data nodes defined in these YANG modules that are writable/creatable/deletable (i.e., config true, which is the default). These data nodes may be considered sensitive or vulnerable in some network environments. Write operations (e.g., edit-config) and delete operations to these data nodes without proper protection or authentication can have a negative effect on network operations. Specifically, the following subtrees and data nodes have particular sensitivities/vulnerabilities in the "ietf-bearer-svc" module:¶
An attacker who is able to access this data node can modify the attributes to influence how a service is delivered to a customer, and this leads to Service Level Agreement (SLA) violations.¶
An attacker who is able to access this data node can modify the attributes of bearer and, thus, hinder how ACs are built.¶
In addition, an attacker could attempt to add a new bearer or delete existing ones. An attacker may also change the requested type or the activation scheduling.¶
The following subtrees and data nodes have particular sensitivities/vulnerabilities in the "ietf-ac-svc" module:¶
This container includes a set of sensitive data that influence how an AC will be delivered. For example, an attacker who has access to these data nodes may be able to manipulate routing policies, QoS policies, or encryption properties.¶
These profiles are defined with "nacm:default-deny-write" tagging [I-D.ietf-opsawg-teas-common-ac].¶
An attacker who has access to these data nodes may be able to manipulate service-specific policies to be applied for an AC.¶
This container is defined with "nacm:default-deny- write" tagging.¶
An attacker who is able to access this data node can modify the attributes of an AC (e.g., QoS, bandwidth, routing protocols, keying material), leading to malfunctioning of services that will be delivered over that AC and therefore to SLA violations. In addition, an attacker could attempt to add a new AC.¶
Some of the readable data nodes in these YANG modules may be considered sensitive or vulnerable in some network environments. It is thus important to control read access (e.g., via get, get-config, or notification) to these data nodes. Specifically, the following subtrees and data nodes have particular sensitivities/vulnerabilities in the "ietf-bearer-svc" module:¶
An attacker can retrieve privacy-related information about location from where the customer is connected. Disclosing such information may be used to infer the identity of the customer.¶
The following subtrees and data nodes have particular sensitivities/vulnerabilities in the "ietf-ac-svc" module:¶
An attacker can retrieve privacy-related information, which can be used to track a customer. Disclosing such information may be considered a violation of the customer-provider trust relationship.¶
An attacker can retrieve the cryptographic keys protecting the underlying connectivity services (routing, in particular). These keys could be used to inject spoofed routing advertisements.¶
Several data nodes ('bgp', 'ospf', 'isis', and 'rip') rely upon [RFC8177] for authentication purposes. As such, the AC service module inherits the security considerations discussed in Section 5 of [RFC8177]. Also, these data nodes support supplying explicit keys as strings in ASCII format. The use of keys in hexadecimal string format would afford greater key entropy with the same number of key- string octets. However, such a format is not included in this version of the AC service model because it is not supported by the underlying device modules (e.g., [RFC8695]).¶
IANA is requested to register the following URIs in the "ns" subregistry within the "IETF XML Registry" [RFC3688]:¶
URI: urn:ietf:params:xml:ns:yang:ietf-bearer-svc Registrant Contact: The IESG. XML: N/A; the requested URI is an XML namespace. URI: urn:ietf:params:xml:ns:yang:ietf-ac-svc Registrant Contact: The IESG. XML: N/A; the requested URI is an XML namespace.¶
IANA is requested to register the following YANG modules in the "YANG Module Names" subregistry [RFC6020] within the "YANG Parameters" registry.¶
Name: ietf-bearer-svc Maintained by IANA? N Namespace: urn:ietf:params:xml:ns:yang:ietf-bearer-svc Prefix: bearer-svc Reference: RFC xxxx Name: ietf-ac-svc Maintained by IANA? N Namespace: urn:ietf:params:xml:ns:yang:ietf-ac-svc Prefix: ac-svc Reference: RFC xxxx¶
This section includes a non-exhaustive list of examples to illustrate the use of the service models defined in this document. An example instance data can also be found at [Instance-Data].¶
An example of a request message body to create a bearer is shown in Figure 22.¶
A bearer-reference is then generated by the controller for this bearer. Figure 23 shows the example of a response message body that is sent by the controller to reply to a GET request:¶
Note that the response also indicates that Sync Phy is supported for this bearer.¶
An example of a request message body to create a simple AC over an existing bearer is shown in Figure 24. The bearer reference is assumed to be known to both the customer and the network provider. Such a reference can be retrieved, e.g., following the example described in Appendix A.1 or using other means (including, exchanged out-of-band or via proprietary APIs).¶
Figure 25 shows the message body of a response received from the controller and which indicates the "cvlan-id" that was assigned for the requested AC.¶
An example of a request to create a simple AC, when the peer SAP is known, is shown in Figure 26. In this example, the peer SAP identifier points to an identifier of an SF. The (topological) location of that SF is assumed to be known to the network controller. For example, this can be determined as part of an on-demand procedure to instantiate an SF in a cloud. That instantiated SF can be granted a connectivity service via the provider network.¶
Let us consider the example of an eNodeB (CE) that is directly connected to the access routers of the mobile backhaul (see Figure 27). In this example, two ACs are needed to service the eNodeB (e.g., distinct VLANs for Control and User Planes).¶
An example of a request to create the ACs to service the eNodeB is shown in Figure 28. This example assumes that static addressing is used for both ACs.¶
Figure 29 shows the message body of a response received from the controller.¶
The example shown Figure 29 is not optimal as it includes many redundant data. Figure 30 shows a more compact request that factorizes all the redundant data.¶
A customer may request adding a new AC by simply referring to an existing per-node AC profile as shown in Figure 31. This AC inherits all the data that was enclosed in the indicated per-node AC profile (IP addressing, routing, etc.).¶
When multiple ACs are requested by the same customer for the same site, the request can tag one of these ACs as "primary" and the other ones as "secondary". An example of such a request is shown in Figure 33. In this example, both ACs are bound to the same "group-id", and the "precedence" data node is set as a function of the intended role of each AC (primary or secondary).¶
Figure 34 shows an example of CEs that are interconnected by a service provider network.¶
Figure 35 depicts an example of the message body of a response to a request to instantiate the various ACs that are shown in Figure 34.¶
This example shows how the AC service model complements [I-D.ietf-teas-ietf-network-slice-nbi-yang] to connect a site to a slice service.¶
First, Figure 36 describes the end-to-end network topology as well the orchestration scopes:¶
The topology is made up of two sites ("site1" and "site2"), interconnected via a Transport Network (e.g. IP/MPLS Network). An SF is deployed within each site in a dedicated IP Subnet.¶
A 5G SMO is responsible for the deployment of SFs and the indirect management of a local Gateway (i.e., CE device).¶
An IETF Network Slice Controller is responsible for the deployment of IETF Network Slices across the TN.¶
SFs are deployed within each site.¶
Figure 37 describes the logical connectivity enforced thanks to both IETF Network Slice and Attachment Circuit models.¶
Figure 38 shows the message body of the request to create the required ACs using the Attachment Circuit module.¶
Figure 39 shows the message body of a response received from the controller.¶
Figure 40 shows the message body of the request to create the a slice service bound to the ACs created using Figure 38. Only references to these ACs are included in the Slice Service request. This example assumes that the module that "glues" the service/AC is also supported by the NSC.¶
This example (Figure 41) shows how the AC service model can be used to connect a Cloud Infrastructure to a service provider network. This example makes the following assumptions:¶
A customer (e.g., Mobile Network Team or partner) has a virtualized infrastructure running in a Cloud Provider. A simplistic deployment is represented here with a set of Virtual Machines running in a Virtual Private Environment. The deployment and management of this infrastructure is achieved via private APIs that are supported by the Cloud Provider: this realization is out of the scope of this document.¶
The connectivity to the Data Center is achieved thanks to a service based on direct attachment (physical connection), which is delivered upon ordering via an API exposed by the Cloud Provider. When ordering that connection, a unique "Connection Identifier" is generated and returned via the API.¶
The customer provisions the networking logic within the Cloud Provider based on that unique connection Identifier (i.e., logical interfaces, IP addressing, and routing).¶
Figure 42 illustrates the pre-provisioning logic for the physical connection to the Cloud Provider. After this connection is delivered to the service provider, the network inventory is updated with "bearer-reference" set to the value of the "Connection Identifier".¶
Next, API workflows can be initiated:¶
Cloud Provider for the configuration as per (3) above.¶
Service provider network via the Attachment Circuit model. This request can be used in conjunction with additional requests based on the L3SM (VPN provisioning) or Network Slice Service model (5G hybrid Cloud deployment).¶
Figure 43 shows the message body of the request to create the required ACs to connect the Cloud Provider Virtualized (VM) using the Attachment Circuit module.¶
Figure 44 shows the message body of the response received from the provider. Note that this Cloud Provider mandates the use of MD5 authentication for establishing BGP connections.¶
The module supports MD5 to basically accommodate the installed BGP base (including by some Cloud Providers). Note that MD5 suffers from the security weaknesses discussed in Section 2 of [RFC6151] and Section 2.1 of [RFC6952].¶
CE-PE routing using BGP is a common scenario in the context of MPLS VPNs and is widely used in enterprise networks. In the example depicted in Figure 45, the CE routers are customer-owned devices belonging to an AS (ASN 65536). CEs are located at the edge of the provider's network (PE, or Provider Edge) and use point-to-point interfaces to establish BGP sessions. The point-to-point interfaces rely upon a physical bearer ("Line-113") to reach the provider network.¶
The attachment circuit in this case use a SAP identifier to refer to the physical interface used for the connection between the PE and the CE. The attachment circuit includes all the additional logical attributes to describe the connection between the two ends, including VLAN information and IP addressing. Also, the configuration details of the BGP session makes use of peer group details instead of defining the entire configuration inside the 'neighbor' data node.¶
This scenario allows the provider to maintain a list of ACs belonging to the same customer without requiring the full service configuration.¶
This section illustrates how to use the AC service model for interconnection purposes. To that aim, we assume a simplified Internet eXchange Point (IXP) configuration without zooming into IXP deployment specifics. Let us assume that networks are interconnected via a Layer 2 facility. BGP is used to exchange routing information and reachability announcements between those networks. The same approach can be used to negotiate interconnection between two networks and without involving an IXP.¶
As depicted in Figure 47, each network connects to the IXP switch via a bearer over which an AC is created. The AC configuration (Figure 48) includes parameters such as VLAN configuration, IP addresses, MTU, and any additional settings required for connectivity. The peering location is inferred from the "bearer-reference".¶
Figure 49 shows the received response with the required information for the activation of the AC.¶
Once the ACs are established, BGP peering sessions can be configured between routers of the participating networks. BGP sessions can be established via a route server or between two networks. For the sake of illustration, let us assume that BGP sessions are established directly between two network. Figure 50 shows an example of a request to add a BGP session to an existing AC. The properties of that AC are not repeated in this request because that information is already communicated during the creation of the AC.¶
Figure 51 provides the example of a response which indicates that the request is awaiting validation. The response includes also a server-assigned reference for this BGP session.¶
Once validation is accomplished, a status update is communicated back to the requestor. The BGP session can then be established over the AC. The BGP session configuration includes parameters such as neighbor IP addresses, ASNs, authentication settings (if required), etc. The configuration is triggered at each side of the BGP connection.¶
This section demonstrates how the AC service model permits to manage connectivity requirements for complex Network Functions (NFs) - containerized or virtualized - that are typically deployed in Telco networks. This integration leverages the concept of "parent AC" to decouple physical and logical connectivity so that several ACs can shares Layer 2 and Layer 3 resources. This approach provides flexibility, scalability, and API stability.¶
NF is used to refer to the SF defined [RFC7665]. NF is used here as this term is widely used outside the IETF.¶
The NFs have the following characteristics:¶
The NF is distributed on a set of compute nodes with scaled-out and redundant instances.¶
The NF has two distinct type of instances: user plane ("nf-up") and routing control plane ("nf-cp").¶
The User plane component can be distributed among the first 8 compute nodes ("compute-01" to "compute-08") to achieve high performance.¶
The Control plane is deployed in a redundant fashion on two instances running on distinct compute nodes ("compute-09" and "compute-10").¶
The NF is attached to distinct networks, each making use of a dedicated VLAN. These VLANs are therefore instantiated as separate ACs. From a realization standpoint, the NF interface connectivity is generally provided thanks to MacVLAN or Single Root I/O Virtualization (SR-IOV). For the sake of simplicity only two VLANs are presented in this example, additional VLANs are configured following a similar logic.¶
Figure 53 describes the physical infrastructure. The compute nodes (customer) are attached to the provider infrastructure thanks to a set of physical links on which attachment circuits are provisioned (i.e., "compute-XX-nicY"). The provider infrastructure can be realized in multiple ways, such as IP Fabric, Layer 2/Layer 3 Edge Routers. This document does not intend to detail these aspects.¶
The NFs are deployed on this infrastructure in the following way:¶
Configuration of a parent AC as a centralized attachment for "vlan 100". The parent AC captures Layer 2 and Layer 3 properties for this VLAN: vlan-id, IP default gateway and subnet, IP address pool for NFs endpoints, static routes with BFD to user plane and BGP configuration to control plane NFs.¶
Configuration of a parent AC as a centralized attachment for "vlan 200". This vlan is for Layer 2 connectivity between NFs (no IP configuration in the provider network).¶
"Child ACs" binding bearers to parent ACs for "vlan 100" and "vlan 200".¶
The deployment deploys the network service to all compute nodes ("compute-01" to "compute-10"), even though the NF is not instantiated on "compute-07"/"compute-08". This approach permits to handle compute failures and scale-out scenarios in a reactive and flexible fashion thanks to a pre-provisioned networking logic.¶
For readability the payload is displayed as single JSON file (Figure 55). In practice, several API calls may take place to initialize these resources (e.g., GET requests from the customer to retrieve the IP address pools for NFs on "vlan 100" thanks to parent configuration and BGP configuration, and POST extra routes for user planes and BFD).¶
Note that no individual IP address is assigned in the data model for the NF user plane instances (i.e., no "customer-address" in the Child AC). The assignment of IP addresses to the NF endpoints is managed by the Cloud Infrastructure IPAM based on the customer-addresses IP address pool "192.0.2.1-200". Like in any standard LAN-facing scenario, it is assumed that the actual binding of IP endpoints to logical attachments (here Child ACs) relies on a dedicated protocol logic (typically, ARP or NDP) and is not captured in the data model. Hence, the IP addresses displayed for NF user plane instances are simply examples of a realization approach. Note also that the Control Plane is defined with static IP address assignment on a given AC/bearer to illustrate another deployment alternative.¶
Assuming a failure of "compute-01", the instance "nf-up-1" can be redeployed to "compute-07" by the NF/Cloud Orchestration. Additionally, the NF can be scaled-out thanks to the creation of an extra instance "nf-up7" on "compute-08". Since connectivity is pre-provisioned, these operations happen without any API calls. In other words, this redeployment is transparent from the perspective of the configuration of the provider network.¶
Finally, the addition or deletion of compute nodes in the deployment ("compute-11", "compute-12", etc.) involves merely changes on Child ACs and possible routing on the parent AC. In any case, the parent AC is a stable identifier, which can be consumed as a reference by end-to-end service models for VPN configuration such as [I-D.ietf-opsawg-ac-lxsm-lxnm-glue], Slice Service [I-D.ietf-teas-ietf-network-slice-nbi-yang], etc. This decoupling to a stable identifier provides great benefits in terms of scalability and flexibility since once the reference with the parent AC is implemented, no API calls to the VPN model are needed for any modification in the cloud.¶
This document leverages [RFC9182] and [RFC9291]. Thanks to Gyan Mishra for the review.¶
Thanks to Ebben Aries for the YANG Doctors review and for providing [Instance-Data].¶
Thanks to Donald Eastlake for the careful rtg-dir review.¶