Internet-Draft | draft-dong-bmwg-network-slicing-00 | July 2025 |
Dong & Contreras | Expires 8 January 2026 | [Page] |
Network slicing aims to provide assurance of specific network performance objectives for network services which require both connectivity and specific performance commitment. Such network services are considered as network slice services. This document provides a benchmarking methodology for network slicing, focusing on evaluating the key functionalities of network slicing mechanisms and the performance of network slice services. The network slicing functionalities includes the data plane, control plane and management plane mechanisms for realizing network slice service, and the performance of network slice service includes the service level agreement (SLA) commitments (bandwidth, delay, and jitter), path constraints and resource guarantee.¶
The tests aim to demonstrate how network slicing can support competing services in a shared network, ensuring that critical network services in one network slice remain unaffected by congestion or unexpected behavior of other traffic in the same network.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].¶
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 8 January 2026.¶
Copyright (c) 2025 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.¶
Network slicing aims to provide assurance of specific network performance objectives for network services which require both connectivity and specific performance commitment. Such network services are considered as network slice services. [RFC9543] describes the general framework for requesting and operating network slices built from IETF technologies. [RFC9543] also introduces the concept of Network Resource Partition (NRP), which is a subset of the buffer/queuing/scheduling resources and associated policies on each of a connected set of links in the underlay network. NRP can be created to support specific SLAs of one or a group of network slice services. In a default deployment, the network resources are not partitioned, which means all the network service are provisioned over the underlay network with shared network resources. Depending on the types of services carried in the network, this may or may not meet all the service requirements.¶
[RFC9732] describes a framework for Enhanced Virtual Private Networks (VPNs) based on Network Resource Partitions (NRPs) to support the needs of applications with specific traffic performance requirements (e.g., low latency, bounded jitter). NRP-based enhanced VPN can be used to deliver network slice services.¶
As outlined in [RFC9543], network slicing is crucial for 5G services and beyond, where diverse service requirements demand tailored performance guarantees. Benchmarking of network slicing is essential to understand its effect on service assurance, particularly in high-demand service environments. Evaluating performance aspects such as resource guarantee and SLA adherence helps service providers to understand different types of network slicing mechanisms and choose the suitable one for their network scenarios and demands.¶
This document provides benchmarking guidelines to evaluate the key functionalities and effectiveness of network slicing, referencing [RFC2544] for benchmarking principles. More specifically, it focuses on the benchmarking of 1) Network Resource partitioning for network slices; 2) Network Slice Topology Control; 3) SR Path with Resource Guarantee 4) Service SLA assurance, including bandwidth, latency etc..¶
Network Slice: As defined in [RFC9543], An IETF Network Slice enables connectivity between a set of SDPs with specific Service Level Objectives (SLOs) and Service Level Expectations (SLEs) over a common underlay network.¶
Network Resource Partition (NRP): As defined in [RFC9543], an NRP is a subset of the buffer/queuing/scheduling resources and associated policies on each of a connected set of links in the underlay network.¶
End-to-End SLA: A guaranteed level of service performance across an entire network path.¶
Benchmarking tests will be conducted in a controlled environment, consisting of:¶
Hardware: Network devices capable of supporting network slicing.¶
Software: Network slice control and management tools.¶
Traffic Generator: Configured to simulate background and high-value service traffic scenarios.¶
Background Traffic: Continuous traffic aimed at saturating network links to simulate congestion.¶
High-Value Traffic: Critical service traffic which requires guaranteed bandwidth, low latency etc., such as voice or video streaming.¶
Bandwidth of network slice service: The amount of bandwidth consumed by a network slice service.¶
Delay (Latency) of network slice service: Time taken for a data packet to traverse the network via a network slice.¶
Jitter of network slice service: Variation of packet arrival time in a network slice, which can affect time-sensitive applications.¶
Network Slice +---+ +---+ +---+ +--- |PE1|----------| P1|------------|PE3|---+ | +-+-+ +-+-+ +-+-+ | | | | | | | +-+-+ +---+ +-+-+ | +--- |PE2|----------| P2|------------|PE4|---+ | +---+ +---+ +---+ | | +-----------------+ | +----------|Traffic Generator|---------------+ +-----------------+ Test Topology¶
Objective: Verify that whether the device supports the partitioning of link resources to form an NRP.¶
Procedure:¶
Create an NRP instance, specify the set of links which belong to the NRP.¶
On each of the link of the NRP, configure the set of bandwidth to be reserved for the NRP.¶
Generate background traffic in the network.¶
Generate test traffic with NRP Selector ID [I-D.ietf-6man-enhanced-vpn-vtn-id] carried in data packet, test the NRP throughput on each link .¶
Expected Results:¶
Objective: Verify the mechanism to control the topology of an NRP, so that network slice traffic will only be forwarded along the paths within the NRP topology.¶
Procedure:¶
Create an NRP instance, reserve link bandwidth on each link of the NRP.¶
Define a logical topology which aligns with the NRP topology using either Flexible-Algorithm [RFC9350] or Multi-Topology [RFC5120] with Segment Routing.¶
Generate background traffic in the network.¶
Generate test traffic with Algorithm/Topology -specific SR SID and NRP Selector ID carried in the packet, check the packet forwarding path and throughput.¶
Expected Results:¶
The NRP instance is created successfully, and the link resources for the NRP are reserved successfully.¶
The Flex-Algo or Multi-topology is created successfully, and aligns with the topology of NRP.¶
The test traffic is sent along the paths within the specified Flex-Algorithm or MT, no traffic goes to links out side the topology of NRP.¶
The throughput of test traffic is the same as the reserved bandwidth on the set of links of the NRP.¶
Objective: Verify that an SR Policy can be associated with an NRP to provide guaranteed performance for network slice service¶
Procedure:¶
Provision an SR Policy [RFC9256] with one or multiple candidate paths, each consists of one or multiple segment lists.¶
Create an NRP instance, reserve link bandwidth on each link of the NRP. The topology of the NRP covers all the candidate paths and segment lists of the SR Policy.¶
Associate the SR Policy with the NRP for resource guarantee.¶
Generate background traffic in the network.¶
Generate test traffic with the SID list and NRP Selector ID carried in the packet, check the packet forwarding path and throughput.¶
Expected Results:¶
The NRP instance is created successfully, and the link resources for the NRP are reserved successfully.¶
The SR Policy is created successfully, and is associated with the NRP.¶
The test traffic is sent along the paths specified by the SID list of the SR Policy.¶
The throughput of test traffic is the same as the reserved bandwidth on the set of links of the NRP.¶
Objective: Verify that the bandwidth required by a network slice service can be guaranteed.¶
Procedure:¶
Provision an L3VPN or EVPN service with the requirement on high bandwidth.¶
Provision an SR Policy with one or multiple candidate paths, each consists of one or multiple SID lists according to the connectivity and bandwidth requirement of the L3VPN or EVPN service.¶
Create an NRP instance, reserve link bandwidth on each link of the NRP to meet the bandwidth requirement of the network slice service. The topology of the NRP covers all the candidate paths of the SR Policy.¶
Associate the SR Policy with the NRP for bandwidth guarantee.¶
Generate background traffic in the network, make the network congested.¶
Generate test traffic of the VPN service, steer the VPN service into the SR Policy. Check the packet forwarding path and throughput.¶
Expected Results:¶
The VPN service is provisioned successfully.¶
The NRP instance is created successfully, and the link resources for the NRP are reserved successfully.¶
The SR Policy is created successfully, and is associated with the NRP.¶
The VPN service test traffic is sent along the paths specified by the SID list of the SR Policy, and the packet is encapsulated with the NRP Selector ID.¶
The throughput of test VPN traffic is the same as the required bandwidth of the service.¶
Objective: Verify that the latency required by a network slice service can be guaranteed.¶
Procedure:¶
Provision an L3VPN or EVPN service with specific requirement on bandwidth and latency.¶
Provision an SR Policy with one or multiple candidate paths, each consists of one or multiple SID lists which can meet the latency requirement of the L3VPN or EVPN service.¶
Create an NRP instance, reserve link bandwidth on each link of the NRP to meet the SLA requirement of the network slice service. The topology of the NRP covers all the candidate paths of the SR Policy.¶
Associate the SR Policy with the NRP for latency guarantee.¶
Generate background traffic in the network, make the network congested.¶
Generate test traffic of the VPN service, steer the VPN service into the SR Policy. Check the packet forwarding path, throughput and latency.¶
Expected Results:¶
The VPN service is provisioned successfully.¶
The NRP instance is created successfully, and the link resources for the NRP are reserved successfully.¶
The SR Policy is created successfully, and is associated with the NRP.¶
The VPN service test traffic is sent along the paths specified by the SID list of the SR Policy, and the packet is encapsulated with the NRP Selector ID.¶
The throughput of test VPN traffic is the same as the required bandwidth of the service, the latency of the traffic meet the required latency of the service.¶
This document makes no request of IANA.¶
Note to RFC Editor: this section may be removed on publication as an RFC.¶
The authors would like to thank Xuesong Geng for the review and valuable suggestions.¶