NeoTec Y. Zhang
Internet-Draft C. Xie
Intended status: Informational China Telecom
Expires: 23 April 2025 20 October 2024
A Use Case of Network Operation for Telecom Cloud
draft-zx-neotec-net4cloud-usecase-00
Abstract
This document discusses the network operation issues of telecom
cloud. It first introduces a typical use case of network for telecom
cloud, then illustrates the requirements for network operation for
telecom cloud from the perspective of TSP, and proposes a general
network operation model for telecom cloud. It also analyzes the gap
based on the current technological status.
Status of This Memo
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 23 April 2025.
Copyright Notice
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.
Zhang & Xie Expires 23 April 2025 [Page 1]
Internet-Draft net4cloud usecase October 2024
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Use Case of Network for Telecom Cloud . . . . . . . . . . . . 3
4. Requirements . . . . . . . . . . . . . . . . . . . . . . . . 7
5. Network Operation Model for Telecom Cloud . . . . . . . . . . 7
6. Gap analysis . . . . . . . . . . . . . . . . . . . . . . . . 9
7. Security Considerations . . . . . . . . . . . . . . . . . . . 10
8. Normative References . . . . . . . . . . . . . . . . . . . . 10
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 10
1. Introduction
Nowadays, the resources required for internet services and
applications mainly exist in the form of cloud, which is distributed,
decentralized and highly dynamic. As a result, the network planning
and operation methods need to be adjusted to meet the demands of
cloud.
Telecom cloud refers to a cloud computing system owned and operated
by Telecom Service Provider (TSP). In general, TSPs own and operate
global wide area networks, 5G and cellular networks. TSPs have also
developed Telecom Cloud, which is a cloud environment optimized for
service delivery through the integration of network and cloud
infrastructure.
To support cloud computing services, TSPs build overlay networks for
corporate customer on the underlay network, and provides
differentiated network services in terms of bandwidth, latency and
security isolation to meet the flexibility needs of customers and
applications. Due to various reasons, overlay network and
traditional underlay network have significant differences in planning
and operation: overlay network can be quickly created and iteratively
upgraded by TSPs, but it has poor awareness of the underlying
network's operational status and potential failures. Traditional
underlay network focuses on traffic and connectivity, physical
pipeline planning between cloud resource pools and network solutions
for individual cloud network products, but not on application
scenarios. The network operation for telecom cloud is facing many
challenges due to the lack of a customer or service-centric approach,
as well as the inability to fully meet the demands of cloud services
in terms of flexibility, openness, and response speed.
For the illustration of network operation for telecom cloud, this
document provides a typical use case of telecom cloud network. It
first introduces the architecture, then lists the requirements for
operations of network for telecom cloud from the perspective of TSPs.
Zhang & Xie Expires 23 April 2025 [Page 2]
Internet-Draft net4cloud usecase October 2024
Additionally, it proposes a network operation model for telecom
cloud, it also analyzes the current gaps in conjunction with the use
case.
2. Terminology
The following terms are used in this document:
CE: Customer Edge
IOAM: Input Output Access Management
MPLS: Multi-protocol Label Switching
OTN: Optical Transport Network
PE: Provider Edge
POP: Point of Presence
PW: Pseudo Wire
SD-WAN: Software-Defined Wide Area Network
SRv6: Segment Routing IPv6
TSP: Telecom Service Provider
VLAN: Virtual Local Area Network
VPC: Virtual Private Cloud
3. Use Case of Network for Telecom Cloud
For the provisioning of cloud services to its customers, a TSP
deploys its large-scale network in a distributed manner, with its
backbone underlay network using MPLS or SRv6+EVPN methods, as shown
in figure 1. PE devices is located at the edge of the backbone
network, responsible for connecting CE devices. This Backbone
Network connects cloud resource pools which distributes over a wide
geographic area, so its users are capable of accessing local telecom
cloud as well as remote telecom cloud. On the access side, CE
devices are located at the edge of customer sites, used to establish
connections with PE devices. When customers need to access the
cloud, they can choose multiple methods, such as: IP RAN, PON, direct
fiber drive, OTN, SD-WAN, 5G slice dedicated line. Telecom cloud
nodes can be deployed on the operator's Backbone Network or at the
network edge. Operators provide cloud services to enterprise
Zhang & Xie Expires 23 April 2025 [Page 3]
Internet-Draft net4cloud usecase October 2024
customers, including elastic computing, storage, networking, and
security infrastructure resources, supporting enterprises to quickly
build and deploy application systems and achieve rapid activation of
internet services.
+------------+ +------------+
| Core Cloud | | Core Cloud |
+------------+ +------------+
| |
+------------+ +----+ +----+ +------------------+
| Edge Cloud |---| CE |---| PE |---| |
+------------+ +----+ +----+ | |
+-----------------+ +----+ | | +--------------+
| Customer Site 1 |---| PE |---| Backbone Network |---| 3rd-Party |
+-----------------+ +----+ | | | Public Cloud |
+-----------------+ +----+ | | +--------------+
| Customer Site 2 |---| PE |---| |
+-----------------+ +----+ +------------------+
Figure 1. Diagram of typical network for Telecom Cloud
In scenarios such as medical education and game streaming, one
customer uploads data to the edge cloud through campus network and
private cloud. Important data, or data that has been preliminarily
processed by the edge cloud, is then uploaded to the core cloud POP.
At the same time, the core cloud also transmits data to the edge
cloud. Ultimately, customers can obtain the required data from
either the edge cloud or the core cloud.
Additionally, the operation of the cloud service requires underlay
network support. The cloud network service carrying methods for
telecom cloud are mainly of two types:
(1) Cloud Access Service
Cloud access services allow customers to seamlessly connect their IT
resources or data centers to telecom cloud resource pools via the
telecom cloud management/control system. These resource pool sites
are typically deployed in proximity to customers to minimize data
transmission latency and deliver a more localized service
experience.This method not only provides network connection services
but also integrates cloud computing resources.
The current activation methods for Cloud Access Service include six
types of network-side access: IP RAN, SD-WAN, PON, direct fiber
connection, 5G slicing dedicated line, and OTN. Each of these
requires their respective network OSS systems to enable access
services for cloud applications.
Zhang & Xie Expires 23 April 2025 [Page 4]
Internet-Draft net4cloud usecase October 2024
For example, customers directly access telecom cloud through the IP
RAN network. The IP RAN technology itself supports multiple service
types, ensuring low latency and high reliability of data
transmission, suitable for enterprise applications with higher
network performance requirements. It can dynamically adjust network
resources according to enterprise customer needs.
In this architecture, customers access the local operator's boundary
equipment (PE) of the Backbone Network through the IP RAN network.
The customer’s boundary equipment (CE) establishes a layer 2 channel
with the ASBR/B-LEAF device through VLAN+PW. The CE connects to the
local IP RAN U equipment in a single or dual uplink mode,
interconnected in a layer 2 VLAN manner. A PW is established between
the customer-accessed IP RAN U equipment and the ASBR/B-LEAF device
to carry the customer’s layer 2 channel. The ASBR device is
interconnected with the local Backbone Network PE equipment within
the local metropolitan area The CE acts as the three-layer gateway of
its own internal network, being neighbors with the Backbone Network
PE equipment at the three-layer level. The above ASBR/B-LEAF devices
reuse the IP RAN cloud dedicated line to provide customers with end-
to-end complete services.
+----------+ +----+ +---+ +-------+ +------+ +----+ +----------+
| Customer |---| CE |---| U |---| IPRAN |---| ASBR |---| PE |---| Backbone |
| Site | +----+ +---+ +-------+ +------+ +----+ | Network |
+----------+ +----------+
Figure 2. Diagram of Cloud Access Service network
In this architecture, the cloud resource pool's cloud dedicated line
switch acts as CE, connecting to the PE provided by the Backbone
Network in the resource pool's city through a pre-configured high-
bandwidth relay link. Customers can conveniently access services
through the nearest access point. The connection between the cloud
dedicated line switch and the Backbone Network PE adopts VLAN sub-
interfaces to achieve flexible network configuration. In the entire
dedicated line service, the cloud dedicated line switch provides
access functions, while the inter-cloud high-speed network handles
networking and traffic scheduling tasks. Different customers within
the cloud resource pool achieve isolation through their own
independent virtual routing and forwarding instances, ensuring
security and privacy.
(2) Inter-Cloud Connection Service
To meet the needs of customers for networking between different cloud
data centers, TSPs utilize cloud resource pools distributed in
different regions to construct overlay network, such as SD-WAN or VPN
on top of the underlay network, as shown in figure 3. The overlay
Zhang & Xie Expires 23 April 2025 [Page 5]
Internet-Draft net4cloud usecase October 2024
network can quickly adjust configurations according to each
customer's and each service's needs without being constrained by
physical network elements in most cases. The Overlay network
directly serves customers on the cloud, providing customer-level/
application-level virtual network connection, management, and
monitoring services.
The overlay network includes intra-cloud and inter-cloud. The intra-
cloud network mainly refers to VPC, a specific VPC acts as the core
and links all intra-cloud network services. VPC allows customers to
build a secure, reliable, and configurable/manageable virtual network
environment in the cloud while also supporting external network
connections with various native cloud gateways and access to third-
party public clouds.
The underlay network provides high reliability and low latency
network connection services for the overlay network on the cloud. To
support the overlay network, it is necessary to open basic network
capabilities to the cloud management system through the underlay
network's central controller and reserve resources such as VPN for
the Overlay network. However, it is not aware of customers and
applications.
The establishment of an overlay network include manual configuration
and automated deployment. Enterprise customers can manually
configure parameters of network devices (such as routers, switches)
according to their actual needs, including tunnel endpoints,
encapsulation protocols, etc. This method only works for small-scale
network environments, but for large enterprises, manual configuration
may be too cumbersome and prone to errors, necessitating automated
deployment through network orchestrators.
+-------+ VxLAN +-------+
Overlay Network | |<------------------------------->| |
| Edge | +-----------+ | Cloud |
+-----+ IPSec | Cloud | VxLAN | Municipal | VxLAN | |
| |<----->| |<------>| Cloud |<---------->| |
+------+ | | +-------+ +-----------+ +-------+
| USER |---| CPE | | POP | | POP | | POP |
+------+ | | +-----+ +--------------+ +------------------+ |
| |---| |---| IP RAN |---| Backbone Network |----+
+-----+ | MAN | | +----------+ | | +----------+ |
| | | | MPLS vpn | | | | MPLS vpn | |
Underlay Network +-----+ | +----------+ | | +----------+ |
+--------------+ +------------------+
Figure 3. Diagram of constructing an Overlay network on Underlay network
Zhang & Xie Expires 23 April 2025 [Page 6]
Internet-Draft net4cloud usecase October 2024
4. Requirements
From the perspective of TSPs, the network operation for telecom cloud
should meet the following requirements:
1) Agile provisioning: Capable of rapidly deploying and configuring
cloud and network resources to promptly meet cloud service needs.
2) Flexible adjustment of bandwidth and other resources: Dynamically
adjust network bandwidth and other resources according to actual
needs within minutes, ensuring efficient resource utilization.
3) High stability and reliability: For new cloud services, such as AI
computing, network failures could lead to training task interruptions
or data loss. The network needs to possess high stability and
reliability to ensure continuous and complete data transmission.
4) Network performance monitoring: The network operation system
should have a unified interface and scheduling ability to ensure
operable and manageable maintenance. The system should be able to
monitor the network's operating status in real-time, quickly detect
and locate network problems of various computing nodes. It should
also support correlation analysis between application traffic and
network status, providing customers with self-healing capabilities
for overlay networks. Through monitoring collaboration between cloud
and network, it realizes predictive and planning functions of the
network.
5) Centralized control: Connect to DC, metropolitan area networks,
backbone networks, and other network domains, apply SDN/NFV and other
new technologies to achieve centralized control and dynamic
management of network elements.
6) Scenario visualization: Efficiently and promptly perceive network
operational data, sense the real-time operating status of the
network, and automatically discover changes in the network in a
timely manner. Realize panoramic visibility of end-to-end service
paths and SLA visibility in the network.
5. Network Operation Model for Telecom Cloud
Telecom cloud must provide enterprises with flexible cloud computing
resources, supporting the deployment and operation of various cloud
applications and services, while also offering customers flexible
network configuration and deployment services. Therefore, its
network operations need to be optimized. This section introduces the
network operation model for telecom cloud. The top layer is the
Telecom Cloud Service Management/Controller System, which includes
Zhang & Xie Expires 23 April 2025 [Page 7]
Internet-Draft net4cloud usecase October 2024
components such as orchestrator, controllers, fault operations,
resource management, etc. The Telecom Cloud Service Management/
Controller System is responsible for comprehensively managing telecom
cloud, building and operating overlay network. At the same time,
through interfaces with the Network Management/Controller System, it
controls and mobilizes the capabilities of the underlying underlay
network to meet the demands of cloud computing above. This model
also supports interoperability with third-party Public Cloud.
+----------------------------+
| Application |
+----------------------------+
^ ^
| |
v v
+---------------------+ +---------------------+
| Telecom Cloud | | |
| +-----------------+ | | Telecom Cloud |
| | VPC | |<-->| Service Management/ |
| |-----------------| | | Control System |
| | Overlay Network | | | |
| +-----------------+ | | |
+---------------------+ +---------------------+
^ ^
| | Interface
v v
+--------------+ +------------------+ +---------------------+
| 3rd-Party |<--->| Underlay Network |<--->| Network Management/ |
| Public Cloud | | | | Control System |
+--------------+ +------------------+ +---------------------+
Figure 4. Network Operation Model for Telecom Cloud
In this model, unified interfaces need to be established between the
Telecom Cloud Service Management/Control System and the Network
Control System to achieve seamless data exchange and command
transmission. This includes standardized API calls, data formats,
and communication protocols.
The main interfaces instructions sent by the Telecom Cloud Service
Management/Control System to the Network Management/Control System
include:
* Network establishment
* Cloud service operation data annoucement
* Network resource scaling on connection, bandwidth, latency, path,
etc.
Zhang & Xie Expires 23 April 2025 [Page 8]
Internet-Draft net4cloud usecase October 2024
* Retrieval of network performance data
* Retrieval of network resource usage data
* Network traffic scheduling
* Network termination
Similarly, the Network Management/Control System can also report back
to the Telecom Cloud Service Management/Control System. Its
interface includes:
* Proactive network fault reporting
* Proactive network performance reporting
* Proactive network resource usage reporting
* Proactive underlay network operation reporting
6. Gap analysis
Although telecom cloud providers have advantages in providing cloud-
network integrated services, most networks, while extensive, are
often rigid in operation, making them difficult to fully adapt to
modern cloud service needs.
As telecom cloud evolves towards multi-cloud and edge cloud models,
new challenges in configuration, automation, and operation emerge.
Moreover, interoperability becomes a key issue, as TSPs typically
rely on equipment from multiple vendors, while public cloud providers
can standardize on a single vendor ecosystem. Telecom cloud services
must seamlessly integrate various software and hardware systems,
creating challenges and opportunities for collaboration, especially
within IETF. Addressing these challenges requires collaborative
efforts to develop standardized solutions to ensure flexibility,
scalability, and interoperability in the telecom cloud environment.
1) Lack of definition for key interfaces and data models: Standardize
the abstraction and interface definition of cloud and network
resources for management across different vendors. Achieve
interoperability and seamless integration of cloud and network
services.
2) Lack of definition for cloud resource models: Abstract and define
cloud resources, creating standardized resource models. Establish
unified interfaces and protocol standards to achieve flexible
Zhang & Xie Expires 23 April 2025 [Page 9]
Internet-Draft net4cloud usecase October 2024
combination and scheduling of virtual network resources, enhancing
resource sharing and collaborative management between cloud and
network.
3) Lack of unified Input Output Access Management (IOAM): Extend IOAM
to enhance real-time performance monitoring in cloud-integrated
networks. Develop standardized methods to manage cross-domain nested
slices, enhancing visualization capabilities from the application
layer to the network layer.
4) Unified cloud-network orchestration mechanism: Develop a
standardized orchestration layer connecting CRD (Custom Resource
Definitions) in Kubernetes with NETCONF/YANG to improve coordination
between computing resources and network resources.
7. Security Considerations
TBD.
8. Normative References
[RFC8466] Wen, B., Fioccola, G., Ed., Xie, C., and L. Jalil, "A YANG
Data Model for Layer 2 Virtual Private Network (L2VPN)
Service Delivery", RFC 8466, DOI 10.17487/RFC8466, October
2018, <https://www.rfc-editor.org/info/rfc8466>.
[RFC8986] Filsfils, C., Ed., Camarillo, P., Ed., Leddy, J., Voyer,
D., Matsushima, S., and Z. Li, "Segment Routing over IPv6
(SRv6) Network Programming", RFC 8986,
DOI 10.17487/RFC8986, February 2021,
<https://www.rfc-editor.org/info/rfc8986>.
Authors' Addresses
Yue Zhang
China Telecom
Beiqijia Town, Changping District
Beijing
102209
China
Email: zhangy390@chinatelecom.cn
Zhang & Xie Expires 23 April 2025 [Page 10]
Internet-Draft net4cloud usecase October 2024
Chongfeng Xie
China Telecom
Beiqijia Town, Changping District
Beijing
102209
China
Email: xiechf@chinatelecom.cn
Zhang & Xie Expires 23 April 2025 [Page 11]