SRv6 Network Programming

Ushering in a New Era of IP Networks

Zhenbin Li, Zhibo Hu, Cheng Li

• Publisher: CRC Press
• Year: 2021
• ISBN:
  9781000400175
• Edition: 1
• Topic: Computer Science
• Subtopic: Computer Science General
• Index: Computer Science

SRv6 Network Programming, beginning with the challenges for Internet Protocol version 6 (IPv6) network development, describes the background, roadmap design, and implementation of Segment Routing over IPv6 (SRv6), as well as the application of this technology in traditional and emerging services.

The book begins with the development of IP technologies by focusing on the problems encountered during MPLS and IPv6 network development, giving readers insights into the problems tackled by SRv6 and the value of SRv6. It then goes on to explain SRv6 fundamentals, including SRv6 packet header design, the packet forwarding process, protocol extensions such as Interior Gateway Protocol (IGP), Border Gateway Protocol (BGP), and Path Computation Element Protocol (PCEP) extensions, and how SRv6 supports existing traffic engineering (TE), virtual private networks (VPN), and reliability requirements. Next, SRv6 network deployment is introduced, covering the evolution paths from existing networks to SRv6 networks, SRv6 network deployment processes, involved O&M technologies, and emerging 5G and cloud services supported by SRv6. Bit Index Explicit Replication IPv6 encapsulation (BIERv6), an SRv6 multicast technology, is then introduced as an important supplement to SRv6 unicast technology. The book concludes with a summary of the current status of the SRv6 industry and provides an outlook for new SRv6-based technologies.

SRv6 Network Programming: Ushering in a New Era of IP Networks collects the research results of Huawei SRv6 experts and reflects the latest development direction of SRv6. With rich, clear, practical, and easy-to-understand content, the volume is intended for network planning engineers, technical support engineers and network administrators who need a grasp of the most cutting-edge IP network technology. It is also intended for communications network researchers in scientific research institutions and universities.

Authors:

Zhenbin Li is the Chief Protocol Expert of Huawei and member of the IETF IAB, responsible for IP protocol research and standards promotion at Huawei.

Zhibo Hu is a Senior Huawei Expert in SR and IGP, responsible for SR and IGP planning and innovation.

Cheng Li is a Huawei Senior Pre-research Engineer and IP standards representative, responsible for Huawei's SRv6 research and standardization.

I

Introduction

CHAPTER 1

SRv6 Background

In this chapter, we expand on the history of Internet technology development, from the competition between Asynchronous Transfer Mode (ATM) and IP to the emergence of Multiprotocol Label Switching (MPLS), the dawning of the All IP 1.0 era and the consequent challenges, as well as the rise of Software Defined Networking (SDN). We then sum up by introducing Segment Routing over IPv6 (SRv6), which is a key technology to enable the All IP 2.0 era.

1.1 Overview of Internet Development

Humans develop in tandem, as opposed to individually, through communication and collaboration. Over our extensive history, various communications modes have emerged, ranging from beacon fires to emails, and from messenger birds to quantum entanglement. The scope of communications has also expanded from people within a certain vicinity to people in distant localities, across a whole country, around the globe, and even in outer space. That said, we as a people have never stopped our pursuit of communication technologies, which in turn have boosted human prosperity. It is, therefore, safe to say that we are no longer content with just human-to-human communication, or put differently, we now aim at achieving the connectivity of everything anytime and anywhere using the Internet. It is within this context that the development of the Internet has profoundly impacted the path of human society.

After years of development, the Internet has become almost as essential as water and electricity. Although the Internet has made life more convenient through the information age, few people know the ins and outs of its technological development history. With this in mind, we briefly summarize the history of Internet technology development in Figure 1.1.

FIGURE 1.1 Internet development milestones.

In 1969, Advanced Research Projects Agency Network (ARPANET) — the major predecessor of the Internet — came into existence.

In 1981, IPv4^[1] was defined.

In 1986, the Internet Engineering Task Force (IETF), dedicated to formulating Internet standards, was founded.

In 1995, IPv6,^[2] the next generation of IPv4, was standardized.

In 1996, MPLS^[3] was proposed.

In 2007, SDN^[4] debuted.

In 2008, the OpenFlow^[5] protocol was introduced.

In 2013, SR^[6] was proposed, including Segment Routing over MPLS (SR-MPLS)^[7] and SRv6.^[8]

In 2014, Virtual eXtensible Local Area Network (VXLAN)^[9] was released.

On November 25, 2019, IPv4 addresses were exhausted.^[10]

1.2 Start of All IP 1.0: A Complete Victory for IP

1.2.1 Competition between ATM and IP

In the initial stage of network development, multiple types of networks, such as X.25, Frame Relay (FR), ATM, and IP, coexisted to meet different service requirements. These networks could not interwork with each other, and also competed, with mainly ATM and IP networks taking center stage.

ATM is a transmission mode that uses fixed-length cell switching. It establishes paths in connection-oriented mode and can provide better Quality of Service (QoS) capabilities than IP. Its design philosophy involves centering on networks and providing reliable transmission, and its design concepts reflect the reliability and manageability requirements of telecommunications networks. This is the reason why ATM was widely deployed on early telecommunications networks.

The design concepts of IP differ greatly from those of ATM. To be more precise, IP is a connectionless communication mechanism that provides the best-effort forwarding capability, and the packet length is not fixed. On top of that, IP networks mainly rely on the transport-layer protocols (e.g., TCP) to ensure transmission reliability, and the requirement for the network layer involves ease of use. To add on to this, the design concept of IP networks embodies the “terminal-centric and best-effort” notion of the computer network. We can therefore say that IP is widely used on computer networks because it meets the corresponding service requirements.

The competition between ATM and IP networks can essentially be represented as a competition between telecommunications and computer networks. In other words, telecommunications practitioners sought to use ATM for network interconnection to protect network investments. On the flip side, computer practitioners aimed at using ATM as only a link-layer technology to provide QoS guarantee for IP networks, while setting aside the task of establishing network connections for IP.

Computer networks subsequently evolved toward broadband, intelligence, and integration, with mainly burst services. Despite this, the QoS requirements that traffic places on computer networks are not as high as those on telecommunications networks, and the length of packets is not fixed. As such, the advantages of ATM — fixed-length cell switching and good QoS capabilities — cannot be brought into full play on computer networks. Not only that, the QoS capabilities of ATM are based on connection-oriented control with a certain packet header overhead. Therefore, ATM is inefficient in carrying computer network traffic, and it yields high transmission and switching costs.

To sum up, as network scale expanded and network services increased in number, ATM networks became more complex than IP networks, while also bearing higher management costs. Within the context of costs versus benefits, ATM networks exited the arena as they were gradually replaced by IP networks.

1.2.2 MPLS: The Key to All IP 1.0

Although with relation to the development of computer networks, the IP network is more fitting than the ATM network, a certain level of QoS guarantee is still required. To compensate for the IP network’s insufficient QoS capabilities, numerous technologies integrating IP and ATM, such as Local Area Network Emulation (LANE), IP over ATM (IPoA),^[11] and tag switching,^[12] have been proposed. However, these technologies only addressed part of the issue, until 1996 when MPLS technology was proposed^[3] to provide a better solution to this issue.

MPLS is considered as a Layer 2.5 technology that runs between Layer 2 and Layer 3. It supports multiple network-layer protocols, such as IPv4 and IPv6, and is compatible with multiple link-layer technologies, such as ATM and Ethernet. Some of its other highlights include the fact that it incorporates ATM’s Virtual Channel Identifier (VCI) and Virtual Path Identifier (VPI) switching concepts, combines the flexibility of IP routing and simplicity of label switching, and adds connection-oriented attributes to connectionless IP networks. By establishing virtual connections, MPLS provides better QoS capabilities for IP networks.

However, this is not the only reason why it was initially proposed. Point in case being that MPLS also forwards data based on the switching of fixed-length 32-bit labels, and therefore it features a higher forwarding efficiency than IP, which forwards data based on the Longest Prefix Match (LPM). That said, as hardware capabilities have and continue to improve, MPLS no longer features distinct advantages in forwarding efficiency. Nevertheless, MPLS provides a good QoS guarantee for IP through connection-oriented label forwarding and also supports Traffic Engineering (TE), Virtual Private Network (VPN), and Fast Reroute (FRR).^[13] These advantages play a key role in the continuous expansion of IP networks, while also catapulting the IP transformation of telecom networks.

In general, the success of MPLS depends mainly on its three important features: TE, VPN, and FRR.

• TE: Based on Resource Reservation Protocol-Traffic Engineering (RSVP-TE),^[14] MPLS labels can be allocated and distributed along the MPLS TE path, and TE features (such as resource guarantee and explicit path forwarding) can be implemented. This overcomes IP networks’ lack of support for TE.
• VPN: MPLS labels can be used to identify VPNs^[15] for isolation of VPN services. As one of the major application scenarios of MPLS, VPN is a key technology for enterprise interconnection and multiservice transport as well as an important revenue source for carriers.
• FRR: The IP network cannot provide complete FRR protection, which in turn means that it is unable to meet the high-reliability requirements of carrier-grade services. MPLS improves the FRR capabilities of IP networks and supports 50 ms carrier-grade protection switching in most failure scenarios.

Because IP networks are cost-effective and MPLS provides good TE, VPN, and FRR capabilities, IP/MPLS networks gradually replaced dedicated networks, such as ATM, FR, and X.25. Ultimately, MPLS was applied to various networks, including IP backbone, metro, and mobile transport, to support multiservice transport and implement the Internet’s All IP transformation. In this book, we refer to the IP/MPLS multiservice transport era as the All IP 1.0 era.

1.3 Challenges Facing All IP 1.0: IP/MPLS Dilemma

Although IP/MPLS drove networks into the All IP 1.0 era, the IPv4 and MPLS combination has also set forth numerous challenges, which are becoming more prominent as network scale expands and cloud services develop, and are thereby hindering the further development of networks.

1.3.1 MPLS Dilemma

From one perspective, MPLS plays an important role in All IP transport, while from another perspective, it complicates inter-domain network interconnection by causing isolated network islands.

To put it more precisely, consider the fact that on the one hand, MPLS is deployed in different network domains, such as IP backbone, metro, and mobile transport networks, forming independent MPLS domains and creating new network boundaries. However, many services require E2E deployment, and this means that services need to be deployed across multiple MPLS domains, which in turn results in complex inter-domain MPLS solutions. In that regard, multiple inter-Autonomous System (AS) solutions, such as Option A, Option B, and Option C,^[15,16] have been proposed for inter-AS MPLS VPN, and each one involves relatively complex service deployment.

On the other hand, as the Internet and cloud computing develop, more and more cloud data centers are built. To meet the requirements of multi-tenant networking, multiple overlay technologies were proposed, among which VXLAN is a typical example. At the same time, quite a few attempts were made to provide VPN services by introducing MPLS to data centers. However, these attempts all wound up in failure due to multiple factors, including numerous network boundaries, complex management, and insufficient scalability.

In Figure 1.2, traffic from end...