CN116941230A

CN116941230A - Invoking SRv segment identifier and differential segment of network behavior including implementation of network slice

Info

Publication number: CN116941230A
Application number: CN202280019642.9A
Authority: CN
Inventors: 拉克什·冈赫; 克拉伦斯·菲尔斯菲斯; 扎法尔·阿里; 帕特里斯·布里塞特
Original assignee: Cisco Technology Inc
Current assignee: Cisco Technology Inc
Priority date: 2021-03-07
Filing date: 2022-02-24
Publication date: 2023-10-24

Abstract

In one embodiment, a segment routing internet protocol version 6 (SRv) micro segment ("uSID") is included in a destination address of a packet transmitted over a network and possibly in other segment identifiers ("SIDs") and invokes corresponding network behavior, including but not limited to implementing a corresponding network slice. In one embodiment, the network node is configured to perform a differential network slice implementation function based on slice representative values provided by the global and/or local uSIDs of the packet. The configuration may be defined by a controller and/or routing protocol advertisement in the network. In response to the received packet, the network node identifies and performs a corresponding network slice implementation function based on the slice representative values provided by the one or more global and/or local uSIDs of the destination address of the received packet. Various encodings within the IPv6 destination address of the encapsulated packet are disclosed.

Description

Invoking SRv segment identifier and differential segment of network behavior including implementation of network slice

Cross Reference to Related Applications

The present application claims the benefit and priority of U.S. non-provisional patent application No. 17/453,353, filed on 3 at 11, 2021, which claims the benefit of U.S. provisional patent application No. 63/157,810, filed on 7 at 3, 2021, each of which is incorporated herein by reference in its entirety.

Technical Field

The present disclosure relates generally to control plane provisioning and data plane transmission and processing of packets in a piecewise routed network, including invoking corresponding network behavior, including, but not limited to, implementing piecewise routed segments and/or differential segments based on corresponding network slices.

Background

The communications industry is rapidly changing to accommodate emerging technologies and ever increasing customer needs. This customer demand for increased performance of new applications and existing applications drives communication network and system providers to employ networks and systems with higher speeds and capacities (e.g., greater bandwidth). In attempting to achieve these goals, a common approach taken by many communication providers is to use packet-switched technology. Packets are typically forwarded in the network based on one or more values representing network nodes or paths.

Drawings

The features of one or more embodiments are set forth with particularity in the appended claims. The embodiments and their advantages may be understood from the following detailed description taken in conjunction with the accompanying drawings, in which:

fig. 1A illustrates a packet switching device according to one embodiment;

FIG. 1B illustrates an apparatus or component thereof according to one embodiment;

FIG. 2A illustrates a network slice map of controller programming or node advertising in accordance with one embodiment;

FIG. 2B illustrates link bandwidth network splitting according to one embodiment;

FIG. 2C illustrates a segment routed packet according to one embodiment;

FIG. 2D illustrates a segment routing identifier (SID) and a micro segment (uSID) according to one embodiment;

FIG. 3A illustrates a process according to one embodiment;

FIG. 3B illustrates a process according to one embodiment;

FIG. 4 illustrates a network operating in accordance with one embodiment;

FIG. 5A illustrates a network operating in accordance with one embodiment;

FIG. 5B illustrates a network operating in accordance with one embodiment;

FIG. 5C illustrates a network operating in accordance with one embodiment; and

fig. 5D illustrates a network operating in accordance with one embodiment.

Detailed Description

1. Summary of the invention

Aspects of the invention are set out in the independent claims, with preferred features set out in the dependent claims. Features of one aspect may be applied to various aspects alone or in combination with other aspects.

Methods, apparatus, computer storage media, mechanisms, means, and so forth associated with control plane provisioning and data plane provisioning, transmission, and processing of packets in a network are disclosed, wherein a segment routing internet protocol version 6 (SRv) identifier (SID) and/or micro-segment (uSID) is included in the packet, thereby causing invocation of a corresponding identified network behavior, including, but not limited to, implementing a corresponding network slice. In one embodiment, such network behavior includes differential processing by network nodes specified in the associated different SIDs and/or usids.

One embodiment includes a method. A particular segment routing node in the network receives a segment routing version 6 (SRv) packet that includes an internet protocol version 6 (IPv 6) destination address as an address of the particular segment routing node, wherein the IPv6 destination address includes a plurality of micro segments (uads), a particular uad of the plurality of uads mapped to a particular network slicing behavior of a plurality of different network slicing behaviors performed by the particular segment routing node on different packets. A particular segment routing node processing the received SRv comprising differentially processing a particular SRv packet according to a particular network slicing behavior, wherein the differential processing differs from processing according to one of a plurality of different network slicing behaviors that is not the particular network slicing behavior; updating an IPv6 destination address comprising removing routing usids identifying particular segment routing nodes and shifting one or more usids in the IPv6 destination address of the received SRv packet into higher order bit positions; and transmitting SRv packets from the particular segment routing node with the updated IPv6 destination address.

One embodiment performs a lookup operation in the network slice mapping data structure that results in the identification of a particular network slice behavior based on a particular uSID.

In one embodiment, the particular uSID is a per-hop behavior (PHB) and routing uSID that indirectly identifies a combination of particular network slice behaviors and is part of an advertised route for a particular segment routing node. In one embodiment, the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the particular uSID concatenated with the first routing uSID is part of an advertised route of the particular segmented routing node; and wherein the removing the routing uSID comprises removing the first routing uSID. In one embodiment, the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the first route uSID is part of an advertised route for a particular segment routing node; and wherein the removing the routing uSID comprises removing the first routing uSID. In one embodiment, the particular uSID is in a lower order bit in the IPv6 destination address than all of the routing uSIDs in the IPv6 destination address of the received SRv packet.

In one embodiment, the IPv6 destination address of the received SRv packet includes a plurality of pairs of per-hop behaviors (PHBs) and routing usids; wherein the plurality of pairs includes a particular pair including a particular uSID and a routing uSID; and wherein the removing the routing uSID comprises removing a particular pairing. In one embodiment, the particular pairing is part of an advertised route of the particular segmented routing node. In one embodiment, the specific pairing includes a specific uSID concatenated with the routing uSID. In one embodiment, the particular pairing includes a routing uSID concatenated with a routing uSID. In one embodiment, the route uSID is part of an advertised route for a particular segment routing node.

In one embodiment, the particular uSID is a global per-hop behavior (PHB) uSID that identifies a PHB to be performed on one or more other segment routing nodes in the network. In one embodiment, a particular uSID is part of an advertised route for a particular segment routing node. In one embodiment, the IPv6 destination address of the received SRv packet includes Flex-Algo uSID; and the processing of the packet includes processing according to a flexible algorithm identified by the Flex-Algo uSID.

In one embodiment, prior to receiving SRv packets, a particular segment routing node configures one or more hardware resources to perform differential processing for each of a plurality of different network slice behaviors. In one embodiment, each of the plurality of network slice behaviors defines a packet processing delay or a link bandwidth capacity. In one embodiment, the one or more resources include queues, ternary Content Addressable Memory (TCAM), and/or memory.

2. Description of the invention

Methods, apparatus, computer storage media, mechanisms, means, and so forth associated with control plane provisioning and data plane provisioning, transmission, and processing of packets in a network are disclosed, wherein a segment routing internet protocol version 6 (SRv) identifier (SID) and/or a micro segment (uSID) is included in an internet protocol version 6 (IPv 6) destination address, and possibly in a segment list of segment routing headers. In one embodiment, these uSIDs and SIDs identify different forwarding and processing information that causes the network node to invoke network behavior of the corresponding identification, etc., including, but not limited to, implementing the corresponding network slice. In one embodiment, such network behavior includes differential processing by network nodes specified in associated different uSIDs and/or SIDs.

Embodiments disclosed herein are generally described using the uSID terminology and corresponding processing (e.g., consistent with SRv network programming, IPv6 Segment Routing Header (SRH), and compressed SRv segment list encoding in SRH). As used herein, the terms micro-segment, micro-segment identifier, micro-SID, and uSID are used interchangeably to refer to embodiments of compressed SIDs (also referred to as compact SIDs). The teachings provided herein regarding the uSIDs are applicable to embodiments that use other forms of compressed SIDs and/or compact forwarding identifiers.

As used herein, the terms SRv segment identifier, SRv6 SID, SRv6 segment, segment identifier, SID, and segment are used interchangeably to refer to 128-bit values (e.g., IPv6 addresses), which may or may not include a ussid. When one of these terms is defined by "uSID" (or the like), SRv SID (e.g., uSID container, 128-bit value, IPv6 address) includes one or more uSIDs. The uSID container is sometimes referred to as a uSID carrier.

The terms "node" and "network node" are used herein to refer to a router or host. The term "route" is used herein to refer to a fully or partially extended prefix/route (e.g., 10.0.0.1 or 10.0 for IPv 4.) that is different from a "path" through the network that refers to a next hop (e.g., a next router) or a full path (e.g., through router a then router B, etc.). Also, the term "prefix" without qualifiers is used herein to refer to a prefix that is fully or partially extended. The use of an ellipsis (".") identifies that an item may include additional values. The term "concatenate" means to concatenate sequentially in the order identified (e.g., "a" concatenated with "B" means "AB" -not "BA").

As used herein, "forwarding information" includes, but is not limited to, information describing how to process (e.g., forward, send, manipulate, modify, change, discard, copy, receive) the corresponding packet. In one embodiment, determining forwarding information is performed via one or more lookup operations (e.g., an ingress lookup operation, an egress lookup operation). Also, the term "processing" when referring to the processing of a packet procedure refers to a broad range of operations performed in response to the packet, such as, but not limited to, forwarding/sending, dropping, manipulating/modifying/changing, receiving, copying, creating, intercepting, consuming, policing, quality of service processing, applying one or more services or application functions (e.g., updating network configuration, forwarding, network, management, operation/management/and/or other information) to the packet or packet switching device, and the like. Also, as used herein, the term "parallel" processing is used in the general sense that at least a portion of two or more operations are performed overlapping in time. The term "interface" as broadly used herein includes interface infrastructure (e.g., buffers, storage locations, forwarding and/or other data structures, processing instructions) used by a network node in performing packet-related processing. Further, as used herein, a "virtual interface" is an interface that is not directly connected to an external cable or optical cable (e.g., a termination interface to a cable) or other communication mechanism, as opposed to a "physical interface".

As described herein, an embodiment includes various elements and limitations, with no one element or limitation contemplated as being a critical element or limitation. Each claim individually recites an aspect of the embodiment in its entirety. Moreover, one or more embodiments described include, but are not limited to, systems, networks, integrated circuit chips, embedded processors, ASICs, other hardware components, methods, and computer-readable media containing instructions, etc. In one embodiment, one or more systems, devices, components, etc. comprise the embodiment, which may include some elements or limitations of a claim being performed by the same or different systems, devices, components, etc. when compared to a different embodiment. In one embodiment, the processing elements include general purpose processors, task specific processors, ASICs with one or more processing cores, and/or any other co-located resource sharing implementations for performing corresponding processing. The embodiments described below embody various aspects and configurations, with the figures illustrating exemplary and non-limiting configurations. Computer-readable media and apparatus (e.g., processors and memory or other devices configured to perform these operations) for performing the method and process block operations are disclosed and are consistent with the extensible scope of embodiments. The term "apparatus" is used herein consistent with its common definition of appliance or device.

The processing of steps, connections, and signals and information illustrated in the figures, including but not limited to any block diagrams and flowcharts, and message sequence charts, is typically performed in the same or a different serial or parallel order and/or by different components and/or processes, threads, etc. and/or over different connections, and in other embodiments in combination with other functions, unless this renders the embodiments ineffective or explicitly or implicitly requires a sequence (e.g., for a sequence of read values, the read values must be processed-the values must be obtained before the read operation, but some associated processing is performed before, concurrently with, and/or after the read operation). Moreover, nothing described or referenced in this document is admitted as prior art to the present application unless explicitly so stated.

The term "one embodiment" is used herein to refer to a particular embodiment, wherein each reference to "one embodiment" may refer to a different embodiment. The term "one embodiment" is used repeatedly herein to describe associated features, elements and/or limitations included in one or more embodiments, but does not establish a cumulative set of associated features, elements and/or limitations that each and every embodiment must include. However, an embodiment may include all of these features, elements, and/or limitations. In addition, the terms "first," "second," etc., as well as "particular" and "specific" are used herein to denote different units (e.g., a first widget or operation, a second widget or operation, a particular widget or operation, a specific widget or operation). The use of these terms herein does not connote an ordering such as one unit, operation, or event occurring or coming before another or another characterization, but rather provides a mechanism to distinguish between the unit of elements. Moreover, the phrase "x-based," "responsive to x" is used to indicate a minimum set of items "x" from which something is derived or caused, where "x" is extensible and does not necessarily describe a complete list of items on which an operation is performed. Additionally, the phrase "coupled to" or "communicatively coupled to" is used to indicate some level of direct or indirect connection between elements and/or devices with or without modification of the coupled signals or communicated information by one or more coupling devices. Moreover, the term "or" is used herein to identify a selection of one or more (including all) of the connection items. In addition, the transitional term "comprising" synonymous with "comprising", "containing" or "characterized by" is inclusive/open-ended and does not exclude additional, unrecited elements, method steps, or the like. Finally, the term "particular machine" when recited in the method claims for performing steps refers to a particular machine within the 35USC ≡101 machine legal category.

Segment routing internet protocol version 6 (SRv 6) network programming enables creation of overlays with underlying optimizations to be deployed in the Segment Routing (SR) domain. Ingress edge SRv network nodes typically encapsulate received packets with an external internet protocol version 6 (IPv 6) header and optionally one or more Segment Routing Headers (SRH). In one embodiment, the destination address in the external IPv6 header includes a plurality of usids including global and/or local usids (e.g., for routing packets and defining processing behavior of the packets).

Network slicing provides the ability to divide a physical network into multiple logical networks of different sizes, structures, and functions so that individual slices can be dedicated to a particular service or customer. Network slicing requires parallel operation while providing slice flexibility in terms of network resource allocation. The implementation of network slicing (e.g., a description of its definition) is an implementation in the underlying network layer. The implementation of network slicing is typically determined by the service requirements and the available capabilities of the underlying, typically shared, infrastructure.

Disclosed herein are various techniques for encoding a network slice in an IPv6 destination address (and possibly in a SID in a segment list in an SRH), and then transmitting and/or processing packets according to the encoded network slice (directly or indirectly). One embodiment includes a network slice value in one or more uSIDs or SIDs representing a corresponding network slice. In one embodiment, the network slice value is a "slice identifier" (SLID), such as an 8-bit value that uniquely identifies a particular slice in the SR domain. In one embodiment, the network node converts the received uSID or SID into a corresponding mapped network slice value. In one embodiment, the network node converts the received network slice values into corresponding mapping uSIDs or SID Per Hop Behaviors (PHBs). In one embodiment, the network slice values are used to identify corresponding network slice Per Hop Behaviors (PHBs) in one or more data structures. In one embodiment, the network node converts the received PHB identifier into a corresponding mapped network slice value. In one embodiment, the network node converts the received PHB identifier into a corresponding aggregated network slice value. In one embodiment, the network node converts the received PHB identifier into a local hardware resource identifier on the node.

In one embodiment, the network node is configured to perform a differential network slice implementation function based on slice representative values provided by the global and/or local uSIDs of the packet. In response to the received packet, the network node identifies and performs a corresponding network slice implementation function based on the slice representative values provided by the one or more global and/or local uSIDs of the destination address of the received packet.

The network split is used for different usage scenarios, subscriber services and customer categories. In one embodiment, for 5G, 6G, and/or other networks, the service provider uses network slicing techniques to deliver ultra-reliable low latency communication (URLLC) services, such as for, but not limited to, telemedicine, online gaming, autonomously connected automobiles, and many other mission critical applications. To provide these guaranteed services and implement the required Service Level Agreements (SLAs), in one embodiment, network resources and network functions are provisioned to ensure that there is no degradation (or minimizes degradation) due to congestion, failures, maintenance, and other problems. In one embodiment, security and privacy guarantees are provided by these services. In one embodiment, the processing and storage of data in the network is a function of the identified network slice.

In one embodiment, the network split is basically an end-to-end division of network resources and network functions such that selected applications/services/connections run isolated from each other for a particular business purpose. In a general sense, a network slice refers to an overlay infrastructure that provides a particular network service according to particular attributes, goals, and constraints. In one embodiment, the infrastructure includes every aspect of the network architecture, including the radio, transport network, mobile core infrastructure, and orchestration infrastructure required to manage and operate the slices.

In one embodiment, the network slices define connectivity resource requirements and associated network behavior such as, but not limited to, bandwidth allocation, latency, jitter, packet loss, availability, security, privacy, hardware queue allocation, deterministic schedulers, hardware and software resource partitioning, and network service functions with other resource behaviors such as computing and storage availability. Each network slice is associated with a set of characteristics and behaviors that separate one type and/or group of flows from another type and/or group of flows of user traffic. Network slicing depends on how data packets used by the network slicing are handled on the nodes per-hop behavior. The nodes may use different behaviors to provide certain guarantees, such as bandwidth guarantees or latency bound guarantees. The guarantees are then used by the service provider to provide service level guarantees (SLAs) for the services provided by the corresponding network slices.

The network cut associated with the network infrastructure typically includes at least some of the following requirements.

Transport slice management, including the ability to create, modify, and delete complete network slices (including any actions required by the transport layer itself). It typically includes per-slice operation, administration and maintenance (OAM) capabilities to allow slice application owners and operators to monitor the health and performance of the slices.

Resource reservation, including the ability to reserve transmission resources for a transmission slice.

Slice isolation, including any single transport slice should be isolated from other transport slices at the appropriate performance, operational, security, and reliability levels specified by the application or end user's service policies.

Abstracting, including the ability to model and build transport infrastructure appropriate to the slicing requirements with resources.

Network slices are sometimes characterized as "hard" or "soft" slices based on the level of resource sharing between different slices. In both cases they generally fulfill the above requirements and/or characteristics. Network slices that have resources dedicated to them and that are not shared with other slices are considered "hard" slices. For example, the transport layer portion of a network slice may have bandwidth dedicated to it. In contrast, "soft" slice resources may be shared between slices while maintaining appropriate Service Layer Agreements (SLAs) and/or other requirements, and returning resources to the network when resources for other uses are no longer needed.

Within the core transport network, segment routing provides a means to share resources in conjunction with DiffServ QoS to create soft slices using shortest path routing and statistical multiplexing. Differentiated services (Diffserv) models allow multiple services to be carried on top of a single physical network by relying on compliance nodes to apply specific forwarding treatments (scheduling and drop policies) to packets carrying corresponding Diffserv code points. DiffServ, however, cannot discern and differentially process the same type of traffic (e.g., voIP traffic) from different tenants with different SLA requirements, or otherwise perform traffic isolation. In one embodiment, slice identification is independent of the topology of the network and QoS/DiffServ policies, thereby enabling scalable network slicing for SRv coverage. In one embodiment, each network slice in the SR domain is uniquely identifiable based on the various techniques and encodings described herein.

In one embodiment, to create hard slices, traffic engineered segment routing policies are constructed using a distributed technique such as Flex-Algo or a centralized technique such as using a segment routing path computation engine (SR-PCE) that provides resources that are entirely dedicated to a particular transport slice.

Fig. 1A-1B and the discussion thereof herein are intended to provide a description of various exemplary packet-switched systems used in processing packets in accordance with one embodiment. One embodiment processes packets in a network based on a SID, a uSID, a compressed SID, and/or a compact forwarding identifier typically represented in a single 128-bit IPv6 address (e.g., the destination address of the packet's external IPv6 header). In one embodiment, the additional SID, the uSID, the compressed SID, and/or the compact forwarding identifier are included in one or more fragment lists of the SRH.

One embodiment of a packet switching device 100 is illustrated in fig. 1A. As shown, packet switching device 100 includes a plurality of line cards 101 and 105, each having one or more network interfaces for sending and receiving packets over a communication link and having one or more processing elements that, in one embodiment, are used in association with SRv SID and uSID, SRv SID and uSID calls include network behavior that implements network slicing, typically represented by a single 128-bit IPv6 address, such as, but not limited to, the destination address of the packet's external IPv6 header.

The packet switching device 100 also has a control plane with one or more processing elements 102 for managing the control plane and/or control plane processing of packets. The packet switching device 100 also includes other cards 104 (e.g., service cards, blades) that include processing elements for processing (e.g., forwarding/transmitting, dropping, manipulating, changing, modifying, receiving, creating, replicating, applying services) packets based on SID, uSID, compressed SID, and/or compact forwarding identifier in one embodiment; and some hardware-based communication mechanisms 103 (e.g., buses, switching fabrics, and/or matrices, etc.) for allowing different entities 101, 102, 104, and 105 thereof to communicate.

Line cards 101 and 105 typically perform actions as both ingress and egress line cards with respect to a number of other specific packets and/or packet flows received by or transmitted from packet switching device 100.

Fig. 1B is a block diagram of an apparatus 120 used in one embodiment. In one embodiment, the apparatus 120 performs one or more processes or portions thereof corresponding to one flow chart illustrated or otherwise described herein and/or illustrated or otherwise described in another figure.

In one embodiment, the apparatus 120 includes one or more processors 121 (typically with on-chip memory), memory 122, storage devices 123, dedicated components 125 (e.g., such as optimized hardware for performing lookup and/or packet processing operations, associated memory, binary and/or ternary content addressable memory, etc.), and interfaces 127 for communicating information (e.g., sending and receiving packets, user interfaces, displaying information, etc.), which are typically communicatively coupled via one or more communication mechanisms 129 (e.g., buses, links, switching fabrics, matrices), wherein the communication paths are typically tailored to meet the needs of a particular application.

Various embodiments of the apparatus 120 may include more or fewer elements. The operation of apparatus 120 is typically controlled by processor 121 using memory 122 and storage device 123 to perform one or more tasks or processes. Memory 122 is one type of computer-readable/computer storage medium and typically includes Random Access Memory (RAM), read Only Memory (ROM), flash memory, integrated circuits, and/or other storage elements. Memory 122 typically stores computer-executable instructions to be executed by processor 121 and/or data which is manipulated by processor 121 for implementing functionality in accordance with an embodiment. Storage 123 is another type of computer-readable medium and typically includes solid state storage media, magnetic disk drives, floppy disks, networking services, tape drives, and other storage devices. Storage devices 123 typically store computer-executable instructions to be executed by processor 121 and/or data which is manipulated by processor 121 for implementing functionality in accordance with an embodiment.

One embodiment implements end-to-end network slicing using per-hop forwarding behavior of data packets defined by SRv6 uSID instructions and slice identifiers and their associated slice profiles for individual nodes and/or links along a packet traversal path to ensure consistent processing of network ranges for the individual end-to-end network slices. In one embodiment, the per-hop behavior SRv uSID instruction and slice identifier are assigned by the controller. In one embodiment, the per-hop behavior SRv uSID instruction and slice identifier are advertised (e.g., flooded) by nodes of the network via a routing protocol (e.g., interior gateway routing protocol "IGRP").

The data packet carries a per-hop behavior SRv uSID instruction that is used by the forwarding plane (e.g., network nodes along the traversal path) to provide corresponding processing in the data plane. These SRv uSID instructions provide the forwarding state of packets consistent with the segmented routing architecture.

In one embodiment, the packet does not include a slice ID, but instead maps the per-hop behavior SRv u sid instruction to a corresponding specific end-to-end network slice (e.g., using a level of indirection, which can help improve the scale of the number of network slices, as many slices can be mapped to one per-hop behavior identifier on a node). In one embodiment, the packet processing includes, but is not limited to, using low latency queuing, rate limiting, security level, privacy, storage functions, service function chains (e.g., firewalls), field OAM for transmission attestation, path tracing, service provisioning, fast reroute protection, reliability, deterministic scheduling for time sensitive slices, and/or physical/logical isolation and resource partitioning. In one embodiment, the desired per-hop behavior is augmented by quality of service processing based on a traffic class field that typically includes a Differentiated Services Code Point (DSCP) field.

In one embodiment, the desired per-hop behavior is augmented by an Interior Gateway Protocol (IGP) flexible algorithm (Flex-Algo). Flex-Algo is typically used to route data packets on the minimum IGP cost or lowest latency path in a network. Flex-Algo is generally not suitable for providing PHB processing that can map a large number of network slices (e.g., one thousand). Network slices may use certain Flex-Algo paths for data packets. Thus, many network slices can be mapped to certain Flex-Algo. Within Flex-Algo, packets carry different PHB instructions to provide the different PHBs required for network slicing.

In one embodiment, packets of the same flow use the same PHB uSID instruction when forwarding the packets over the network; thus, multi-path (ECMP) routing is uniformly handled according to equal cost, especially when ECMP hashing based on triplets of IPv6 source address, IPv6 destination address, and flow label is used.

In one embodiment, the mapping SLID of the per-hop behavior SRv uSID instruction is used by a network controller (e.g., a Software Defined Network (SDN) controller) to provide an end-to-end SLA and mapped to a network service. In one embodiment, the mapping SLID of the per-hop behavior SRv uSID instruction is used for Constrained Shortest Path First (CSPF) path computation per slice, performance delay/loss measurement, survival monitoring, OAM (in general), and the like.

In one embodiment, per-hop behavior SRv instructions (and their corresponding slice identifiers) are used to "stitch" SRv/IPv 6 networks with MPLS (or other) networks to provide end-to-end multi-domain network slices. Upon detecting a problem on a network node, packet traffic is automatically diverted in the segment routing data plane to provide multi-domain matching per-hop behavior in an MPLS (or other) network. One embodiment performs multi-domain "stitching" from an MPLS (or other) network to a SRv/IPv 6 network. In one embodiment, an IPv6 provider edge router (6 VPE) maps its IPv6 per-hop behavior to MPLS core per-hop behavior and vice versa.

In one embodiment, each hop SRv uSID instruction (e.g., uSID) is allocated on each hop along each packet path for data plane processing of the packet to implement an end-to-end network slice and its associated slice identifier (SLID). The nodes use the slice profile to configure and provide resource allocation, scheduling/queuing, security, privacy, quarantine, and other packet forwarding processing specific to the individual network slices in the data plane. One or more uSIDs may be mapped to the same SLID that provides scalability. In one embodiment, the forwarding fast path of the network node implements only a portion of the different behaviors (e.g., ten different behaviors in the network that support thousands of different network slices managed by the network controller in one embodiment). In one embodiment, indirection (e.g., mapping) of PHB instructions (e.g., a uSID) to network slices implemented on network nodes provides scalability because the processing resources required to support the desired PHBs are reduced by one or more orders of magnitude. In one embodiment, the aggregation of network slices is represented as PHB identifiers as an aggregation identifier. In one embodiment, a network slice is used as a resource partition on a node, where PHB on the node identifies the partitioned resource of the network slice.

In one embodiment, a network slice Per Hop Behavior (PHB) is determined based on a slice profile. These slice profiles are typically configured with identification slice IDs (SLIDs), names and characteristics (such as bandwidth, latency, queues, priority, ACLs, etc.) on the network nodes. These characteristics identify the PHB required to implement a network slice and packet traffic processing on the node for the corresponding network slice.

Methods are generally described in this document as using a per-hop behavior identifier or a slice identifier in the IPv6 destination address field. However, one embodiment includes such an identifier in another field, such as but not limited to an IPv6 hop-by-hop option, an IPv6 end-to-end option, a Type Length Value (TLV) in SRH, or another field in an IPv6 header or SRH.

In one embodiment, the IPv6 destination address is copied from a segmented list of SRHs of the packet. In this case, the per-hop behavior identifier or slice identifier instruction is carried in the segment list of the SRH. In one embodiment, the node that handles the external IPv6 header also copies the next-hop address and per-hop behavior instructions from the segment list of the SRH. In one embodiment, per-hop behavior is maintained in an IPv6 destination address that is modified, where next hop routing instructions are copied from the segment list of the SRH.

In one embodiment, the node derives the next hop destination address from an IPv6 extension header carrying this next hop information. In one embodiment, the PHB identifier or slice identifier is also carried in the IPv6 extension header with the next hop address information.

In one embodiment, the PHB instruction carries an optional timestamp value (e.g., thirty-two bits encoded similar to the PHB instruction, either just before or just after the PHB uad instruction in the IPv6 destination address field, or at the beginning or end of the IPv6 destination address) that directly or indirectly identifies the deadline or lifetime of the packet in the network. In one embodiment, if the deadline or lifetime is exceeded, the packet is handled differently by the intermediate or egress node (e.g., dropped/dropped by a per-hop behavior on the node). In one embodiment, such behavior is particularly useful and/or desirable in Time Sensitive Networks (TSNs) where the network service cannot tolerate excessive latency.

FIG. 2A illustrates a controller programmed or node advertised PHB based on a uSID to SLID mapping (200) in accordance with one embodiment. As shown, each mapping 201-203 includes a uSID (e.g., uSID10- > SLID 1, uSID15- > SLID 2, and uSID25- > SLID 3) mapped to a SLID, where each SLID has a corresponding name (for facilitating user identification).

In one embodiment, each network slice has resource allocation and/or guarantees for the network node, such as, but not limited to: link bandwidth capacity, hardware resources (e.g., queues, content addressable memory entries, memory), hardware queue mapping, network processing unit or co-processor mapping, forwarding table mapping, BCDL (bulk code download) route update priority, TI-LFA (topology independent-loop free replacement) fast reroute protection, scheduling, service function chaining, field operation, administration and maintenance (oam) behavior (such as time stamping and logging of interface/node identifiers), security, reliability, time sensitivity (e.g., considered deterministic scheduling in one embodiment), isolation, partitioning, and/or quality of service (QoS) applications.

In one embodiment, for corresponding per-hop behavior, an end-to-end specific network function is enabled. In one embodiment, these network functions include performance monitoring (e.g., delay or packet loss, throughput), OAM functions, and/or Service Function Chains (SFCs). In one embodiment, end-to-end performance is measured for network slices and services provided by routing one or more performance measurement probe query packets using the same PHB instructions as those used for customer data traffic. In one embodiment, the identified performance measurement attributes are collected and published to a network controller (e.g., by a provider edge node). By providing end-to-end performance measurements of network slices, the use of a uSID in implementing PHBs along a particular path through the network provides advantages over existing networks.

Fig. 2B illustrates a bandwidth network slice map 210 of a link 219 according to one embodiment. As shown, each mapping 211-213 includes a ussd (e.g., ussid 10- > spid 1, ussid 15- > spid 2, and ussid 25- > spid 3) mapped to a spid, where each spid has a corresponding name (for facilitating user identification), a slice latency attribute (e.g., low latency, high bandwidth, best effort), and an allocated bandwidth (e.g., twenty percent, sixty percent).

In one embodiment, a routing protocol (e.g., IGP, BGP-LS) is used to advertise (e.g., flood) network SLIDs for each link. The ingress network node or controller performs path computation on each network slice using the SLID of the respective link. In one embodiment, when assigned by each network node, the corresponding uSID is advertised with the SLID of each link.

Fig. 2C illustrates SRv packet 220 including a uSID when implementing a PHB along a particular path through a network, according to one embodiment. As shown, the external IPv6 header 221 includes an IPv6 destination address that includes per-hop behavior (PHB) upid instructions, an optional Segment Routing Header (SRH) 222, and an encapsulated packet 226 that is transmitted over the network according to PHB based on the upid and/or SID included in the external IPv6 header 221 and optional SRH 222 (if present).

In one embodiment, the IPv6 destination address in the external IPv6 header 221 includes all the usids required to define the PHBs that implement end-to-end network slicing along the desired forwarding path. Thus, in one embodiment, SRv packet 220 does not include optional Segment Routing Header (SRH) 222. In one embodiment, SRv packet 220 includes an optional Segment Routing Header (SRH) 222 with an IPv6 destination address in the segment list and/or carrying other information in a field of SRH 222.

In one embodiment, the IPv6 destination address in the external IPv6 header 221 does not include all the usids required to define the PHBs that implement end-to-end network slicing along the desired forwarding path. Thus, in one embodiment, SRv packet 220 includes an optional Segment Routing Header (SRH) 222 that includes one or more SIDs including one or more usids that cumulatively define the PHBs that implement end-to-end network slices along the desired forwarding path (e.g., in addition to the usids in the IPv6 destination address). In one embodiment, when the IPv6 destination address does not include the next uv-id (e.g., all uv-ids after a valid uv-id are E-o-C uv-ids), the segment routing process of network node pair SRv6 packet 200 includes updating the IPv6 destination address in the outer IPv6 header 220 of SRv6 packet 200 to the address of the next SID in the segment list of SRH 222. In one embodiment, the optional Segment Routing Header (SRH) 222 also includes an IPv6 destination address in the SID list and/or carries other information in a field of the SRH 222.

Fig. 2D illustrates a SID and a uSID (e.g., in a 128-bit uSID container) for processing (including forwarding) packets in a network, including invoking a per-hop behavior (PHB) in accordance with various embodiments.

SRv6 is a segment routing feature that implements source routing in the IPv6 data plane. As shown, the SRv Segment (SID) 230 is represented as a 128-bit SID address, comprising a locator portion (block) 231 (e.g., a highest order bit block), a function portion 232, and an ARG portion 233 for optional variables and/or padding. The first SID is an IPv6 destination address of an external IPv6 header of the SRv packet; if so, the additional SIDs are carried in the segment list of the Segment Routing Header (SRH) (i.e., IPv6 extension header). SID is a topology (e.g., internet gateway protocol IGP) or service (e.g., virtual Private Network (VPN), network Function Virtualization (NFV)) type.

In one embodiment, one or more SRv6 micro-segments (uSIDs) are encoded in a single SID (e.g., 128-bit IPv6 address). As shown, the uSID container 240 includes a uSID block 241 (e.g., an IPv6 prefix) followed by one or more usids 242, wherein any remaining portion of the uSID container 240 is filled with end of container (E-o-C) usids 243 (e.g., 0x 0000) or fill bits. In one embodiment, when all desired uSIDs are within a single IPv6 address (e.g., uSID container 240), the SRH is not included in the SRv packet. One embodiment uses the uSIDs in a manner that takes advantage of the benefits of IP, such as, but not limited to, longest prefix match forwarding, prefix summarization, identification entropy, and the like.

FIG. 2D also illustrates some of the extensible number of variants of the semantics of a particular uSID and a particular SID 280-290 within a uSID container 250-270 in accordance with various embodiments.

As shown, the uSID container 250 includes a uSID block 251 (e.g., an IPv6 prefix) followed by one or more combined PHBs and routing usids 252, where each uSID identifies a particular PHB for a particular node (e.g., according to a corresponding network slicing operation) and is used to forward packets into the particular node. In one embodiment, the prefix comprising the uSID block and the highest order uSID 252 is the advertising address of the network node that will execute the PHB identified based on the highest order uSID 252. One embodiment using such a uSID that defines PHB and network routing is described below with respect to FIG. 4. In one embodiment, the uSID indirectly identifies a corresponding slice identifier (SLID).

As shown, in accordance with a corresponding one of the embodiments, the uSID container 260 includes a uSID block 261 (e.g., an IPv6 prefix) and one or more global per-hop behaviors (PHBs) usids 263 located at different locations within the uSID container 260. The global SRv PHB uSID identifies the PHB to be performed on the SRv6 packet by the at least SRv node identified by the routing uSID (262, 265) in the uSID container 260. In one embodiment, global PHB uSID 263 immediately follows uSID block 261 (e.g., with zero forwarding usids 262 and one or more routing usids 265), a corresponding embodiment will be described below with respect to fig. 5A. In one embodiment, global PHB uSID 263 is located after one or more routes uSID 262 following uSID block 261 in uSID container 260, a corresponding embodiment is described below with respect to fig. 5B. In one embodiment, the uSID container 260 includes a uSID 264 that identifies a particular flexible algorithm (Flex-Algo) to be used in processing packets in a network.

As shown, the upid container 270 includes a upid block 271 (e.g., an IPv6 prefix) and one or more paired routes and PHB upids 272. In one embodiment, pairing 272 is a routing uSID followed by a PHB uSID, a corresponding embodiment is described below with respect to FIG. 5C. In one embodiment, pairing 272 is a PHB uSID followed by a routing uSID, a corresponding embodiment is described below with respect to FIG. 5D.

As shown, SID 280 includes a Locator (LOC) 281, a combined PHB and routing value 283 (e.g., in the SR Function (FUNCT) portion of SID 280), and zero or more arguments or padding 284 (e.g., in the SR Argument (ARG) portion of SID 280). In one embodiment, SID 280 includes a Flex-Algo value 282 (e.g., in the SR function (function) portion of SID 280) that identifies a particular flexible algorithm (Flex-Algo) for use in processing packets in the network.

As shown, SID 290 includes a Locator (LOC) 291, a routing value 293 in the SR Function (FUNCT) portion of SID 290, and a PHB value 294 encoded in the Argument (ARG) portion of SID 290. In one embodiment, SID 290 includes a Flex-Algo value 292 (e.g., in the SR function (function) portion of SID 290) that identifies a particular flexible algorithm (Flex-Algo) for use in processing packets in the network.

FIG. 3A illustrates a process according to one embodiment. Processing begins with process block 300. In process block 302, a node in the network is configured as a PHB implementing its respective network slice, each PHB being associated with a SLID. In one embodiment, each network slice has resource allocation and/or guarantees for the network node, such as, but not limited to: link bandwidth capacity, hardware resources (e.g., queues, content addressable memory entries, memory), hardware queue mapping, network processing unit or co-processor mapping, forwarding table mapping, BCDL (bulk code download) route update priority, TI-LFA (topology independent-loop free replacement) fast reroute protection, scheduling, service function chaining, field operation, administration and maintenance (oam) behavior (such as time stamping and logging of interface/node identifiers), security, reliability, time sensitivity (e.g., considered deterministic scheduling in one embodiment), isolation, partitioning, and/or quality of service (QoS) applications. In process block 304, the uSID and/or SID prefix is assigned and mapped to the SLID, or announced locally and then by a network node (e.g., using an IGP flood record TLV) and/or by a network controller (e.g., a Software Defined Network (SDN) controller). In process block 306, the network node is programmed to invoke a corresponding PHB (e.g., according to the forwarding behavior of the network slice) based on the uSID and/or SID prefix. In process block 308, the ingress network node is programmed with a corresponding segment routing policy (e.g., defined by an ordered SID and/or a uSID) to cause the desired path forwarding and PHB on the network node. In one embodiment, these segment routing policies are calculated by the respective ingress network node based on the received advertisement of the SID/uSID and its associated SLID. In one embodiment, the network controller assigns a SID/uSID and its associated SLID, calculates a segment routing policy, and programs each ingress network node to correspondingly apply the segment routing policy to the received packet. The process of the flowchart of FIG. 3A is complete, as indicated by process block 309.

FIG. 3B illustrates a process according to one embodiment. Processing begins at process block 310. In process block 312, a corresponding segment routing policy is applied to the received packet (e.g., based on packet classification) by the ingress node, which generally includes encapsulating the received packet in a SRv packet that includes one or more SIDs (e.g., each SID may include one or more usids) defining the PHB applied to the encapsulated SRv packet by the network node traversing the network according to the calculated path through the network. In one embodiment, the SRv packet does not include a SLID, but rather includes SID/uSID per hop behavior instructions that map to a particular end-to-end network slice (e.g., as an indirect method). In one embodiment, a particular per-hop behavior may be bound to one or more end-to-end network slices.

In process block 314, the SRv packet is forwarded to the egress node over the network according to the segment routing policy, where the network node performs its corresponding PHB identified by the (valid) SID/uSID, with the IPv6 destination address in the outer header (and possibly the segment routing header) of the SRv6 packet typically updated to implement the remainder of the segment routing policy (e.g., including traversing to the next network node). In one embodiment, an ingress line card of a network node determines an associated SLID based on a valid PHB SRv6 uSID instruction of a received packet. The SLID is used to provide forwarding behavior to the packet and is programmed in hardware based on a local slice profile configured for the link receiving the packet. In one embodiment, packets with a particular PHB may be quickly rerouted to another matching PHB in response to an identified failure condition.

In one embodiment, the end-to-end forwarding behavior (e.g., a segment routing policy) provides layer 3 virtual private network (L3 VPN) and/or Ethernet Virtual Private Network (EVPN) services. In one embodiment, the same network slice is associated with multiple Virtual Private Networks (VPNs) having common service guarantee requirements and is mapped to the same PHB/segment routing policy. In one embodiment, VPNs with different requirements are carried in different network slices and mapped to different PHB/segment routing policies.

In process block 316, the egress node processes the received SRv packet in accordance with the PHB of the segment routing policy identified by the (valid) SID/uSID and decapsulates the original packet, where the processing typically includes forwarding the decapsulated original packet from the egress node in accordance with its corresponding PHB.

The process of the flowchart of FIG. 3B is complete, as indicated by process block 319.

Fig. 4 illustrates a network 400 operating in accordance with one embodiment. As shown, the network 400 includes: ingress node 2 (420) programmed to apply the segment routing PHB policy identified in the destination address of SRv packet 402; intermediate node 3 (430), programmed with a network slice mapping table 432; intermediate node 4 (440), programmed with a network slice mapping table 442; and an egress node 5 (450) programmed with a network slice mapping table 452. Each of the uads in each of the mapping tables 432, 442, 452 identifies (a) the address of the corresponding network node 430, 440, 450, and (b) its corresponding SLID 1 (low latency (LL) slice), SLID 2 (high bandwidth (HB) slice), SLID 3 (best effort (BE) slice), etc. It follows that the advertising address of the network node 3 (430) comprises prefixes (2001:db8:uS10:), (2001:db8:uS100:), and (2001:db8:uS200:); the advertised address of network node 4 (440) includes prefixes (2001:db8:uS15:), (2001:db8:uS150:), and (2001:db8:uS250:); and the advertised address of the network node 5 (450) includes prefixes (2001:db8:uS25:), (2001:db8:uS500:), and (2001:db8:uS600:).

Referring specifically to packets 402-404 shown in FIG. 4, their IPv6 destination addresses (and/or SIDs in the segment list in the SRH) include a uSID container that includes a uSID block (2001: db8:), followed by one or more combined PHBs and a routing uSID (both the traversal order identifying the nodes of network 500 and their corresponding PHBs applied to the packet), followed by (:) (e.g., zero or more E-o-C uSIDs, each E-o-C uSID typically having a value of 0x 0000).

As shown in fig. 4, network node 2 (420) receives and processes original IP packet 401, including applying a corresponding segment routing policy (e.g., encapsulating original packet 401 in segment routing packet 402 including the illustrated IPv6 destination address) to cause the desired transmission and end-to-end low latency PHB through network 400 via network nodes 430, 440, and 450.

In one embodiment, this processing of network node 2 (420) includes generating SRv packets 402 with corresponding end-to-end PHB and routing uSID SRv6 policies (2001:db8:uS10:uS15:uS25:), in the destination address of SRv packets 402, and (low latency) correspondingly transmitting SRv packets 402 with IPv6 destination addresses (2001:db8:uS10:uS15:uS25:) to network 400.

SRv6 packet 402 is forwarded to network node 3 based on its IPv6 destination address (430). Based on this IPv6 destination address with a valid uSID of uS10 (2001:db8:us10:us15:us25: the network node 3 (430) identifies a low latency PHB (SLID 1) via a lookup operation in the network slice map 432. Such low latency processing includes: updating the IPv6 destination address of the received packet 402 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 403; and (low latency) correspondingly sends SRv packets 403 with IPv6 destination addresses (2001:db8:uS15:uS25::) to the network 400.

SRv6 packet 403 is forwarded to network node 4 based on its IPv6 destination address (440). Based on this IPv6 destination address (2001:db8:uS15:uS25::) with the valid uSID of uS10, network node 4 (440) identifies a low-latency PHB (SLID 1) via a lookup operation in network slice mapping table 442. Such low latency processing includes: updating the IPv6 destination address of the received packet 403 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 404; and (low latency) correspondingly sends SRv packet 404 with IPv6 destination address (2001:db8:us25:) to network 400.

SRv6 packet 404 is forwarded to network node 5 based on its IPv6 destination address (450). Based on this IPv6 destination address (2001:db8:uS25::) with the valid uSID of uS25, network node 5 (450) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 452. Such low latency processing includes decapsulating and (low latency) correspondingly transmitting the original IP packet 401 into the network 400.

Each of fig. 5A-5D illustrates a network 500 operating in accordance with a corresponding one of the embodiments. As shown in each of fig. 5A to 5D, the network 500 includes: an ingress node 2 (520) programmed to apply a segment routing PHB policy identified in a destination address of a corresponding transmitted SRv packet (502, 512, 582, 592); intermediate node 3 (530), programmed with a network slice mapping table 532; intermediate node 5 (540), programmed with a network slice mapping table 542; and an egress node 5 (550) programmed with a network slice mapping table 552.

In one embodiment, a network controller (e.g., an SDN controller) provisions SRv for each node a 6 PHB uSID instruction (e.g., a particular global uSID) and its matching slice ID and slice profile. In one embodiment, the network controller allocates and causes programming of the mapping tables 532, 542, and 552 to provide per-hop behavior (PHB) at the various nodes that can help to implement end-to-end network slicing. In one embodiment, the uSIDs provisioned and installed in each mapping table 532, 542, and 552 include a global uSID: PHBuSID1: map to SLID 1 (low latency (LL) slice); PHBuSID2, which maps to SLID 2 (high bandwidth (HB) slice); and PHBuSID3, which maps to SLID 3 (best effort (BE) slice), etc.

In one embodiment, the global uSIDs programmed in the mapping tables 532, 542, 552 identify slicing behavior to be performed on packets, with additional routing uSIDs included in the uSID container (e.g., IPv6 destination address, SID) to identify ordered forwarding between segmented routing nodes in the network 500. Fig. 5A-5D each illustrate a different one of the embodiments using a PHB uSID (identifying slicing behavior) and a routing uSID (identifying ordered forwarding between segmented routing nodes in network 500).

FIG. 5A illustrates a network 500 operating in accordance with one embodiment in which a global PHB uSID occurs in higher order bits than a routing uSID in a uSID container. By separating the PHB and routing functions into different uSIDs, one embodiment uses different combinations of routing uSIDs and PHB uSIDs. Thus, in one embodiment, the same routing uSID may be used to route packets to a node while applying a different PHB (e.g., the PHB identified in the global PHB uSID).

Referring specifically to packets 502-504 shown in FIG. 5A, their IPv6 destination addresses (and/or SIDs in the segment list in the SRH) include a uSID container that includes a uSID block (2001: db8:), followed by a global PHB uSID (identifying the end-to-end behavior to be performed on the packet), followed by one or more uSIDs (identifying the traversal order of the nodes of network 500), followed by (:) (e.g., zero or more E-o-C uSIDs, each E-o-C uSID typically having a value of 0x 0000).

In one embodiment, the advertising addresses of the network nodes 530-550 have the following format: (uSIDBlock: globalphBuSID: routinguSIDofNetworkNode:). In one embodiment, the advertising address of the network node 3 (530) includes prefixes (2001:db8:PHBuSID 1:uS10:), (2001:db8:PHBuSID 2:uS10:), and (2001:db8:PHBuSID 3:uS10:). In one embodiment, the advertising address of the network node 4 (540) includes prefixes (2001:db8:PHBuSID 1:uS15:), (2001:db8:PHBuSID 2:uS15:), and (2001:db8:PHBuSID 3:uS15:). In one embodiment, the advertising address of the network node 5 (550) includes prefixes (2001:db8:PHBuSID 1:uS25:), (2001:db8:PHBuSID 2:uS25:), and (2001:db8:PHBuSID 3:uS25:).

As shown in fig. 5A, network node 2 (520) receives and processes original IP packet 501, including applying a corresponding segment routing policy (e.g., encapsulating original packet 501 in segment routing packet 502 including the illustrated IPv6 destination address) to cause the desired transmission and end-to-end low latency PHB through network 500 via network nodes 530, 540, and 550.

In one embodiment, this processing of network node 2 (520) includes generating SRv packets 502 with a corresponding end-to-end PHB uSID SRv6 policy (2001:db8: PHBuSID1: u10: uS15: uS 25:), in the destination address of SRv packets 502, and (low latency) correspondingly transmitting SRv packets 502 with IPv6 destination addresses (2001:db8: PHBuSID1: uS10: uS15: uS 25:) to network 500.

SRv6 packet 502 is forwarded to network node 3 based on its IPv6 destination address (530). Based on the IPv6 destination address with the valid uSID of uS10 (2001:db8:phbusid1:us10:us15:us25: the network node 3 (530) identifies a low latency PHB (SLID 1) via a lookup operation in the network slice mapping table 532 based on PHBuSID 1. Such low latency processing includes: updating the IPv6 destination address of the received packet 502 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 503; and (low latency) correspondingly sends SRv packet 503 with an IPv6 destination address (2001:db8: PHBuSID1: uS15: uS 25:) to network 500.

SRv6 packet 503 is forwarded to network node 5 based on its IPv6 destination address (540). Based on the IPv6 destination address with the valid uSID of uS10 (2001:db8:phbusid1:us15:us25: the network node 5 (540) identifies a low latency PHB (SLID 1) via a lookup operation in the network slice mapping table 542 based on PHBuSID 1. Such low latency processing includes: updating the IPv6 destination address of the received packet 503 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 504; and (low latency) correspondingly sends SRv packets 504 with IPv6 destination address (2001:db8: PHBuSID1: uS 25:) to network 400.

SRv6 packet 504 is forwarded to network node 5 based on its IPv6 destination address (550). Based on this IPv6 destination address (2001:db8:PHBuSID 1:uS25::) with the valid uSID of uS25, network node 5 (550) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 552 based on PHBuSID 1. Such low latency processing includes decapsulating and (low latency) correspondingly transmitting the original IP packet 501 into the network 500.

Fig. 5B illustrates a network 500 operating in accordance with one embodiment in which a global PHB uSID occurs immediately after a routing uSID in a uSID container. By separating the PHB and routing functions into different uSIDs, one embodiment uses different combinations of routing uSIDs and PHB uSIDs. Thus, in one embodiment, the same routing uSID may be used to route packets to a node while applying a different PHB (e.g., the PHB identified in the global PHB uSID).

Referring specifically to the packet 512 shown in FIG. 5B, its IPv6 destination address (and/or SIDs in the segment list in the SRH) includes a uSID container that includes a uSID block (2001: db8:), followed by one or more uSIDs (corresponding to segment routing nodes 530, 540, 550 for traversing the packet 512, 513, 514 through the network 500), followed by one or more global PHB uSIDs (identifying the end-to-end behavior to be performed on the packet), followed by zero or more E-o-C uSIDs (:).

It follows that the advertising addresses of the network nodes 530-550 are typically in the following format: (uSIDBlock: routinguSIDofNetworkNode:). In one embodiment, the advertised address of network node 3 (530) includes a prefix (2001:db8:uS10:). In one embodiment, the advertised address of network node 4 (540) includes a prefix (2001:db8:uS15:). In one embodiment, the advertised address of network node 5 (550) includes a prefix (2001:db8:uS25:).

As shown in fig. 5B, network node 2 (520) receives and processes original IP packet 511, including applying a corresponding segment routing policy (e.g., encapsulating original packet 511 in segment routing packet 512 including the illustrated IPv6 destination address) to cause the desired transmission and end-to-end low latency PHB through network 500 via network nodes 520, 530, and 540. This processing by the network node 2 includes generating SRv packets 512 with a corresponding end-to-end PHB uSID SRv6 policy (2001:db8:u10:uS15:uS25:PHBuSID 1:), in the destination address of the SRv packets 512, and (low latency) correspondingly transmitting SRv packets 512 with IPv6 destination addresses (2001:db8:uS10:uS15:uS25:PHBuSID 1:) to the network 500.

SRv6 packet 512 is forwarded to network node 3 based on its IPv6 destination address (530). Based on the IPv6 destination address with the valid uSID of uS10 (2001:db8:us10:us15:us25:phbusid1: the network node 3 (530) identifies a low latency PHB (SLID 1) via a lookup operation in the network slice mapping table 532 based on PHBuSID 1. Such low latency processing includes: updating the IPv6 destination address of the received packet 512 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 513; and (low latency) will correspondingly have an IPv6 destination address (2001:db8:us15:us25:phbusid1: a) SRv packet 513 is sent to the network 500.

SRv6 packet 513 is forwarded to network node 5 based on its IPv6 destination address (540). Based on the IPv6 destination address with the valid uSID of uS10 (2001:db8:u15:us25:phbusid1: the network node 4 (540) identifies a low latency PHB (SLID 1) via a lookup operation in the network slice map 542 based on PHBuSID 1. Such low latency processing includes: updating the IPv6 destination address of the received packet 513 by removing the valid uSID and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating SRv packet 514; and (low latency) correspondingly sends SRv packet 514 with IPv6 destination address (2001:db8:u25:phbusid1::) to network 400.

SRv6 packet 514 is forwarded to network node 5 based on its IPv6 destination address (550). Based on this IPv6 destination address (2001:db8:uS25:PHBuSID 1::) with the valid uSID of uS25, network node 5 (550) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 552 based on PHBuSID 1. Such low latency processing includes decapsulating and (low latency) correspondingly transmitting the original IP packet 511 into the network 500.

Each of fig. 5C-5D illustrates a network 500 operating in accordance with a corresponding one of the embodiments, wherein the uSID container includes a pairing of routes and PHB usids. For ease of illustration, global PHB uSID is used; thus, the mapping tables 532, 542, and 552 are illustrated as having the same mapping. In one embodiment, the local PHB uSIDs are used in one embodiment of each of FIGS. 5C-5D to specify the respective PHBs performed on the packet by each of the respective nodes of the network 500.

Fig. 5C illustrates one embodiment of ordered pairing of routing uSID > using < PHB uSID.

Referring specifically to packets 582-584 shown in FIG. 5C, their IPv6 destination address (and/or SIDs in the segment list in the SRH) includes a uSID container that includes a uSID block (2001: db8:), followed by < PHB uSID, one or more ordered pairs of routing uSIDs > (identifying the node traversal order and PHBs to be applied by the corresponding node), followed by (:) (e.g., zero or more E-o-C uSIDs, each E-o-C uSID typically having a value of 0x 0000).

As shown in fig. 5C, network node 2 (520) receives and processes original IP packet 581, including applying a corresponding segment routing policy (e.g., encapsulating original packet 581 in segment routing packet 582 including the illustrated IPv6 destination address) to cause a desired transmission PHB through network 500 via network nodes 530, 540, and 550.

In one embodiment, this processing by network node 2 (520) includes generating SRv packets 582 with corresponding end-to-end PHB uSID SRv6 policies (2001:db8:PHBuSID 1:u10:PHBuSID 1:uS15:PHBuSID 1:uS25:)) in destination addresses of SRv packets 582, and (low latency) correspondingly transmitting SRv packets 582 with IPv6 destination addresses (2001:db8:PHBuSID 1:uS10:PHBuSID1:uS15:PHBuSID 1:uS25:)) to network 500.

SRv6 packet 582 is forwarded to network node 3 based on its IPv6 destination address (2001:db8:PHBuSID 1:uS10:PHBuSID1:uS15:PHBuSID 1:uS25::) (530). The network node 530 receives SRv a packet 582 that includes a valid uSID ordered pair of < PHBuSID1, uS10 >. Based on PHBuSID1, network node 3 (530) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 532. Such low latency processing includes: updating the IPv6 destination address of the received packet 582 by removing the uSID in the valid uSID ordered pair and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating a SRv packet 583; and (low latency) correspondingly sends SRv packets 583 with IPv6 destination addresses (2001:db8: PHBuSID1: uS15: PHBuSID1: uS 25:) to network 500.

SRv6 packet 583 is forwarded to network node 4 based on its IPv6 destination address (2001:db8:PHBuSID 1:uS15:PHBuSID1:uS 25::) (540). The network node 540 receives SRv a packet 583 that includes a valid uSID ordered pair of < PHBuSID1, uS15 >. Based on PHBuSID1, network node 4 (540) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 542. Such low latency processing includes: updating the IPv6 destination address of the received packet 583 by removing the uSID in the valid uSID ordered pair and shifting the remaining uSID and adding the E-o-C uSID in the low order bits, thereby generating a SRv packet 584; and (low latency) correspondingly sends SRv packets 584 with IPv6 destination addresses (2001:db8: phbusid1: us 25:) to the network 500.

SRv6 packet 584 is forwarded to network node 5 (550) based on its IPv6 destination address (2001:db8:PHBuSID 1:uS 25:). The network node 550 receives SRv a packet 584 that includes a valid uSID ordered pair of < PHBuSID1, uS25 >. Based on PHBuSID1, network node 5 (550) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 552. Such low latency processing includes decapsulating and (low latency) correspondingly transmitting original IP packet 581 into network 500.

Fig. 5D illustrates one embodiment of ordered pairing using < route uSID, PHB uSID >.

Referring specifically to the packet 592-594 shown in FIG. 5D, its IPv6 destination address (and/or SIDs in the segment list in the SRH) includes a uSID container that includes a uSID block (2001: db8:), followed by < route uSID, PHB uSID > (one or more ordered pairs that identify the node traversal order and PHB to be applied by the corresponding node), followed by (:) (e.g., zero or more E-o-C uSIDs, each E-o-C uSID typically having a value of 0x 0000).

In one embodiment, the advertising addresses of the network nodes 530-550 have the following format: (uSIDBlock: routinguSIDofNetworkNode:). In one embodiment, the advertised address of network node 3 (530) includes a prefix (2001:db8:uS10:). In one embodiment, the advertised address of network node 4 (540) includes a prefix (2001:db8:uS15:). In one embodiment, the advertised address of network node 5 (550) includes a prefix (2001:db8:uS25:).

In one embodiment, the advertising addresses of the network nodes 530-550 have the following format: (uSIDBlock: routinguSIDofNetworkNode: globalpBuSID) or (uSIDBlock: routinguSIDofNetworkNode: PHBuSID) that increases the number of addresses advertised by network nodes 530-550 by the number of different global or local PHB uSIDs used by the network nodes 530-550 and generally correspondingly increases the size of RIBs and FIBs on each network node 530-550 to support forwarding packets to each advertising address.

As shown in fig. 5D, network node 2 (520) receives and processes original IP packet 591, including applying a corresponding segment routing policy (e.g., encapsulating original packet 591 in segment routing packet 592 including the illustrated IPv6 destination address) to cause desired transmissions and PHBs through network 500 via network nodes 530, 540, and 550.

In one embodiment, this processing of network node 2 (520) includes generating a request with a corresponding end-to-end PHB uSID SRv6 policy in the destination address of SRv packet 592 (2001:db8:u10:phbusid1:u15:phbusid1:u25:phbusid1::), and (low latency) will have IPv6 destination addresses (2001:db8:u10:phbusid1:u15:phbusid1:u25:phbusid1::) SRv packets 592 are transmitted to network 500.

SRv6 packet 592 is forwarded to network node 3 based on its IPv6 destination address (2001:db8:uS10:PHBuSID 1:uS15:PHBuSID1:uS25:PHBuSID 1::) (530). The network node 530 receives SRv a packet 592 comprising a valid uSID ordered pair of < uS10, PHBuSID1 >. Based on PHBuSID1, network node 3 (530) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 532. Such low latency processing includes: updating the IPv6 destination address of received packet 592 by removing the uSID in the valid uSID ordered pair and shifting the remaining uSID and adding two E-o-C usids in the low order bits, thereby generating SRv packet 593; and (low latency) correspondingly sends SRv packets 593 with IPv6 destination addresses (2001:db8:uS15:phBuSID 1:uS25:phBuSID 1::) to network 500.

SRv6 packet 593 is forwarded to network node 4 based on its IPv6 destination address (2001:db8:uS15:PHBuSID 1:uS25:PHBuSID 1::) (540). The network node 540 receives SRv a packet 593 that includes valid uSID ordered pairs of < uS15, PHBuSID1 >. Based on PHBuSID1, network node 4 (540) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 542. Such low latency processing includes: updating the IPv6 destination address of the received packet 593 by removing the uSID in the valid uSID ordered pair and shifting the remaining uSID and adding two E-o-C usids in the low order bits, thereby generating a SRv packet 594; and (low latency) correspondingly sends SRv packets 594 with an IPv6 destination address (2001:db8:uS 25:PHBuSID 1::) to the network 500.

SRv6 packet 594 is forwarded to network node 5 (550) based on its IPv6 destination address (2001:db8:uS 25:PHBuSID 1:). The network node 550 receives SRv a packet 594 that includes a valid uSID ordered pair of < uS15, PHBuSID1 >. Based on PHBuSID1, network node 5 (550) identifies a low latency PHB (SLID 1) via a lookup operation in network slice map 552. Such low latency processing includes decapsulating and (low latency) correspondingly transmitting the original IP packet 591 into the network 500.

In summary, in one embodiment, a segment routing internet protocol version 6 (SRv) micro segment ("uSID") is included in a destination address of a packet transmitted over a network, and possibly in other segment identifiers ("SIDs"), and invokes corresponding network behavior, including but not limited to implementing a corresponding network slice. In one embodiment, the network node is configured to perform a differential network slice implementation function based on slice representative values provided by the global and/or local uSIDs of the packet. The configuration may be defined by a controller and/or routing protocol advertisement in the network. In response to the received packet, the network node identifies and performs a corresponding network slice implementation function based on the slice representative values provided by the one or more global and/or local uSIDs of the destination address of the received packet. Various encodings within the IPv6 destination address of the encapsulated packet are disclosed.

In view of the many possible embodiments to which the principles of this disclosure may be applied, it will be understood that the embodiments and aspects thereof described herein with respect to the drawings/figures are illustrative only and should not be taken as limiting the scope of the disclosure. For example, and as will be apparent to one of ordinary skill in the art, many of the process block operations can be reordered to be performed before, after, or substantially concurrently with other operations. Moreover, many different forms of data structures may be used in various embodiments. The disclosure described herein contemplates all such embodiments as may come within the scope of the following claims and equivalents thereof.

Claims

1. A method, comprising:

receiving, by a particular segment routing node in a network, a segment routing version 6 (SRv) packet including an internet protocol version 6 (IPv 6) destination address, the IPv6 destination address being an address of the particular segment routing node, wherein the IPv6 destination address includes a plurality of micro segments (uads), a particular uad of the plurality of uads mapping to a particular network slicing behavior of a plurality of different network slicing behaviors performed by the particular segment routing node on different packets;

Processing the received SRv by the particular segment routing node, comprising: differentially processing the particular SRv packet according to the particular network slice behavior, wherein the differential processing is different from processing according to one of the plurality of different network slice behaviors that is not the particular network slice behavior; updating the IPv6 destination address, comprising: removing a routing uSID identifying the particular segmented routing node and shifting one or more uSIDs in the IPv6 destination address of the received SRv packet into higher order bit positions; and transmitting the SRv packet with the updated IPv6 destination address from the particular segment routing node.

2. The method according to claim 1, comprising: a lookup operation is performed in the network slice mapping data structure resulting in an identification of the particular network slice behavior based on the particular uSID.

3. The method of claim 2, wherein the particular uSID is a per-hop behavior (PHB) and routing uSID that indirectly identifies a combination of the particular network slice behavior, and is part of an advertised route for the particular segmented routing node.

4. The method of claim 2, wherein the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the particular uSID concatenated with the first routing uSID is part of an advertised route of the particular segmented routing node; and wherein the removing the routing uSID comprises removing the first routing uSID.

5. The method of claim 2, wherein the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the first route uSID is part of an advertised route of the particular segmented routing node; and wherein the removing the routing uSID comprises removing the first routing uSID.

6. The method of claim 5, wherein the particular uad is in a lower order bit in the IPv6 destination address than all routing uads in the IPv6 destination address of the received SRv packet.

7. The method of any of claims 2-6, wherein the IPv6 destination address of the received SRv packet includes a plurality of pairs of per-hop behaviors (PHBs) and routing usids; wherein the plurality of pairs includes a particular pair including the particular uSID and the routing uSID; and wherein said removing said routing uSID comprises removing said particular pairing.

8. The method of claim 7, wherein the particular pairing is part of an advertised route of the particular segmented routing node.

9. The method of claim 8, wherein the particular pairing comprises the particular uuid concatenated with the routing uusid.

10. The method of claim 8, wherein the particular pairing comprises the routing uasid concatenated with the routing uasid.

11. The method of claim 7, wherein the routing uasid is part of an advertised route for the particular segmented routing node.

12. The method of any of claims 1-11, wherein the particular uSID is a global per-hop behavior (PHB) uSID that identifies a PHB to be performed on one or more other segment routing nodes in the network.

13. The method of claim 12, wherein the particular ussd is part of an advertised route of the particular segmented routing node.

14. The method of any of claims 1-13, wherein the IPv6 destination address of the received SRv packet comprises a Flex-Algo u sid; and processing the packet includes processing according to a flexible algorithm identified by the Flex-Algo uSID.

15. The method of any of claims 1 to 14, wherein, prior to receiving the SRv packet, the particular segment routing node configures one or more hardware resources to perform differential processing for each of the plurality of different network slice behaviors.

16. The method of claim 15, wherein each of the plurality of network slice behaviors defines a packet processing delay or a link bandwidth capacity.

17. The method of claim 16, wherein the one or more resources comprise queues, ternary Content Addressable Memory (TCAM), or memory.

18. A packet switching device, comprising:

a plurality of hardware interfaces that transmit and receive packets; and

one or more network processors associated with a memory;

wherein the packet switching device performs a packet processing operation, wherein the packet processing operation comprises:

receiving, via one of the hardware interfaces, a segment routed version 6 (SRv) packet comprising an internet protocol version 6 (IPv 6) destination address, the IPv6 destination address being an address of the packet switching device, wherein the IPv6 destination address comprises a plurality of micro segments (uads), a particular uad of the plurality of uads mapping to a particular network slicing behavior of a plurality of different network slicing behaviors performed by the packet switching device on different packets;

Processing the received SRv by the particular packet switching device includes: differentially processing the particular SRv packet according to the particular network slice behavior, wherein the differential processing is different from processing according to one of the plurality of different network slice behaviors that is not the particular network slice behavior; updating the IPv6 destination address, comprising: removing a routing uSID identifying the packet switching device and shifting one or more uSIDs in the IPv6 destination address of the received SRv packet into higher order bit positions; and sending the SRv packet with the updated IPv6 destination address from a hardware interface of the plurality of hardware interfaces.

19. The packet switching device of claim 18, wherein the particular ussd is a per-hop behavior (PHB) and routing ussd that indirectly identifies a combination of the particular network slice behaviors and is part of an advertised route for the particular packet switching device.

20. The packet switching device of claim 18, wherein the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the particular uSID concatenated with a first routing uSID is part of an advertised route of the packet switching device; and wherein the removing the routing uSID comprises removing the first routing uSID.

21. The packet switching device of claim 18, wherein the particular uSID is a Per Hop Behavior (PHB) uSID; wherein the first routing uSID is part of an advertised route for the particular packet switching device; and wherein the removing the routing uSID comprises removing the first routing uSID.

22. The packet switching device of claim 18, wherein the IPv6 destination address of the received SRv packet includes a plurality of pairs of per-hop behaviors (PHBs) and routing usids; wherein the plurality of pairs includes a particular pair including the particular uSID and the routing uSID; and wherein said removing said routing uSID comprises removing said particular pairing.

23. An apparatus, comprising:

means for receiving, by a particular segment routing node in a network, a segment routing version 6 (SRv 6) packet comprising an internet protocol version 6 (IPv 6) destination address, the IPv6 destination address being an address of the particular segment routing node, wherein the IPv6 destination address comprises a plurality of micro segments (uads), a particular uad of the plurality of uads being mapped to a particular network slicing behavior of a plurality of different network slicing behaviors performed by the particular segment routing node on different packets;

Means for processing, by the particular segment routing node, the received SRv, including differentially processing the particular SRv packet according to the particular network slice behavior, wherein the differential processing is different from processing according to one of the plurality of different network slice behaviors that is not the particular network slice behavior; updating the IPv6 destination address, comprising: removing a routing uSID identifying the particular segmented routing node and shifting one or more uSIDs in the IPv6 destination address of the received SRv packet into higher order bit positions; and transmitting the SRv packet with the updated IPv6 destination address from the particular segment routing node.

24. The apparatus of claim 23, further comprising means for performing the method of any one of claims 2 to 16.

25. A computer program, computer program product or computer readable medium comprising instructions which, when executed by a computer, cause the computer to perform the steps of the method according to any of claims 1 to 16.