What Is SRV6 TE Policy?
SRV6 TE Policy utilizes the source routing mechanism of SR to direct packet forwarding based on a predefined list of segments encapsulated by the headend. The concept of SRv6 TE Policy can be likened to navigating through a map, as illustrated in Figure below, to elaborate on SRv6 TE Policy.
The entire process can be succinctly described in five steps:
Topology collection: Gathering information about junctions, lanes, traffic rates, and traffic lights.
Path computation: Calculating paths considering various constraints and multi-dimensional SLAs, such as lowest fees, shortest time, shortest distance, and freeway-preferred.
Information delivery: Sending computed path information to the corresponding user terminal.
Path selection: Requesting the user to choose a path based on the destination address and preferences, where each path comprises multiple roads and key junctions.
Driving guidance: Providing instructions for each path segment as the user approaches junctions, indicating actions like going straight, turning left, turning right, or making a U-turn.
The operational process of SRv6 TE Policy within a network mirrors the navigation map. Figure below illustrates how SRV6 TE Policy functions.
The SRV6 TE Policy workflow involves five steps:
- A forwarder reports network topology information to a network controller through BGP-LS, encompassing TE attributes like node information (like junction information), link details (like road information), link cost (like traffic rate), bandwidth (like lane information), and latency (like traffic light information).
- The controller analyses topology information and computes SLA-compliant paths based on service requirements.
- The controller employs a BGP SR-Policy extension to convey path information to the headend, which then generates SRV6 TE Policies containing essential details like the headend address, destination address, and color.
- The headend selects an appropriate SRV6 TE Policy for guiding forwarding.
- The forwarder executes instructions bound to the SID advertised by itself to forward data.
Figure above depicts how a series of SRV6 SIDs can be encapsulated into an SRH to precisely guide packet forwarding along a planned path, ensuring fine-grained E2E control and meeting SLA requirements.
SRv6 utilizes programmable 128-bit IPv6 addresses, expressing diverse network functions through SRv6 instructions identifying forwarding paths and Value-Added Service (VAS) devices. Notably, SRv6’s exceptional extensibility enables the support of new network functions without necessitating protocol changes, expediting the deployment of innovative network services. SRv6 TE Policy plays a crucial role in meeting E2E service requirements and contributing to SRv6 network programming.
SRV6 TE Policy Structure and Advantages
The SRV6 TE Policy structure aims for enhanced reliability and bandwidth utilization, as depicted in Figure below.
It comprises three elements:
Headend: The node where an SRV6 TE Policy originates.
Color: An extended community attribute defining an application-level network SLA policy.
Endpoint: The destination address of a SRv6 TE Policy.
Advantages of the SRV6 TE Policy structure
Flexible traffic steering: Color and endpoint information are added to a SRv6 TE Policy through configuration, enabling the headend to steer traffic based on matching color and endpoint values in associated routes.
High reliability: A single SRV6 TE Policy can include multiple candidate paths, with the highest preference serving as the primary path and the second-highest preference acting as a backup.
Load balancing: Candidate paths can support both Equal-Cost Multi-Path (ECMP) and Unequal Cost Multi-Path (UCMP) modes, facilitating effective load balancing.
SRV6 TE Policy Implementation
SRV6 TE Policies can accommodate common traditional services, illustrated in a scenario involving EVPN L3VPNV4 over SRv6 TE Policy. Figure below outlines the data forwarding process in this context.
The implementation process involves:
- Begin by instructing the controller to transmit an SRV6 TE Policy featuring a color value of 123 and endpoint address 2001:DB8:40:4 (PE2’s address) to the headend PE1. The SRV6 TE Policy, tailored for this purpose, comprises a sole candidate path with a single segment list <2001:DB8:2:1, 2001:DB8:3:1, 2001:DB8:4::1>.
- Subsequently, PE2 communicates the BGP EVPN route 10.2.2.2/32, equipped with the color value 123 and a next-hop address of 2001:DB8:40:4/128 (PE2’s address), to PE1.
- Upon receipt of the route, PE1 correlates it with the appropriate SRV6 TE Policy based on both color and next-hop address.
- When a regular unicast packet arrives from CE1, PE1 scrutinizes the routing table of the relevant VPN instance. Detecting that the matched VPN route has been associated with an SRV6 TE Policy, PE1 inserts a Segmentation Routing Header (SRH) into the packet. This SRH encapsulates the segment list of the SRV6 TE Policy, culminating in the End.DT4 SID corresponding to the VPN route. PE1 then encapsulates an IPv6 header, searches the routing table, and forwards the packet accordingly.
- The transit nodes, P1 and P2, progressively forward the packet hop by hop, guided by the information provided in the SRH.
- Once the packet reaches PE2, it searches the local SID table using the IPv6 destination address 2001:DB8:4::1. A matching End SID is found. Following the instructions linked to the SID, PE2 decrements the Segment Left (SL) value by 1 and updates the IPv6 Destination Address field to the VPN SID 2001:DB8:4:100.
- Utilizing the VPN SID, PE2 queries the local SID table to identify a matching End.DT4 SID. Adhering to the instructions associated with this SID, PE2 removes the SRH and IPv6 header from the packet. Subsequently, it searches the routing table of the VPN instance linked to the End.DT4 SID 2001:DB8:4:100, based on the destination address in the inner packet. Finally, PE2 forwards the packet to CEZ.
SRV6 TE Policy Reliability Design
Figure below outlines the SRV6 TE Policy reliability design, incorporating TI-LFA for protecting segment lists.
While TI-LFA suffices in most scenarios, employing a SRv6 BE path is recommended for extreme reliability. This ensures services switch to the SRV6 BE path for best-effort forwarding if the SRv6 TE Policy encounters a failure.
Different technologies are applied to protect services against failures at various locations, including link detection, ECMP, IP Fast Reroute (FRR), and TI-LFA FRR. The choice depends on the failure points and their relevance to the segment list of the SRV6 TE Policy. Overall, the design aims to enhance the reliability and resilience of SRV6 TE Policy implementations.
Conclusion
In conclusion, SRV6 TE Policy represents a sophisticated approach to guiding packet forwarding in networks, providing flexibility, reliability, and efficient bandwidth utilization. Its structured design and implementation scenarios showcase its adaptability to diverse service requirements within modern network architectures.