Next steps
Pointers into the rest of the Prelude TE documentation once your first peer is up.
Your first BGP-LS session is up and the topology is populating. From here, the documentation is organized by what you want to do next rather than by reading order. Pick whichever fits your day.
Add more peers, safely
A single peer is enough to bring up a topology, but production deployments usually want at least two for redundancy.
- BGP Peers / Peer redundancy — when to add a second peer, how to pair route reflectors.
- Peers — the full life cycle of a peer record: create, enable, disable, force reconnect.
- BGP Peers / Authentication — MD5 configuration.
Send topology changes downstream
Streaming the consolidated topology to the rest of your stack is the most common production use case.
- Outputs / NATS — how to declare a NATS output, the topic format, and the payload shape your subscribers receive.
- Concepts / Data flow — a recap of the guarantees the stream gives you (withdrawals are always published, updates are deduplicated).
Understand the Segment Routing data
If your network runs SR-MPLS, SRv6, or Flex-Algo, Prelude TE exposes those attributes as first-class fields on every node, link, and prefix.
- Topology / Nodes — node attributes, SR-MPLS capabilities and Node SIDs, SRv6 locators, Flex-Algo participation.
- Topology / Links — metrics, bandwidth, admin groups, SRLGs, SR-MPLS Adjacency SIDs, SRv6 End.X SIDs.
- Topology / Prefixes — SR-MPLS Prefix SIDs with decoded flags, SRv6 locator metadata.
- Topology / Flex-Algo — Flexible Algorithm Definitions, constraints, per-algorithm views.
Drive the engine from an AI tool
Prelude TE exposes its full REST API over the Model Context Protocol. With the MCP server and the companion skill installed, your AI tool can add a peer, query the topology, and configure an output without you leaving its chat.
Operate in production
- Monitoring / Health — the endpoints to wire into your readiness and liveness probes.
- Monitoring / Metrics — what the engine exposes for Prometheus scraping.
Reference
- API reference — every endpoint, generated from the same OpenAPI spec the engine ships with.
- Configuration — every environment variable.
- Glossary — the terms used across the documentation.