Peering Analytics & Session Monitoring — with Auto-Outreach

A BGP session that quietly went down three days ago is the most expensive kind of failure: route paths shifted, transit traffic spiked, the relationship with the peer eroded a little — and nobody noticed because the session never came back to generate a fresh alarm. RoutePulse closes that gap with full BMP RFC 7854 session monitoring across all 15 TLV types plus Loc-RIB RFC 9069 support, live uptime badges, flap history, per-peer prefix counts, down-since timestamps, and — since v4.30 — an automated peering-manager outreach engine that detects 3-day-down sessions and emails the peer’s NOC contact directly from a GOLINE-branded template.

The Sessions Manager presents a complete inventory of 48+ BGP peers across 5 routers (2 core: Juniper MX + Huawei NetEngine; 3 IXP: SwissIX, MIX-IT, MINAP) with a KPI strip showing Established / Active / Idle / Connect counts, state filter chips, and full-text search. Per-router sections carry gradient headers and ring gauges showing the established-to-total ratio. The new /peers/health sub-page surfaces auto-outreach status badges in-line: HEALTHY, WATCHING (down < 3d), OUTREACH SENT (3d-down detected, NOC notified), ACKNOWLEDGED (peer replied). Combined with transit-only AS identification, peering candidate scoring, IXP community-based traffic attribution, and per-direction P95 billing, the module turns peering management from reactive ticket-driven work into proactive relationship operations.

BMP RFC 7854 + Loc-RIB RFC 9069 session monitoring

The collector implements all 15 BMP TLV types plus Loc-RIB (RFC 9069), giving operators a real-time view of every BGP peer state transition the moment it happens at the router — not the moment an SNMP poll cycle gets around to it. Session uptime badges, flap detection with full history, per-peer received/originated prefix counts, AS-PATH evolution, and down-since timestamps are all derived directly from the BMP stream. The same data feeds the anomaly detectors and the 8-gate Conviction Engine for downstream blackhole decisions.

peering-manager auto-outreach (v4.30+)

When a session has been down for the configured threshold (default 3 days), RoutePulse looks up the peer’s NOC contact via PeeringDB + RIPE WHOIS, renders a GOLINE-branded email referencing the specific peer (ASN, peer description, IXP if applicable, down-since timestamp, session details, prior flap history), and sends it via the operations mailbox with full audit trail. The status badge on /peers/health turns from WATCHING to OUTREACH SENT, and progresses to ACKNOWLEDGED when a reply is detected. No more “I thought we had a session with them” surprises three months later.

Traffic attribution & per-direction P95 billing

IXP community-based traffic attribution correctly splits SwissIX, MIX-IT, and MINAP traffic using BGP community tags (rather than next-hop heuristics that fail with route servers). Per-direction P95 billing is calculated for each transit provider and IXP, producing the exact numbers operators need for monthly cost allocation and capacity planning. The peering candidate scoring engine cross-references PeeringDB IXP presence with observed traffic volume to surface ASNs co-located at your IXPs that could become direct peers — with concrete bytes/sec savings projections.

Peer health, comparison & multi-vendor SNMP

The peer comparison tool selects any two UP peers for side-by-side analysis of prefix coverage, common prefixes, and AS-PATH differences. Peer health monitoring tracks uptime percentages, active flaps, the most-unstable-peer ranking, and a per-peer flap event timeline. SNMP BGP state monitoring complements BMP across all vendors (Juniper, Huawei, MikroTik, Standard MIBs); inline-editable peer descriptions support operator notes; and a Refresh Prefixes button on MikroTik IXP routers reloads prefix counts on demand via the RouterOS API.

Key Capabilities

• Full BMP RFC 7854 collector implementing all 15 TLV types plus Loc-RIB RFC 9069 for real-time per-peer state visibility
• Sessions Manager with KPI strip (Established / Active / Idle / Connect), state filter chips, and full-text search across 48+ peers
• Per-router sections with gradient headers and ring gauges showing established-to-total session ratio across 5 routers
• 2 core routers (Juniper MX, Huawei NetEngine) + 3 IXP routers (SwissIX, MIX-IT, MINAP)
• peering-manager auto-outreach (v4.30+): 3-day-down detection, NOC contact lookup via PeeringDB + RIPE WHOIS
• GOLINE-branded email templates auto-rendered with peer ASN, description, IXP, down-since timestamp, flap history
• Status badge on /peers/health: HEALTHY / WATCHING / OUTREACH SENT / ACKNOWLEDGED with full audit trail
• BMP session uptime badges, flap detection + history, down-since timestamps, per-peer received/originated prefix counts
• Transit-only AS identification discovering ASNs reachable only through transit, not via IXP peering
• Peering candidate scoring identifying co-located ASNs at your IXPs with concrete bytes/sec saving projections
• IXP community-based traffic attribution correctly splitting SwissIX, MIX-IT, and MINAP traffic via BGP communities
• Per-direction P95 billing per transit provider and IXP for accurate cost allocation
• Peer comparison tool for side-by-side prefix coverage, common prefixes, and AS-PATH differences
• SNMP BGP state monitoring across all vendors: Juniper, Huawei, MikroTik, Standard MIBs
• Refresh Prefixes button for MikroTik IXP routers via RouterOS API for on-demand prefix reloads

Engineered and operated by the GOLINE SOC & Network Engineering team.

Explore all RoutePulse features →