Platform · Deep Collection

See what monitoring can't.

Up/down tells you a light went off. Phantom tells you why. It collects protocol-level state from every multi-vendor device over SNMP and SSH, and keeps it as time-series history you can rewind.

Coverage

The protocols that actually explain outages.

BGP

BGP path & prefixes

Advertised/received prefixes with AS-path, local-pref, MED, communities and next-hop — see not just what was chosen, but why.

M

Multicast & mroute

Groups, members, (S,G) trees, RPF, PIM and outgoing-interface lists, with source→receiver path discovery.

VRF

VRF & route-targets

Per-tenant tables with route distinguishers and import/export targets — every route attributed to the right customer.

MPLS

MPLS & LDP

Labels, FECs, MPLS interfaces and LDP neighbors — the invisible label plane, made auditable.

RIB

Routing tables

Full RIB per device with next-hop, protocol, admin distance and metric — snapshotted to detect path changes.

SLA

Circuits & IP-SLA

Latency, packet loss and jitter against contracted thresholds, for objective SLA evidence.

IF

Interfaces

Status, traffic, errors and utilization per port, with per-interface monitoring controls.

L2

LLDP · ARP · FDB

Neighbor adjacencies for topology, plus IP↔MAC↔port mapping for endpoint tracking.

Route-flap & stability

Detect flapping and profile it — duty cycle, period and confidence — to separate a one-off from an oscillation.

Time-series by design

Every record kept with history — rewind the network.

Each snapshot is stamped first_seen / last_seen, so "what changed, and when" is a query — not a war room. Diff any two points in time across routing, BGP, multicast, VRF and more.

See your BGP paths and multicast trees.

Book a demo