diff --git a/docker-compose.yml b/docker-compose.yml
index 8129489..eabffc3 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -419,6 +419,22 @@ services:
     ports:
       - "5053:5053"
 
+  # GoBGP -- pulls the full real Internet routing table (roadmap E1) from the
+  # AS57355 lab route server and BMP-exports it to the OpenBMP collector, where
+  # it lands in PostgreSQL ip_rib as a monitored peer. Config + MRT fallback
+  # script live in ./gobgp (see gobgp/README.md). Receive-only, local AS 65001.
+  gobgp:
+    restart: unless-stopped
+    container_name: obmp-gobgp
+    image: jauderho/gobgp:v4.5.0
+    depends_on:
+      - collector
+    # gobgpd reads /config/gobgpd.conf; the same mount carries mrt-refresh.sh
+    # and the cached MRT dumps it downloads.
+    volumes:
+      - ./gobgp:/config
+    command: ["gobgpd", "-f", "/config/gobgpd.conf", "-t", "toml"]
+
   whois:
     restart: unless-stopped
     container_name: obmp-whois
diff --git a/gobgp/README.md b/gobgp/README.md
new file mode 100644
index 0000000..88b1b54
--- /dev/null
+++ b/gobgp/README.md
@@ -0,0 +1,99 @@
+# GoBGP global Internet table feed (roadmap E1)
+
+This service runs [GoBGP](https://github.com/osrg/gobgp) to pull the **full real
+Internet routing table** (IPv4 ~1M + IPv6 ~200k routes) from Łukasz Bromirski's
+lab route server (**AS57355**) and BMP-export every received route to the
+OpenBMP collector. The table lands in PostgreSQL `ip_rib` as a monitored peer.
+
+- Image: `jauderho/gobgp:v4.5.0` — community-maintained, multi-arch, tracks
+  upstream GoBGP releases (rebuilt within an hour of each release). Chosen
+  because the official `osrg/gobgp` image is published less consistently.
+- Local AS: **65001** (private). Router-id: `10.40.40.250`.
+- The session is **receive-only** — we announce nothing to the route server.
+
+## Files
+
+| File             | Purpose                                                        |
+|------------------|----------------------------------------------------------------|
+| `gobgpd.conf`    | GoBGP daemon config (global, neighbors, BMP export). TOML.     |
+| `mrt-refresh.sh` | MRT full-table fallback loader (cron-driven).                  |
+| `mrt/`           | Created at runtime; cached RouteViews RIB dumps.               |
+
+## Bring it up
+
+The `gobgp` service is defined in the repo `docker-compose.yml`, on the same
+default compose network as `collector`, and `depends_on` it.
+
+```sh
+docker compose config            # validate compose is well-formed
+docker compose up -d gobgp       # start (collector must be running)
+docker logs -f obmp-gobgp
+```
+
+> The live BGP cutover is performed by a human — bringing the container up is
+> all that is needed; GoBGP initiates the eBGP-multihop sessions automatically.
+
+## Confirm the session and route count
+
+```sh
+# session state — expect both neighbors in "Establ"
+docker exec obmp-gobgp gobgp neighbor
+
+# received route counts — expect ~1M IPv4, ~200k IPv6
+docker exec obmp-gobgp gobgp global rib summary -a ipv4
+docker exec obmp-gobgp gobgp global rib summary -a ipv6
+```
+
+## How the data appears in OpenBMP
+
+GoBGP opens an outbound **BMP** session to `obmp-collector:5000` with
+`route-monitoring-policy = "pre-policy"` (Adj-RIB-In, pre import-policy —
+consistent with the rest of the OpenBMP fleet).
+
+In OpenBMP / PostgreSQL the source is identified by the **BMP router**, which
+GoBGP reports using its `router-id` (`10.40.40.250`) and `local-as` (`65001`):
+
+- `routers` table — a row with `ip_address` / name derived from `10.40.40.250`.
+- `bgp_peers` table — two peer rows for `85.232.240.179` and
+  `2001:1a68:2c:2::179`, both `peer_as = 57355`.
+- `ip_rib` — every prefix from the global table, attributed to those peers.
+
+To find it in Grafana/SQL, filter on `peer_as = 57355` or the router-id above.
+
+## MRT fallback
+
+AS57355 is a **single volunteer-run host with no SLA** — it can and does go
+away. `mrt-refresh.sh` keeps the global table in `ip_rib` warm when the live
+feed is down:
+
+1. If any AS57355 session is `Established`, the script does nothing — the live
+   feed is authoritative and must not be overwritten with a stale dump.
+2. Otherwise it downloads the latest full RIB dump from RouteViews
+   (`https://archive.routeviews.org/route-views/bgpdata/YYYY.MM/RIBS/rib.YYYYMMDD.HHMM.bz2`,
+   published every 2 hours UTC) and runs `gobgp mrt inject global <file>`,
+   which installs every prefix into the running daemon. BMP export to the
+   collector then happens automatically.
+
+The script is idempotent (re-uses an already-downloaded dump), guarded by a
+`flock` against overlapping runs, and prunes to the 4 most recent dumps.
+
+### Schedule it (host crontab, 2-hour cadence)
+
+```cron
+0 */2 * * * docker exec obmp-gobgp /config/mrt-refresh.sh >> /var/log/gobgp-mrt.log 2>&1
+```
+
+Run it once manually to verify:
+
+```sh
+docker exec obmp-gobgp /config/mrt-refresh.sh
+```
+
+## Caveats
+
+- **No SLA.** AS57355 is a volunteer lab route server; treat the live feed as
+  best-effort and rely on the MRT fallback for continuity.
+- eBGP-multihop TTL is set to 64 — the route server is many hops away.
+- A full table is ~1M+ prefixes; expect a noticeable load spike in the
+  collector and PostgreSQL when the session first establishes or an MRT dump
+  is injected.
diff --git a/gobgp/gobgpd.conf b/gobgp/gobgpd.conf
new file mode 100644
index 0000000..0041ac3
--- /dev/null
+++ b/gobgp/gobgpd.conf
@@ -0,0 +1,70 @@
+# GoBGP daemon configuration -- OpenBMP "global Internet table" feed (roadmap E1)
+#
+# Pulls the full real Internet routing table (IPv4 ~1M + IPv6 ~200k routes)
+# from Lukasz Bromirski's lab route server (AS57355) and BMP-exports every
+# received route to the OpenBMP collector, where it lands in PostgreSQL ip_rib.
+# Peering spec: https://lukasz.bromirski.net/post/bgp-w-labie-3/
+#
+# Receive-only: we announce NOTHING -- AS57355 explicitly asks peers not to
+# send prefixes. Local AS is 65001 (the value the route server expects).
+# Per the spec: eBGP multihop, no password, keepalive 3600 / hold-time 7200.
+# TOML syntax targets GoBGP v3.x / v4.x.
+
+[global]
+  [global.config]
+    as = 65001
+    router-id = "10.40.40.250"
+    # Listen for inbound BGP on the standard port. We only originate
+    # outbound sessions, but the daemon still needs a listen port.
+    port = 179
+
+# --- Neighbor: route server, IPv4 feed --------------------------------------
+# The IPv4 transport session carries the full IPv4 table only.
+[[neighbors]]
+  [neighbors.config]
+    neighbor-address = "85.232.240.179"
+    peer-as = 57355
+    description = "AS57355 Bromirski lab route-server (IPv4 feed)"
+  [neighbors.timers.config]
+    keepalive-interval = 3600
+    hold-time = 7200
+  [neighbors.ebgp-multihop.config]
+    enabled = true
+    multihop-ttl = 64
+  [neighbors.transport.config]
+    # we initiate the session; no local-address pinning
+    passive-mode = false
+  [[neighbors.afi-safis]]
+    [neighbors.afi-safis.config]
+      afi-safi-name = "ipv4-unicast"
+
+# --- Neighbor: route server, IPv6 feed --------------------------------------
+# The IPv6 transport session carries the full IPv6 table only.
+[[neighbors]]
+  [neighbors.config]
+    neighbor-address = "2001:1a68:2c:2::179"
+    peer-as = 57355
+    description = "AS57355 Bromirski lab route-server (IPv6 feed)"
+  [neighbors.timers.config]
+    keepalive-interval = 3600
+    hold-time = 7200
+  [neighbors.ebgp-multihop.config]
+    enabled = true
+    multihop-ttl = 64
+  [neighbors.transport.config]
+    passive-mode = false
+  [[neighbors.afi-safis]]
+    [neighbors.afi-safis.config]
+      afi-safi-name = "ipv6-unicast"
+
+# --- BMP export to the OpenBMP collector ------------------------------------
+# GoBGP connects OUT to the collector. "obmp-collector" resolves on the shared
+# compose network; port 5000 is the collector's BMP listener.
+# route-monitoring-policy = "pre-policy" exports the Adj-RIB-In (received
+# routes, pre import-policy) -- consistent with the rest of the OpenBMP fleet.
+[[bmp-servers]]
+  [bmp-servers.config]
+    address = "obmp-collector"
+    port = 5000
+    route-monitoring-policy = "pre-policy"
+    statistics-timeout = 3600
diff --git a/gobgp/mrt-refresh.sh b/gobgp/mrt-refresh.sh
new file mode 100755
index 0000000..ec857f5
--- /dev/null
+++ b/gobgp/mrt-refresh.sh
@@ -0,0 +1,104 @@
+#!/usr/bin/env bash
+#
+# mrt-refresh.sh -- MRT full-table fallback loader for the OpenBMP GoBGP feed.
+#
+# Roadmap E1. The live route server (AS57355) is a single volunteer-run host
+# with no SLA. When it is unreachable, the global table in PostgreSQL ip_rib
+# would otherwise age out. This script downloads the latest RouteViews full
+# MRT RIB dump and injects it into the running gobgpd so the table stays warm.
+#
+# Designed to be idempotent and cron-safe at a 2-hour cadence:
+#   - it only downloads a dump it does not already have,
+#   - it only injects when the live route server session is NOT established,
+#   - concurrent runs are guarded by a flock.
+#
+# Run it INSIDE the gobgp container (it shells out to the local `gobgp` CLI):
+#   docker exec obmp-gobgp /config/mrt-refresh.sh
+#
+# Example crontab entry on the docker host (every 2 hours):
+#   0 */2 * * * docker exec obmp-gobgp /config/mrt-refresh.sh >> /var/log/gobgp-mrt.log 2>&1
+
+set -euo pipefail
+
+# --- tunables ---------------------------------------------------------------
+MRT_DIR="${MRT_DIR:-/config/mrt}"
+RV_BASE="${RV_BASE:-https://archive.routeviews.org/route-views/bgpdata}"
+GOBGP="${GOBGP:-gobgp}"
+LOCKFILE="${LOCKFILE:-/tmp/gobgp-mrt-refresh.lock}"
+# RouteViews publishes a full RIB dump every 2 hours; dumps land a few minutes
+# after the even hour, so we look back a safe margin.
+LOOKBACK_HOURS="${LOOKBACK_HOURS:-4}"
+
+log() { echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] $*"; }
+
+# --- single-instance guard --------------------------------------------------
+exec 9>"${LOCKFILE}"
+if ! flock -n 9; then
+  log "another mrt-refresh run is in progress; exiting"
+  exit 0
+fi
+
+mkdir -p "${MRT_DIR}"
+
+# --- skip if the live route server is up ------------------------------------
+# If any AS57355 neighbor is Established, the live feed is authoritative and
+# we must NOT inject a stale MRT dump on top of it.
+if ${GOBGP} neighbor 2>/dev/null | grep -qiE 'Establ'; then
+  log "a BGP session is Established; live feed is healthy, skipping MRT inject"
+  exit 0
+fi
+log "no Established BGP session; proceeding with MRT fallback"
+
+# --- locate the most recent available RIB dump -----------------------------
+# RouteViews RIB dumps:
+#   <RV_BASE>/YYYY.MM/RIBS/rib.YYYYMMDD.HHMM.bz2
+# RIB dumps are taken at even hours (00,02,04,...,22) UTC.
+found_url=""
+found_file=""
+now_epoch="$(date -u +%s)"
+for ((h = 0; h <= LOOKBACK_HOURS; h++)); do
+  ts_epoch=$(( now_epoch - h * 3600 ))
+  hh="$(date -u -d "@${ts_epoch}" +%H)"
+  # only even hours carry RIB dumps
+  if (( 10#${hh} % 2 != 0 )); then
+    continue
+  fi
+  ym="$(date -u -d "@${ts_epoch}" +%Y.%m)"
+  ymd="$(date -u -d "@${ts_epoch}" +%Y%m%d)"
+  fname="rib.${ymd}.${hh}00.bz2"
+  url="${RV_BASE}/${ym}/RIBS/${fname}"
+  if curl -fsI --max-time 30 "${url}" >/dev/null 2>&1; then
+    found_url="${url}"
+    found_file="${fname}"
+    break
+  fi
+done
+
+if [[ -z "${found_url}" ]]; then
+  log "ERROR: no RouteViews RIB dump found within ${LOOKBACK_HOURS}h lookback"
+  exit 1
+fi
+
+dest="${MRT_DIR}/${found_file}"
+
+# --- download (idempotent) --------------------------------------------------
+if [[ -s "${dest}" ]]; then
+  log "already have ${found_file}; reusing cached copy"
+else
+  log "downloading ${found_url}"
+  tmp="${dest}.partial.$$"
+  curl -fsSL --max-time 600 -o "${tmp}" "${found_url}"
+  mv -f "${tmp}" "${dest}"
+  log "downloaded $(du -h "${dest}" | cut -f1) -> ${dest}"
+fi
+
+# --- inject into the running gobgpd -----------------------------------------
+# `gobgp mrt inject global` reads the bz2 dump directly and installs every
+# prefix into the global RIB; BMP export to the collector follows automatically.
+log "injecting ${found_file} into gobgpd global RIB"
+${GOBGP} mrt inject global "${dest}"
+log "MRT inject complete"
+
+# --- housekeeping: keep only the 4 most recent dumps ------------------------
+( cd "${MRT_DIR}" && ls -1t rib.*.bz2 2>/dev/null | tail -n +5 | xargs -r rm -f )
+log "done"