Add platform roadmap: multi-lab CML integration and production deployment

Four-track roadmap covering configuration centralization (inventory.yaml),
CML API automation (virl2_client), production ISP deployment (multi-vendor
IOS-XR + Junos), and packaging for distribution.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

cml/build-cml-image.sh

@@ -0,0 +1,53 @@
+#!/bin/bash
+# Build the ExaBGP Docker image and export it for CML 2.9 import.
+#
+# Usage:
+#   ./cml/build-cml-image.sh
+#
+# Output:
+#   /tmp/obmp-exabgp.tar  — upload this to CML via:
+#     Tools > Node and Image Definitions > Image Definitions > Manage Image Uploads
+#
+# After upload, also import the node + image definition YAMLs:
+#   Tools > Node and Image Definitions > Import > cml/exabgp-node-definition.yaml
+#   Tools > Node and Image Definitions > Import > cml/exabgp-image-definition.yaml
+
+set -e
+cd "$(dirname "$0")/.."
+
+echo "=== Building ExaBGP Docker image ==="
+docker build -t obmp-exabgp:latest ./exabgp/
+
+echo ""
+echo "=== Exporting image to /tmp/obmp-exabgp.tar ==="
+docker save -o /tmp/obmp-exabgp.tar obmp-exabgp:latest
+
+echo ""
+echo "=== Image details ==="
+SIZE=$(du -h /tmp/obmp-exabgp.tar | cut -f1)
+echo "  File: /tmp/obmp-exabgp.tar ($SIZE)"
+
+SHA=$(sha256sum /tmp/obmp-exabgp.tar | awk '{print $1}')
+echo "  SHA256: $SHA"
+
+IMAGE_ID=$(docker image inspect obmp-exabgp:latest --format='{{.Id}}')
+echo "  Image ID: $IMAGE_ID"
+
+echo ""
+echo "=== Next steps ==="
+echo "1. Update cml/exabgp-image-definition.yaml with:"
+echo "   sha256: $SHA"
+echo ""
+echo "2. Upload to CML:"
+echo "   a. Tools > Node and Image Definitions > Import"
+echo "      Upload: cml/exabgp-node-definition.yaml"
+echo "   b. Tools > Node and Image Definitions > Import"
+echo "      Upload: cml/exabgp-image-definition.yaml"
+echo "   c. Tools > Node and Image Definitions > Image Definitions > Manage Image Uploads"
+echo "      Upload: /tmp/obmp-exabgp.tar"
+echo ""
+echo "3. In your CML lab topology:"
+echo "   a. Drag 'ExaBGP Route Injector' from the node palette"
+echo "   b. Draw links to CORE-01 and CORE-02"
+echo "   c. Edit the boot.sh in the node config to set correct IPs"
+echo "   d. Start the node"

cml/build-xrd-image.sh

@@ -0,0 +1,62 @@
+#!/bin/bash
+# Export the XRd control-plane Docker image for CML 2.9 import.
+#
+# Usage:
+#   ./cml/build-xrd-image.sh
+#
+# The XRd image already exists locally (ios-xr/xrd-control-plane:25.1.1).
+# This script just exports it to a .tar file for CML upload.
+
+set -e
+
+IMAGE="ios-xr/xrd-control-plane:25.1.1"
+OUTPUT="/tmp/xrd-control-plane.tar"
+
+echo "=== Verifying XRd image exists ==="
+if ! docker image inspect "$IMAGE" >/dev/null 2>&1; then
+    echo "ERROR: Image $IMAGE not found locally."
+    echo "Check with: docker images | grep xrd"
+    exit 1
+fi
+
+echo "  Image: $IMAGE"
+SIZE=$(docker image inspect "$IMAGE" --format='{{.Size}}' | numfmt --to=iec 2>/dev/null || echo "unknown")
+echo "  Size: $SIZE"
+
+echo ""
+echo "=== Exporting image to $OUTPUT ==="
+echo "  (This may take a minute for ~1.3GB image...)"
+docker save -o "$OUTPUT" "$IMAGE"
+
+echo ""
+echo "=== Export complete ==="
+TAR_SIZE=$(du -h "$OUTPUT" | cut -f1)
+echo "  File: $OUTPUT ($TAR_SIZE)"
+
+SHA=$(sha256sum "$OUTPUT" | awk '{print $1}')
+echo "  SHA256: $SHA"
+
+echo ""
+echo "=== Next steps ==="
+echo "1. Update cml/xrd-image-definition.yaml with:"
+echo "   sha256: $SHA"
+echo ""
+echo "2. Upload to CML:"
+echo "   a. Tools > Node and Image Definitions > Import"
+echo "      Upload: cml/xrd-node-definition.yaml"
+echo "   b. Tools > Node and Image Definitions > Import"
+echo "      Upload: cml/xrd-image-definition.yaml"
+echo "   c. Tools > Node and Image Definitions > Image Definitions > Manage Image Uploads"
+echo "      Upload: $OUTPUT"
+echo "      (For large files, consider SCP to CML server instead)"
+echo ""
+echo "3. In your CML lab topology:"
+echo "   a. Drag 'XRd Control-Plane (IOS-XR)' from the node palette"
+echo "   b. Draw links to CORE-01 (→Gi0/0/0/0) and CORE-02 (→Gi0/0/0/1)"
+echo "   c. Edit xrd-startup.cfg if needed (IPs, BMP target, etc.)"
+echo "   d. Start the node (allow ~3-5 min for XRd boot)"
+echo ""
+echo "4. After boot, verify via XRd console:"
+echo "   show isis adjacency"
+echo "   show bgp summary"
+echo "   show bmp server 1"

cml/exabgp-image-definition.yaml

@@ -0,0 +1,10 @@
+id: obmp-exabgp.latest
+node_definition_id: obmp-exabgp
+description: |-
+  OpenBMP ExaBGP Route Injector
+  Python 3.11 + ExaBGP + Flask API for BGP route injection testing.
+label: ExaBGP Route Injector
+disk_image: obmp-exabgp.tar
+read_only: false
+schema_version: 0.0.1
+# sha256: <UPDATE after running: sha256sum /tmp/obmp-exabgp.tar>

cml/exabgp-node-definition.yaml

@@ -0,0 +1,112 @@
+id: obmp-exabgp
+boot:
+  timeout: 60
+  completed:
+    - "ExaBGP Route Injector"
+  uses_regex: false
+sim:
+  linux_native:
+    libvirt_domain_driver: docker
+    driver: ubuntu
+    ram: 512
+    cpus: 1
+    cpu_limit: 100
+    video:
+      memory: 1
+general:
+  nature: server
+  description: OpenBMP ExaBGP Route Injector (Docker container)
+  read_only: false
+configuration:
+  generator:
+    driver: null
+  provisioning:
+    files:
+      - editable: false
+        name: config.json
+        content: |-
+          {
+            "docker": {
+              "image": "obmp-exabgp:latest",
+              "mounts": [
+                "type=bind,source=cfg/boot.sh,target=/cml-boot.sh"
+              ],
+              "misc_args": [],
+              "env": [
+                "EXABGP_LOCAL_AS=65100",
+                "EXABGP_PEER_AS=65020",
+                "EXABGP_API_PORT=5050"
+              ]
+            },
+            "shell": "/bin/bash",
+            "day0cmd": [ "/bin/bash", "/cml-boot.sh" ],
+            "busybox": false
+          }
+      - editable: true
+        name: boot.sh
+        content: |-
+          #!/bin/bash
+          # CML boot script for ExaBGP container
+          # Configures data-plane interfaces before starting ExaBGP
+          #
+          # Interface mapping (assigned by CML topology links):
+          #   eth0 = first connected interface (data-plane link 1)
+          #   eth1 = second connected interface (data-plane link 2)
+          #   ...additional interfaces as connected in topology
+          #
+          # Edit the IPs below to match your topology addressing.
+          # These are examples using 10.120.x.x/30 point-to-point links.
+
+          # --- Data-plane interface configuration ---
+          # Link to CORE-01: ExaBGP=10.120.1.2/30, CORE-01=10.120.1.1/30
+          ip address add 10.120.1.2/30 dev eth0
+          ip link set dev eth0 up
+
+          # Link to CORE-02: ExaBGP=10.120.2.2/30, CORE-02=10.120.2.1/30
+          ip address add 10.120.2.2/30 dev eth1
+          ip link set dev eth1 up
+
+          # --- Set environment for ExaBGP peering ---
+          export EXABGP_LOCAL_IP=10.120.1.2
+          export EXABGP_PEER_1=10.120.1.1
+          export EXABGP_PEER_2=10.120.2.1
+
+          # --- Start ExaBGP ---
+          exec /bin/bash /exabgp/startup.sh
+    media_type: raw
+    volume_name: cfg
+device:
+  interfaces:
+    serial_ports: 1
+    physical:
+      - eth0
+      - eth1
+      - eth2
+      - eth3
+    has_loopback_zero: false
+    default_count: 2
+ui:
+  label_prefix: exabgp-
+  icon: server
+  label: ExaBGP Route Injector
+  visible: true
+  group: Others
+  description: |-
+    OpenBMP ExaBGP Route Injector
+    BGP route injection for OpenBMP testing.
+    AS 65100 (eBGP) peering with IOS-XR routers (AS 65020).
+    Flask API on port 5050 for route management.
+inherited:
+  image:
+    ram: true
+    cpus: false
+    data_volume: false
+    boot_disk_size: false
+    cpu_limit: false
+  node:
+    ram: true
+    cpus: false
+    data_volume: false
+    boot_disk_size: false
+    cpu_limit: false
+schema_version: 0.0.1

cml/xrd-image-definition.yaml

@@ -0,0 +1,10 @@
+id: xrd-control-plane.25.1.1
+node_definition_id: xrd-control-plane-rr
+description: |-
+  Cisco XRd Control-Plane 25.1.1
+  IOS-XR containerized routing daemon for BGP/IS-IS/BMP workloads.
+label: XRd Control-Plane 25.1.1
+disk_image: xrd-control-plane.tar
+read_only: false
+schema_version: 0.0.1
+# sha256: <UPDATE after running: sha256sum /tmp/xrd-control-plane.tar>

cml/xrd-node-definition.yaml

@@ -0,0 +1,179 @@
+id: xrd-control-plane-rr
+boot:
+  timeout: 300
+  completed:
+    - "IOS XR RUN"
+  uses_regex: false
+sim:
+  linux_native:
+    libvirt_domain_driver: docker
+    driver: ubuntu
+    ram: 2048
+    cpus: 2
+    cpu_limit: 100
+    video:
+      memory: 1
+general:
+  nature: router
+  description: Cisco XRd Control-Plane - IOS-XR containerized routing daemon
+  read_only: false
+configuration:
+  generator:
+    driver: null
+  provisioning:
+    files:
+      - editable: false
+        name: config.json
+        content: |-
+          {
+            "docker": {
+              "image": "ios-xr/xrd-control-plane:25.1.1",
+              "mounts": [
+                "type=bind,source=cfg/boot.sh,target=/cml-boot.sh",
+                "type=bind,source=cfg/xrd-startup.cfg,target=/etc/xrd/startup.cfg"
+              ],
+              "misc_args": [
+                "--privileged"
+              ],
+              "env": [
+                "XR_STARTUP_CFG=/etc/xrd/startup.cfg",
+                "XR_MGMT_INTERFACES=linux:eth0,chksum",
+                "XR_INTERFACES=linux:eth1,xr_name=Gi0/0/0/0;linux:eth2,xr_name=Gi0/0/0/1;linux:eth3,xr_name=Gi0/0/0/2;linux:eth4,xr_name=Gi0/0/0/3"
+              ]
+            },
+            "shell": "/bin/bash",
+            "day0cmd": [ "/bin/bash", "/cml-boot.sh" ],
+            "busybox": false
+          }
+      - editable: true
+        name: boot.sh
+        content: |-
+          #!/bin/bash
+          # CML boot wrapper for XRd control-plane.
+          # XRd handles its own init — this script configures
+          # data-plane interfaces before XRd starts.
+          #
+          # Interface mapping (set via XR_INTERFACES env var):
+          #   eth0  = MgmtEth0/RP0/CPU0/0 (CML mgmt)
+          #   eth1  = Gi0/0/0/0 (data-plane link 1, e.g. to CORE-01)
+          #   eth2  = Gi0/0/0/1 (data-plane link 2, e.g. to CORE-02)
+          #   eth3+ = Gi0/0/0/2+ (additional links)
+          #
+          # Linux-level IP config is handled by XRd via startup.cfg.
+          # Just ensure interfaces are up.
+          for iface in eth0 eth1 eth2 eth3 eth4; do
+            [ -d /sys/class/net/$iface ] && ip link set dev $iface up
+          done
+
+          # XRd entrypoint
+          exec /usr/sbin/xrd
+      - editable: true
+        name: xrd-startup.cfg
+        content: |-
+          !! XRd Control-Plane - Third Route Reflector (RR3)
+          !! Peers with CORE-01 and CORE-02 as RR mesh (non-client iBGP)
+          !! Sends BMP to OpenBMP collector at 10.40.40.202:5000
+          !!
+          hostname XRd-RR3
+          !
+          interface Loopback0
+           ipv4 address 10.10.255.30 255.255.255.255
+          !
+          interface Gi0/0/0/0
+           description to-CORE-01
+           ipv4 address 10.120.3.2 255.255.255.252
+           no shutdown
+          !
+          interface Gi0/0/0/1
+           description to-CORE-02
+           ipv4 address 10.120.4.2 255.255.255.252
+           no shutdown
+          !
+          router isis 1
+           is-type level-2-only
+           net 49.0001.0100.1000.0030.00
+           address-family ipv4 unicast
+            metric-style wide
+           !
+           interface Loopback0
+            passive
+            address-family ipv4 unicast
+            !
+           !
+           interface Gi0/0/0/0
+            point-to-point
+            address-family ipv4 unicast
+            !
+           !
+           interface Gi0/0/0/1
+            point-to-point
+            address-family ipv4 unicast
+            !
+           !
+          !
+          router bgp 65020
+           bgp router-id 10.10.255.30
+           address-family ipv4 unicast
+           !
+           neighbor 10.10.255.0
+            remote-as 65020
+            update-source Loopback0
+            address-family ipv4 unicast
+            !
+           !
+           neighbor 10.10.255.20
+            remote-as 65020
+            update-source Loopback0
+            address-family ipv4 unicast
+            !
+           !
+          !
+          bmp server 1
+           host 10.40.40.202 port 5000
+           description OpenBMP
+           update-source Gi0/0/0/0
+           flapping-delay 60
+           initial-delay 5
+           stats-reporting-period 300
+           initial-refresh delay 30 spread 2
+          !
+          ssh server v2
+          end
+    media_type: raw
+    volume_name: cfg
+device:
+  interfaces:
+    serial_ports: 1
+    physical:
+      - eth0
+      - eth1
+      - eth2
+      - eth3
+      - eth4
+    has_loopback_zero: false
+    default_count: 3
+ui:
+  label_prefix: xrd-
+  icon: router
+  label: XRd Control-Plane (IOS-XR)
+  visible: true
+  group: Cisco
+  description: |-
+    Cisco XRd Control-Plane (IOS-XR 25.1.1)
+    Containerized IOS-XR routing daemon for control-plane workloads.
+    Full BGP, IS-IS, BMP, NETCONF support.
+    Configured as third Route Reflector (RR3) with BMP to OpenBMP.
+inherited:
+  image:
+    ram: true
+    cpus: true
+    data_volume: false
+    boot_disk_size: false
+    cpu_limit: false
+  node:
+    ram: true
+    cpus: true
+    data_volume: false
+    boot_disk_size: false
+    cpu_limit: false
+schema_version: 0.0.1

docker-compose.yml

@@ -92,7 +92,13 @@ services:
       - ${OBMP_DATA_ROOT}/grafana/provisioning:/etc/grafana/provisioning/
     environment:
       - GF_SECURITY_ADMIN_PASSWORD=openbmp
-      - GF_AUTH_ANONYMOUS_ENABLED=true
+      - GF_AUTH_ANONYMOUS_ENABLED=false
+      - GF_SERVER_ROOT_URL=https://bmp.apodacalab.com/grafana/
+      - GF_SERVER_SERVE_FROM_SUB_PATH=true
+      - GF_AUTH_PROXY_ENABLED=true
+      - GF_AUTH_PROXY_HEADER_NAME=Remote-User
+      - GF_AUTH_PROXY_HEADER_PROPERTY=username
+      - GF_AUTH_PROXY_AUTO_SIGN_UP=true
       - GF_USERS_HOME_PAGE=d/obmp-home/obmp-home
       - GF_INSTALL_PLUGINS=agenty-flowcharting-panel,grafana-piechart-panel,grafana-worldmap-panel,grafana-simple-json-datasource,vonage-status-panel
 
@@ -275,8 +281,9 @@ services:
       - NET_RAW
       - NET_ADMIN
     environment:
-      - TRAFFIC_GEN_API_PORT=5051
+      - TRAFFIC_GEN_PORT=5051
       - TRAFFIC_GEN_MODE=sender
+      - RESPONDER_URL=http://172.30.0.10:5053
 
   traffic-gen-ui:
     restart: unless-stopped
@@ -287,6 +294,26 @@ services:
     network_mode: host
     # Serves on port 5002 (host network, defined in nginx.conf)
 
+  traffic-gen-responder:
+    restart: unless-stopped
+    container_name: obmp-traffic-gen-responder
+    build:
+      context: ./traffic-gen
+      dockerfile: Dockerfile
+    cap_add:
+      - NET_RAW
+      - NET_ADMIN
+    environment:
+      - TRAFFIC_GEN_PORT=5053
+      - TRAFFIC_GEN_MODE=responder
+      - TRAFFIC_GEN_RESPONDER_MODE=echo
+      - TRAFFIC_GEN_INTERFACE=eth0
+    networks:
+      traffic-test-net:
+        ipv4_address: 172.30.0.10
+    ports:
+      - "5053:5053"
+
   whois:
     restart: unless-stopped
     container_name: obmp-whois
@@ -305,3 +332,30 @@ services:
       - POSTGRES_DB=openbmp
       - POSTGRES_HOST=obmp-psql
       - POSTGRES_PORT=5432
+
+  authelia:
+    restart: unless-stopped
+    container_name: obmp-authelia
+    image: authelia/authelia:4.38
+    ports:
+      - "9091:9091"
+    volumes:
+      - ${OBMP_DATA_ROOT}/authelia:/config
+    environment:
+      - TZ=UTC
+
+  portal:
+    restart: unless-stopped
+    container_name: obmp-portal
+    image: nginx:alpine
+    ports:
+      - "8080:80"
+    volumes:
+      - ./portal:/usr/share/nginx/html:ro
+
+networks:
+  traffic-test-net:
+    driver: bridge
+    ipam:
+      config:
+        - subnet: 172.30.0.0/24

docs/ROADMAP.md

@@ -0,0 +1,269 @@
+# OpenBMP Platform Roadmap
+
+## Context
+
+This BMP monitoring platform is being developed against CML virtual labs (IOS-XR) and will be deployed into an ISP production network running IOS-XR and Juniper routers/route reflectors. The two tracks share a common foundation: configuration must be environment-agnostic so the same stack runs identically against virtual or production routers.
+
+Currently, router IPs, AS numbers, and credentials are hardcoded across 8+ files, tightly coupling the stack to a single CML lab. This roadmap addresses both the multi-lab development workflow and production deployment.
+
+---
+
+## Track A: Configuration Centralization (Foundation for Both Tracks)
+
+### A1. Create `inventory.yaml` — unified topology inventory
+
+**File**: `inventory.yaml` (new)
+
+Single source of truth for all environments. Structure:
+
+```yaml
+platform:
+  host_ip: 10.40.40.202
+  bmp_port: 5000
+  exabgp_port: 5050
+
+environments:
+  cml-lab1:
+    type: cml            # cml | production
+    description: "CML RR cluster - 9 IOS-XR virtual routers"
+    cml_server: "https://10.40.40.174"
+    cml_user: webui
+    bgp_as: 65020
+    netconf: { user: webui, password: cisco, port: 830 }
+    exabgp:
+      local_as: 65100
+      peers:
+        - { ip: 10.100.0.100, name: CORE-01, peer_as: 65020 }
+        - { ip: 10.100.0.200, name: CORE-02, peer_as: 65020 }
+    routers:
+      CORE-01:  { mgmt: 10.100.0.100, loopback: 10.10.255.0, role: rr, vendor: iosxr, gnmi: true }
+      CORE-02:  { mgmt: 10.100.0.200, loopback: 10.10.255.20, role: rr, vendor: iosxr, gnmi: true }
+      R9K-01:   { mgmt: 10.100.0.1, loopback: 10.10.255.1, role: client, vendor: iosxr }
+      # ...
+
+  cml-lab2:
+    type: cml
+    description: "Second CML Lab (TBD topology)"
+    cml_server: "https://<lab2-ip>"
+    routers: {}
+
+  production:
+    type: production
+    description: "ISP production network"
+    bgp_as: <prod-as>
+    netconf: { user: <prod-user>, port: 830 }
+    routers:
+      # IOS-XR and Juniper RRs + routers
+      PROD-RR1: { mgmt: x.x.x.x, role: rr, vendor: iosxr, gnmi: true }
+      PROD-RR2: { mgmt: x.x.x.x, role: rr, vendor: junos }
+      # ...
+```
+
+Key design decisions:
+- `vendor: iosxr | junos` — drives NETCONF dialect, gNMI paths, and config templates
+- `type: cml | production` — CML environments have `cml_server` for API automation; production does not
+- Credentials in `inventory.yaml` (gitignored) or pulled from env vars
+
+### A2. Create `config_loader.py` — Python inventory helper
+
+**File**: `config_loader.py` (new)
+
+Functions: `get_env(name)`, `get_all_routers()`, `get_routers_by_vendor(vendor)`, `get_exabgp_peers()`, `get_gnmi_targets()`, `get_routers_for_env(env_name)`
+
+### A3. Refactor hardcoded Python scripts
+
+Replace `ROUTERS` dicts/lists with `config_loader` calls:
+- `exabgp/route_diversity_config.py` (line 47)
+- `exabgp/bgpls_config.py` (line 35)
+- `gnmi/gnmi_grpc_config.py` (line 25)
+
+### A4. Expand `.env` and parameterize `docker-compose.yml`
+
+Add to `.env`:
+```env
+OBMP_DATA_ROOT=/var/openbmp
+DOCKER_HOST_IP=10.40.40.202
+EXABGP_LOCAL_IP=10.40.40.202
+EXABGP_LOCAL_AS=65100
+EXABGP_PEER_AS=65020
+EXABGP_PEER_1=10.100.0.100
+EXABGP_PEER_2=10.100.0.200
+```
+
+Replace hardcoded IPs in `docker-compose.yml` (Kafka listener, ExaBGP env vars).
+
+### A5. Telegraf config parameterization
+
+Replace hardcoded gNMI addresses in `telegraf/telegraf.conf` with env var substitution. Pass `GNMI_TARGETS` from docker-compose.yml.
+
+### A6. Fix InfluxDB datasource URL
+
+`obmp-grafana/provisioning/datasources/influxdb-ds.yml`: replace `http://10.40.40.202:8086` with `http://obmp-influxdb:8086`.
+
+---
+
+## Track B: Multi-Lab CML Development
+
+### B1. Dynamic ExaBGP multi-peer support
+
+**File**: `exabgp/startup.sh`
+
+Accept `EXABGP_PEERS` env var (comma-separated `ip:as:description`), generate N neighbor blocks. Keep `PEER_1`/`PEER_2` fallback.
+
+### B2. CML API client module
+
+**File**: `cml/cml_client.py` (new)
+
+Python module using `virl2_client` SDK:
+- Connect to CML server (creds from `inventory.yaml`)
+- Upload node/image definitions
+- Import/export topology YAML
+- Start/stop/destroy labs
+- Get node status
+
+### B3. Topology template system
+
+**File**: `cml/templates/xrd_rr.j2` (new)
+
+Jinja2 templates for XRd startup config. Parameterize: hostname, loopback, link IPs, IS-IS NET, BGP AS, neighbor IPs, BMP target.
+
+### B4. CLI deployment tool
+
+**File**: `cml/deploy.py` (new)
+
+```bash
+python3 cml/deploy.py --env cml-lab1 status
+python3 cml/deploy.py --env cml-lab1 upload-images
+python3 cml/deploy.py --env cml-lab2 create
+python3 cml/deploy.py --env cml-lab2 start
+python3 cml/deploy.py --env cml-lab2 destroy
+```
+
+### B5. Update build scripts with API push
+
+`cml/build-cml-image.sh` and `cml/build-xrd-image.sh` get `--push <env-name>` flag.
+
+---
+
+## Track C: Production ISP Deployment
+
+### C1. Multi-vendor NETCONF support
+
+Current scripts assume IOS-XR NETCONF only. For Juniper RRs:
+- `config_loader.py` provides `vendor` field per router
+- NETCONF scripts branch on vendor for dialect differences (`device_params='iosxr'` vs `device_params='junos'`)
+- Route diversity, BGP-LS config scripts get Junos templates alongside IOS-XR
+
+### C2. Multi-vendor gNMI paths
+
+Telegraf gNMI subscriptions currently use OpenConfig paths which work for both IOS-XR and Junos, but:
+- Verify Juniper gNMI support on target hardware
+- Add vendor-specific path overrides in `inventory.yaml` if needed
+- Telegraf can subscribe to multiple targets with different configs via `[[inputs.gnmi]]` blocks
+
+### C3. BMP considerations for production
+
+- BMP collector (port 5000) accepts connections from any router — no changes needed
+- Production routers need BMP config pushed (manual or via NETCONF automation)
+- Consider: separate BMP server IDs per environment for dashboard filtering
+- Juniper BMP config differs from IOS-XR — add Junos BMP config templates
+
+### C4. Dashboard multi-environment awareness
+
+- Add a Grafana template variable for environment filtering (by router name prefix or a tag)
+- Consider a "Network Overview" dashboard that shows all environments side-by-side
+- Existing dashboards work as-is — router dropdowns will show all BMP-reporting routers
+
+### C5. Security hardening for production
+
+- Move credentials out of `inventory.yaml` into environment variables or a secrets manager
+- Authelia config: stronger passwords, TOTP enforcement, session timeouts
+- PostgreSQL: restrict access, enable SSL
+- Kafka: consider authentication if exposed beyond localhost
+- BMP port: firewall to only accept connections from known router management IPs
+
+### C6. Scalability considerations
+
+- Monitor PostgreSQL disk usage and query performance with production-scale RIBs
+- TimescaleDB compression policies for historical data (ip_rib_log, ls_*_log)
+- Kafka topic partitioning if message throughput is high
+- Consider read replicas or materialized views for heavy Grafana queries
+
+---
+
+## Track D: Packaging & Distribution
+
+### D1. Configuration templates
+
+- `inventory.yaml.example` — documented example with placeholder values
+- `.env.example` — all environment variables with descriptions
+
+### D2. Bootstrap script
+
+`setup.sh` that:
+- Creates required directories (`$OBMP_DATA_ROOT/authelia`, etc.)
+- Copies example configs if originals don't exist
+- Validates inventory.yaml syntax
+- Generates Telegraf config from inventory
+
+### D3. Published Docker images
+
+Push custom images to a registry (Docker Hub or GHCR):
+- `obmp-exabgp`
+- `obmp-exabgp-ui`
+- `obmp-traffic-gen`
+- `obmp-traffic-gen-ui`
+- `obmp-portal`
+
+Replace `build:` with `image:` in docker-compose.yml (keep build as override).
+
+### D4. Documentation
+
+- `docs/quickstart.md` — 5-minute setup guide
+- `docs/adding-a-lab.md` — how to add a CML lab environment
+- `docs/production-deployment.md` — production hardening checklist
+- `docs/architecture.md` — system diagram, data flow, port map
+
+---
+
+## Implementation Order
+
+| Priority | Step | Track | Description |
+|----------|------|-------|-------------|
+| 1 | A1 | Foundation | Create `inventory.yaml` |
+| 2 | A2 | Foundation | Create `config_loader.py` |
+| 3 | A3 | Foundation | Refactor hardcoded Python scripts |
+| 4 | A4 | Foundation | Parameterize `.env` + docker-compose |
+| 5 | A5-A6 | Foundation | Telegraf + InfluxDB datasource fixes |
+| 6 | B1 | CML Dev | Dynamic ExaBGP multi-peer |
+| 7 | B2-B4 | CML Dev | CML API client + deploy CLI |
+| 8 | C1 | Production | Multi-vendor NETCONF (Junos support) |
+| 9 | C3 | Production | Junos BMP config templates |
+| 10 | C5 | Production | Security hardening |
+| 11 | D1-D2 | Packaging | Config templates + bootstrap script |
+| 12 | D3 | Packaging | Publish Docker images to registry |
+| 13 | D4 | Packaging | Documentation |
+
+Steps 1-5 (Track A) unblock everything else. Steps 6-7 and 8-10 can proceed in parallel once the foundation is in place.
+
+---
+
+## Verification
+
+1. **Config centralization**: Change a router IP in `inventory.yaml`, verify all scripts pick it up
+2. **ExaBGP multi-peer**: Set 3+ peers, restart, verify BGP sessions establish
+3. **CML API**: `deploy.py --env cml-lab1 status` connects and lists nodes
+4. **BMP multi-source**: Router from lab 2 sends BMP, appears in `SELECT * FROM routers` and Grafana
+5. **Junos support**: NETCONF script connects to a Juniper router, pushes config
+6. **Production dry-run**: Point a test router from the ISP network at the collector, verify end-to-end
+7. **Clean deploy**: Clone repo on a fresh host, run `setup.sh`, `docker compose up`, confirm stack starts
+
+---
+
+## Risks
+
+- **Router name collisions**: Enforce unique hostnames across all environments
+- **Address space overlap**: Each environment needs distinct management subnets
+- **Juniper BMP differences**: Junos BMP implementation may differ in supported tables/TLVs — test early
+- **Production scale**: 500K-route labs are slow; production full tables will stress PostgreSQL more
+- **Credentials in inventory**: Must be gitignored; consider env var fallback for CI/CD

exabgp-ui/src/App.vue

@@ -41,6 +41,7 @@
           <AnnounceForm v-else-if="activeTab === 'inject'" @routes-changed="fetchRoutes" />
           <PeerStatus v-else-if="activeTab === 'peers'" :peers="peers" />
           <ChurnControl v-else-if="activeTab === 'churn'" />
+          <FullTable v-else-if="activeTab === 'full-table'" @routes-changed="fetchRoutes" />
         </div>
       </main>
     </div>
@@ -63,6 +64,7 @@ import RouteTable from './components/RouteTable.vue'
 import AnnounceForm from './components/AnnounceForm.vue'
 import PeerStatus from './components/PeerStatus.vue'
 import ChurnControl from './components/ChurnControl.vue'
+import FullTable from './components/FullTable.vue'
 
 const health = ref(null)
 const routes = ref([])
@@ -75,6 +77,7 @@ const tabs = [
   { id: 'inject', label: 'Inject' },
   { id: 'peers', label: 'Peers' },
   { id: 'churn', label: 'Churn' },
+  { id: 'full-table', label: 'Full Table' },
 ]
 
 async function fetchHealth() {

exabgp-ui/src/api.js

@@ -1,4 +1,4 @@
-const BASE = '/api'
+const BASE = '/exabgp/api'
 
 async function req(method, path, body) {
   const opts = { method, headers: { 'Content-Type': 'application/json' } }
@@ -18,4 +18,7 @@ export const api = {
   announce: payload => req('POST', '/announce', payload),
   withdraw: prefixes => req('POST', '/withdraw', { prefixes }),
   withdrawAll: () => req('POST', '/withdraw/all'),
+  fullTableStart: (count, batchSize) => req('POST', '/full-table/start', { count, batch_size: batchSize }),
+  fullTableStatus: () => req('GET', '/full-table/status'),
+  fullTableStop: () => req('POST', '/full-table/stop'),
 }

exabgp-ui/src/components/FullTable.vue

@@ -0,0 +1,477 @@
+<template>
+  <div class="full-table">
+    <h2 class="section-title">Full Table Injection</h2>
+    <p class="section-desc">
+      Inject a realistic IPv4 routing table into ExaBGP for stress testing.
+      Routes are generated with varied AS paths, prefix lengths, and communities matching real DFZ distribution.
+    </p>
+
+    <div class="config-card">
+      <!-- Level selector -->
+      <div class="form-group">
+        <label>Table Size</label>
+        <div class="level-grid">
+          <button
+            v-for="level in levels"
+            :key="level.count"
+            class="level-btn"
+            :class="{ selected: selectedCount === level.count }"
+            :disabled="injecting"
+            @click="selectedCount = level.count"
+          >
+            <span class="level-count">{{ level.label }}</span>
+            <span class="level-desc">{{ level.desc }}</span>
+          </button>
+        </div>
+      </div>
+
+      <!-- Custom count -->
+      <div class="form-group">
+        <label>Custom Count</label>
+        <input
+          v-model.number="selectedCount"
+          type="number"
+          min="100"
+          max="950000"
+          step="1000"
+          :disabled="injecting"
+          class="mono-input"
+        />
+      </div>
+
+      <!-- Action buttons -->
+      <div class="action-row">
+        <button v-if="!injecting" class="btn-start" @click="startInjection" :disabled="!selectedCount">
+          <span>&#9654;</span> Inject {{ formatNum(selectedCount) }} Routes
+        </button>
+        <button v-else class="btn-stop" @click="stopInjection">
+          <span>&#9632;</span> Stop Injection
+        </button>
+        <button
+          v-if="!injecting && lastCompleted"
+          class="btn-withdraw"
+          @click="withdrawAll"
+          :disabled="withdrawing"
+        >
+          {{ withdrawing ? 'Withdrawing...' : 'Withdraw All' }}
+        </button>
+      </div>
+    </div>
+
+    <!-- Status display -->
+    <div v-if="injecting || statusMsg" class="status-card">
+      <div class="status-header">
+        <span class="status-dot" :class="injecting ? 'dot-active' : 'dot-idle'"></span>
+        <span class="status-text">{{ statusMsg || 'Idle' }}</span>
+      </div>
+
+      <!-- Progress bar -->
+      <div v-if="state.total > 0" class="progress-section">
+        <div class="progress-labels">
+          <span>{{ formatNum(state.injected) }} / {{ formatNum(state.total) }}</span>
+          <span>{{ state.progress_pct || 0 }}%</span>
+        </div>
+        <div class="progress-track">
+          <div class="progress-fill" :style="{ width: (state.progress_pct || 0) + '%' }"></div>
+        </div>
+      </div>
+
+      <!-- Stats row -->
+      <div v-if="state.total > 0" class="stats-row">
+        <div class="stat-item">
+          <span class="stat-label">Rate</span>
+          <span class="stat-val">{{ formatNum(state.rate_pps || 0) }}/s</span>
+        </div>
+        <div class="stat-item">
+          <span class="stat-label">Elapsed</span>
+          <span class="stat-val">{{ state.elapsed_sec || 0 }}s</span>
+        </div>
+        <div class="stat-item">
+          <span class="stat-label">Active Routes</span>
+          <span class="stat-val">{{ formatNum(state.active_routes || 0) }}</span>
+        </div>
+      </div>
+
+      <!-- Error -->
+      <div v-if="state.error" class="inject-error">{{ state.error }}</div>
+    </div>
+  </div>
+</template>
+
+<script setup>
+import { ref, onUnmounted } from 'vue'
+import { api } from '../api.js'
+
+const emit = defineEmits(['routes-changed'])
+
+const levels = [
+  { count: 1000,   label: '1K',   desc: 'Quick test' },
+  { count: 10000,  label: '10K',  desc: 'Light load' },
+  { count: 50000,  label: '50K',  desc: 'Medium load' },
+  { count: 100000, label: '100K', desc: 'Stress test' },
+  { count: 500000, label: '500K', desc: 'Heavy load' },
+  { count: 900000, label: '900K', desc: 'Full DFZ' },
+]
+
+const selectedCount = ref(10000)
+const injecting = ref(false)
+const statusMsg = ref('')
+const lastCompleted = ref(false)
+const withdrawing = ref(false)
+const state = ref({})
+
+let pollTimer = null
+
+function formatNum(n) {
+  if (n == null) return '0'
+  return Number(n).toLocaleString()
+}
+
+async function startInjection() {
+  try {
+    statusMsg.value = 'Starting injection...'
+    injecting.value = true
+    lastCompleted.value = false
+    state.value = {}
+    await api.fullTableStart(selectedCount.value, 1000)
+    startPolling()
+  } catch (e) {
+    statusMsg.value = `Start failed: ${e.message}`
+    injecting.value = false
+  }
+}
+
+async function stopInjection() {
+  try {
+    await api.fullTableStop()
+    statusMsg.value = 'Stop requested...'
+  } catch (e) {
+    statusMsg.value = `Stop failed: ${e.message}`
+  }
+}
+
+async function withdrawAll() {
+  withdrawing.value = true
+  try {
+    const data = await api.withdrawAll()
+    statusMsg.value = `Withdrew ${data.count} routes`
+    lastCompleted.value = false
+    state.value = {}
+    emit('routes-changed')
+  } catch (e) {
+    statusMsg.value = `Withdraw failed: ${e.message}`
+  } finally {
+    withdrawing.value = false
+  }
+}
+
+function startPolling() {
+  stopPolling()
+  pollStatus()
+  pollTimer = setInterval(pollStatus, 2000)
+}
+
+function stopPolling() {
+  if (pollTimer) {
+    clearInterval(pollTimer)
+    pollTimer = null
+  }
+}
+
+async function pollStatus() {
+  try {
+    const data = await api.fullTableStatus()
+    state.value = data
+
+    if (data.active) {
+      statusMsg.value = `Injecting: ${formatNum(data.injected)} / ${formatNum(data.total)} (${data.rate_pps || 0}/s)`
+    } else if (data.error) {
+      statusMsg.value = `Error: ${data.error}`
+      injecting.value = false
+      stopPolling()
+    } else if (data.injected > 0) {
+      statusMsg.value = `Complete: ${formatNum(data.injected)} routes in ${data.elapsed_sec}s (${data.rate_pps}/s)`
+      injecting.value = false
+      lastCompleted.value = true
+      stopPolling()
+      emit('routes-changed')
+    }
+  } catch (e) {
+    // keep polling
+  }
+}
+
+onUnmounted(() => {
+  stopPolling()
+})
+</script>
+
+<style scoped>
+.full-table {
+  display: flex;
+  flex-direction: column;
+  gap: 18px;
+  max-width: 680px;
+}
+
+.section-title {
+  font-size: 14px;
+  font-weight: 600;
+  color: var(--muted);
+  text-transform: uppercase;
+  letter-spacing: 0.08em;
+}
+
+.section-desc {
+  color: var(--muted);
+  font-size: 13px;
+  line-height: 1.6;
+  margin-top: -8px;
+}
+
+.config-card {
+  background: var(--card-bg);
+  border: 1px solid var(--border);
+  border-radius: var(--radius);
+  padding: 20px;
+  display: flex;
+  flex-direction: column;
+  gap: 18px;
+}
+
+.form-group {
+  display: flex;
+  flex-direction: column;
+  gap: 6px;
+}
+
+label {
+  font-size: 12px;
+  font-weight: 600;
+  color: var(--muted);
+  text-transform: uppercase;
+  letter-spacing: 0.05em;
+}
+
+.level-grid {
+  display: grid;
+  grid-template-columns: repeat(3, 1fr);
+  gap: 8px;
+}
+
+.level-btn {
+  display: flex;
+  flex-direction: column;
+  align-items: center;
+  gap: 2px;
+  padding: 12px 8px;
+  background: var(--bg);
+  border: 1px solid var(--border);
+  border-radius: var(--radius);
+  color: var(--text);
+  transition: all 0.15s;
+}
+
+.level-btn:hover:not(:disabled) {
+  border-color: var(--accent);
+  background: rgba(79, 156, 249, 0.08);
+}
+
+.level-btn.selected {
+  border-color: var(--accent);
+  background: rgba(79, 156, 249, 0.15);
+  box-shadow: 0 0 0 1px var(--accent);
+}
+
+.level-count {
+  font-size: 18px;
+  font-weight: 700;
+  font-family: 'Cascadia Code', 'Fira Code', 'Consolas', monospace;
+  color: var(--accent);
+}
+
+.level-desc {
+  font-size: 11px;
+  color: var(--muted);
+}
+
+.mono-input {
+  font-family: 'Cascadia Code', 'Fira Code', 'Consolas', monospace;
+  background: var(--bg);
+  color: var(--text);
+  border: 1px solid var(--border);
+  border-radius: var(--radius);
+  padding: 7px 10px;
+  font-size: 13px;
+  outline: none;
+  max-width: 200px;
+}
+
+.mono-input:focus {
+  border-color: var(--accent);
+}
+
+.action-row {
+  display: flex;
+  gap: 10px;
+  padding-top: 4px;
+  border-top: 1px solid var(--border);
+}
+
+.btn-start {
+  padding: 9px 22px;
+  background: rgba(72, 187, 120, 0.15);
+  color: #48bb78;
+  border: 1px solid rgba(72, 187, 120, 0.3);
+  font-weight: 700;
+  font-size: 14px;
+  display: flex;
+  align-items: center;
+  gap: 7px;
+}
+
+.btn-start:hover:not(:disabled) {
+  background: rgba(72, 187, 120, 0.25);
+}
+
+.btn-stop {
+  padding: 9px 22px;
+  background: rgba(252, 129, 129, 0.15);
+  color: #fc8181;
+  border: 1px solid rgba(252, 129, 129, 0.3);
+  font-weight: 700;
+  font-size: 14px;
+  display: flex;
+  align-items: center;
+  gap: 7px;
+  animation: pulse-border 1.5s ease-in-out infinite;
+}
+
+.btn-stop:hover {
+  background: rgba(252, 129, 129, 0.25);
+}
+
+.btn-withdraw {
+  padding: 9px 18px;
+  background: rgba(246, 173, 85, 0.15);
+  color: #f6ad55;
+  border: 1px solid rgba(246, 173, 85, 0.3);
+  font-weight: 600;
+  font-size: 13px;
+}
+
+.btn-withdraw:hover:not(:disabled) {
+  background: rgba(246, 173, 85, 0.25);
+}
+
+.status-card {
+  background: var(--card-bg);
+  border: 1px solid var(--border);
+  border-radius: var(--radius);
+  padding: 16px 18px;
+  display: flex;
+  flex-direction: column;
+  gap: 12px;
+}
+
+.status-header {
+  display: flex;
+  align-items: center;
+  gap: 10px;
+}
+
+.status-dot {
+  width: 10px;
+  height: 10px;
+  border-radius: 50%;
+  flex-shrink: 0;
+}
+
+.dot-active {
+  background: #48bb78;
+  box-shadow: 0 0 8px #48bb78;
+  animation: pulse-dot 1s ease-in-out infinite;
+}
+
+.dot-idle {
+  background: var(--muted);
+}
+
+.status-text {
+  font-size: 14px;
+  font-weight: 600;
+  color: var(--text);
+  font-family: 'Cascadia Code', 'Fira Code', 'Consolas', monospace;
+}
+
+.progress-section {
+  display: flex;
+  flex-direction: column;
+  gap: 5px;
+}
+
+.progress-labels {
+  display: flex;
+  justify-content: space-between;
+  font-size: 11px;
+  color: var(--muted);
+}
+
+.progress-track {
+  height: 6px;
+  background: var(--border);
+  border-radius: 3px;
+  overflow: hidden;
+}
+
+.progress-fill {
+  height: 100%;
+  background: var(--accent);
+  border-radius: 3px;
+  transition: width 0.5s ease;
+}
+
+.stats-row {
+  display: flex;
+  gap: 24px;
+}
+
+.stat-item {
+  display: flex;
+  flex-direction: column;
+  gap: 2px;
+}
+
+.stat-label {
+  font-size: 10px;
+  color: var(--muted);
+  text-transform: uppercase;
+  letter-spacing: 0.05em;
+}
+
+.stat-val {
+  font-size: 15px;
+  font-weight: 700;
+  color: var(--text);
+  font-family: 'Cascadia Code', 'Fira Code', 'Consolas', monospace;
+}
+
+.inject-error {
+  font-size: 12px;
+  color: #fc8181;
+  padding: 6px 10px;
+  background: rgba(252, 129, 129, 0.08);
+  border-radius: 4px;
+  border: 1px solid rgba(252, 129, 129, 0.2);
+}
+
+@keyframes pulse-dot {
+  0%, 100% { opacity: 1; }
+  50% { opacity: 0.5; }
+}
+
+@keyframes pulse-border {
+  0%, 100% { border-color: rgba(252, 129, 129, 0.3); }
+  50% { border-color: rgba(252, 129, 129, 0.6); }
+}
+</style>

exabgp-ui/vite.config.js

@@ -2,6 +2,7 @@ import { defineConfig } from 'vite'
 import vue from '@vitejs/plugin-vue'
 
 export default defineConfig({
+  base: '/exabgp/',
   plugins: [vue()],
   server: {
     proxy: {

exabgp/api/server.py

@@ -48,11 +48,15 @@ peer_states = {}
 # ExaBGP command helpers
 # ---------------------------------------------------------------------------
 
+_quiet_mode = False
+
+
 def _send(cmd: str):
     """Write a command to ExaBGP via stdout."""
     with _stdout_lock:
         sys.stdout.write(cmd + '\n')
         sys.stdout.flush()
+    if not _quiet_mode:
         log.info('→ ExaBGP: %s', cmd)
 
 
@@ -162,7 +166,22 @@ def api_withdraw_all():
 # ---------------------------------------------------------------------------
 
 sys.path.insert(0, '/exabgp')
-from scenarios import SCENARIOS
+from scenarios import SCENARIOS, generate_full_internet
+
+# ---------------------------------------------------------------------------
+# Full-table background injection
+# ---------------------------------------------------------------------------
+
+_injection_state = {
+    'active': False,
+    'total': 0,
+    'injected': 0,
+    'elapsed_sec': 0,
+    'rate_pps': 0,
+    'error': None,
+    'stop_requested': False,
+}
+_injection_lock = threading.Lock()
 
 
 @app.route('/scenarios', methods=['GET'])
@@ -223,6 +242,131 @@ def get_peers():
     return jsonify({'peers': peer_states})
 
 
+# ---------------------------------------------------------------------------
+# Full-table injection endpoints
+# ---------------------------------------------------------------------------
+
+def _injection_worker(count, batch_size):
+    """Background thread: generate and inject full internet table."""
+    global _quiet_mode
+    try:
+        _quiet_mode = True  # suppress per-route logging
+        log.info('Generating %d full-table prefixes...', count)
+        routes = generate_full_internet(count)
+        with _injection_lock:
+            _injection_state['total'] = len(routes)
+        log.info('Generated %d routes, starting injection at batch_size=%d', len(routes), batch_size)
+
+        start_time = time.time()
+        for i, route in enumerate(routes):
+            with _injection_lock:
+                if _injection_state['stop_requested']:
+                    log.info('Injection stopped by user at %d/%d', i, len(routes))
+                    break
+
+            prefix = route['prefix']
+            announce_route(
+                prefix,
+                next_hop=route.get('next_hop', 'self'),
+                as_path=route.get('as_path', []),
+                communities=route.get('communities', []),
+                med=route.get('med'),
+                local_pref=route.get('local_pref'),
+            )
+
+            # Update progress periodically (every batch_size routes)
+            if (i + 1) % batch_size == 0:
+                elapsed = time.time() - start_time
+                with _injection_lock:
+                    _injection_state['injected'] = i + 1
+                    _injection_state['elapsed_sec'] = round(elapsed, 1)
+                    _injection_state['rate_pps'] = round((i + 1) / elapsed, 1) if elapsed > 0 else 0
+                log.info('Injection progress: %d/%d (%.0f/s)',
+                         i + 1, len(routes), (i + 1) / elapsed if elapsed > 0 else 0)
+
+        elapsed = time.time() - start_time
+        with _injection_lock:
+            _injection_state['injected'] = min(i + 1, len(routes))
+            _injection_state['elapsed_sec'] = round(elapsed, 1)
+            _injection_state['rate_pps'] = round(_injection_state['injected'] / elapsed, 1) if elapsed > 0 else 0
+            _injection_state['active'] = False
+        log.info('Injection complete: %d routes in %.1fs (%.0f/s)',
+                 _injection_state['injected'], elapsed,
+                 _injection_state['injected'] / elapsed if elapsed > 0 else 0)
+
+    except Exception as e:
+        log.error('Injection error: %s', e)
+        with _injection_lock:
+            _injection_state['error'] = str(e)
+            _injection_state['active'] = False
+    finally:
+        _quiet_mode = False
+
+
+@app.route('/full-table/start', methods=['POST'])
+def start_full_table():
+    """Start background injection of a full IPv4 routing table.
+
+    POST body (all optional):
+      count:      Number of prefixes (default 900000)
+      batch_size: Progress update interval (default 1000)
+    """
+    with _injection_lock:
+        if _injection_state['active']:
+            return jsonify({
+                'error': 'Injection already in progress',
+                'state': dict(_injection_state),
+            }), 409
+
+    data = request.get_json(force=True) if request.data else {}
+    count = int(data.get('count', 900000))
+    batch_size = int(data.get('batch_size', 1000))
+
+    with _injection_lock:
+        _injection_state.update({
+            'active': True,
+            'total': count,
+            'injected': 0,
+            'elapsed_sec': 0,
+            'rate_pps': 0,
+            'error': None,
+            'stop_requested': False,
+        })
+
+    t = threading.Thread(target=_injection_worker, args=(count, batch_size), daemon=True)
+    t.start()
+
+    log.info('Started full-table injection: %d prefixes', count)
+    return jsonify({
+        'status': 'started',
+        'count': count,
+        'message': f'Generating and injecting {count} prefixes in background. GET /full-table/status to track progress.',
+    })
+
+
+@app.route('/full-table/status', methods=['GET'])
+def full_table_status():
+    """Get current full-table injection progress."""
+    with _injection_lock:
+        state = dict(_injection_state)
+    if state['total'] > 0:
+        state['progress_pct'] = round(state['injected'] / state['total'] * 100, 1)
+    else:
+        state['progress_pct'] = 0
+    state['active_routes'] = len(active_routes)
+    return jsonify(state)
+
+
+@app.route('/full-table/stop', methods=['POST'])
+def stop_full_table():
+    """Stop an in-progress full-table injection."""
+    with _injection_lock:
+        if not _injection_state['active']:
+            return jsonify({'error': 'No injection in progress'}), 400
+        _injection_state['stop_requested'] = True
+    return jsonify({'status': 'stop_requested', 'injected_so_far': _injection_state['injected']})
+
+
 # ---------------------------------------------------------------------------
 # ExaBGP event loop (main thread)
 # ---------------------------------------------------------------------------

exabgp/inject.py

@@ -12,6 +12,9 @@ Usage:
   inject.py withdraw-all
   inject.py scenario <name>
   inject.py withdraw-scenario <name>
+  inject.py full-table [--count N] [--follow]     # inject full IPv4 table (background)
+  inject.py full-table-status                     # show injection progress
+  inject.py full-table-stop                       # stop injection
   inject.py churn [--count N] [--interval SEC]   # cycle announce/withdraw for ip_rib_log population
   inject.py monitor                               # live-refresh terminal view
 
@@ -29,8 +32,8 @@ import requests
 API = os.environ.get('EXABGP_API', 'http://localhost:5050')
 
 
-def _post(path, data=None):
+def _post(path, data=None, timeout=10):
-    r = requests.post(f'{API}{path}', json=data or {}, timeout=10)
+    r = requests.post(f'{API}{path}', json=data or {}, timeout=timeout)
     r.raise_for_status()
     return r.json()
 
@@ -174,6 +177,101 @@ def cmd_withdraw_scenario(args):
     print(f"Withdrew scenario '{args.name}': {data['count']} routes withdrawn")
 
 
+def cmd_full_table(args):
+    """Inject a full IPv4 routing table for stress testing."""
+    count = args.count
+    print(f"Starting full-table injection: {count} prefixes")
+    print("This generates routes in background. Use 'inject.py full-table-status' to track.\n")
+
+    data = _post('/full-table/start', {'count': count, 'batch_size': args.batch_size}, timeout=120)
+    print(data.get('message', 'Started'))
+
+    if args.follow:
+        print()
+        _follow_injection()
+
+
+def cmd_full_table_status(args):
+    """Show full-table injection progress."""
+    data = _get('/full-table/status')
+    active = data.get('active', False)
+    total = data.get('total', 0)
+    injected = data.get('injected', 0)
+    pct = data.get('progress_pct', 0)
+    rate = data.get('rate_pps', 0)
+    elapsed = data.get('elapsed_sec', 0)
+    error = data.get('error')
+    active_routes = data.get('active_routes', 0)
+
+    if error:
+        print(f"ERROR: {error}")
+    elif active:
+        bar_len = 40
+        filled = int(bar_len * pct / 100)
+        bar = '#' * filled + '-' * (bar_len - filled)
+        print(f"[{bar}] {pct:.1f}%")
+        print(f"  Injected: {injected:,} / {total:,}  ({rate:.0f} routes/s)")
+        print(f"  Elapsed:  {elapsed:.0f}s")
+        print(f"  Active routes in ExaBGP: {active_routes:,}")
+    elif total > 0:
+        print(f"Injection complete: {injected:,} / {total:,} routes in {elapsed:.0f}s ({rate:.0f}/s)")
+        print(f"Active routes in ExaBGP: {active_routes:,}")
+    else:
+        print("No injection running or completed.")
+        print(f"Active routes: {active_routes:,}")
+
+
+def cmd_full_table_stop(args):
+    """Stop an in-progress full-table injection."""
+    try:
+        data = _post('/full-table/stop')
+        print(f"Stop requested. Injected so far: {data.get('injected_so_far', '?'):,}")
+    except requests.exceptions.HTTPError as e:
+        if e.response.status_code == 400:
+            print("No injection in progress.")
+        else:
+            raise
+
+
+def _follow_injection():
+    """Poll injection status until complete."""
+    import shutil
+    lines_printed = 0
+    try:
+        while True:
+            data = _get('/full-table/status')
+            active = data.get('active', False)
+            total = data.get('total', 0)
+            injected = data.get('injected', 0)
+            pct = data.get('progress_pct', 0)
+            rate = data.get('rate_pps', 0)
+            elapsed = data.get('elapsed_sec', 0)
+            active_routes = data.get('active_routes', 0)
+
+            # Move cursor up to overwrite
+            if lines_printed > 0:
+                print(f"\033[{lines_printed}A", end='')
+
+            bar_len = 40
+            filled = int(bar_len * pct / 100)
+            bar = '#' * filled + '-' * (bar_len - filled)
+            output_lines = [
+                f"  [{bar}] {pct:.1f}%",
+                f"  Injected: {injected:,} / {total:,}  ({rate:.0f} routes/s)  elapsed: {elapsed:.0f}s",
+                f"  Active routes: {active_routes:,}",
+            ]
+            print('\n'.join(output_lines))
+            lines_printed = len(output_lines)
+
+            if not active:
+                print(f"\nDone! {injected:,} routes injected in {elapsed:.0f}s")
+                break
+
+            time.sleep(2)
+    except KeyboardInterrupt:
+        print("\n\nFollowing stopped (injection continues in background).")
+
+
 def cmd_churn(args):
     """
     Cycle announce/withdraw on the 'churn' scenario to generate ip_rib_log
@@ -236,6 +334,17 @@ def main():
     p = sub.add_parser('withdraw-scenario', help='Withdraw a named scenario')
     p.add_argument('name')
 
+    p = sub.add_parser('full-table', help='Inject full IPv4 routing table (background)')
+    p.add_argument('--count', type=int, default=900000, metavar='N',
+                   help='Number of prefixes to inject (default: 900000)')
+    p.add_argument('--batch-size', type=int, default=1000, metavar='N',
+                   help='Progress update interval (default: 1000)')
+    p.add_argument('--follow', '-f', action='store_true',
+                   help='Follow progress until complete')
+
+    sub.add_parser('full-table-status', help='Show full-table injection progress')
+    sub.add_parser('full-table-stop', help='Stop full-table injection')
+
     p = sub.add_parser('churn', help='Cycle announce/withdraw to populate ip_rib_log')
     p.add_argument('--count', type=int, default=0, metavar='N',
                    help='Number of cycles (0 = infinite)')
@@ -255,6 +364,9 @@ def main():
         'withdraw-all': cmd_withdraw_all,
         'scenario': cmd_scenario,
         'withdraw-scenario': cmd_withdraw_scenario,
+        'full-table': cmd_full_table,
+        'full-table-status': cmd_full_table_status,
+        'full-table-stop': cmd_full_table_stop,
         'churn': cmd_churn,
     }
 

exabgp/route_diversity_config.py

@@ -0,0 +1,658 @@
+#!/usr/bin/env python3
+"""
+Route Diversity Configuration Script
+=====================================
+Adds loopbacks, static routes, route-policies, and BGP redistribution
+to R9K-01 through R9K-07 to create locally-originated routes that
+produce meaningful RR Loc-RIB diffs between CORE-01 and CORE-02.
+
+IS-IS topology (natural asymmetry — no metric tuning needed):
+  CORE-01 → R9K-01, R9K-02, R9K-03, R9K-04, R9K-05
+  CORE-02 → R9K-05, R9K-06, R9K-07
+  R9K-04 ↔ R9K-06 (cross-link)
+  R9K-05 dual-homed to both COREs
+
+Overlapping prefixes from CORE-01-side and CORE-02-side routers produce
+next-hop diffs because each RR picks the client with lowest IGP cost.
+
+Address plan:
+  Loopbacks:     10.110.{router_id}.{1,2}/32  (unique per router)
+  Overlap LBs:   10.110.{100-103}.1/32        (shared across router pairs)
+  Static routes: 10.111.{router_id}.0/24      (unique per router)
+  Overlap statics: 10.111.{100-103}.0/24      (shared across router pairs)
+
+Usage:
+  python3 route_diversity_config.py              # apply all config
+  python3 route_diversity_config.py --verify-only # just check current state
+  python3 route_diversity_config.py --rollback    # remove all added config
+"""
+
+from ncclient import manager
+import xml.etree.ElementTree as ET
+import sys
+import argparse
+
+# YANG namespaces
+IFMGR_NS   = 'http://cisco.com/ns/yang/Cisco-IOS-XR-ifmgr-cfg'
+IPV4IO_NS  = 'http://cisco.com/ns/yang/Cisco-IOS-XR-ipv4-io-cfg'
+STATIC_NS  = 'http://cisco.com/ns/yang/Cisco-IOS-XR-ip-static-cfg'
+BGP_NS     = 'http://cisco.com/ns/yang/Cisco-IOS-XR-ipv4-bgp-cfg'
+ISIS_NS    = 'http://cisco.com/ns/yang/Cisco-IOS-XR-clns-isis-cfg'
+RPL_NS     = 'http://cisco.com/ns/yang/Cisco-IOS-XR-policy-repository-cfg'
+
+# ──────────────────────────────────────────────────────────────────────
+# Router definitions
+# ──────────────────────────────────────────────────────────────────────
+
+ROUTERS = {
+    'R9K-01': {
+        'mgmt': '10.100.0.1',
+        'loopbacks': [
+            ('Loopback10', '10.110.1.1', '255.255.255.255'),
+            ('Loopback11', '10.110.1.2', '255.255.255.255'),
+            ('Loopback100', '10.110.100.1', '255.255.255.255'),  # overlap with R9K-06
+            ('Loopback103', '10.110.103.1', '255.255.255.255'),  # overlap with R9K-04, R9K-07
+        ],
+        'statics': [
+            ('10.111.1.0', 24, 100),       # unique, tag=100 → LP=200
+            ('10.111.100.0', 24, 100),      # overlap with R9K-06 (tag=200)
+            ('10.111.103.0', 24, 100),      # overlap with R9K-04, R9K-07 (same tag)
+        ],
+    },
+    'R9K-02': {
+        'mgmt': '10.100.0.2',
+        'loopbacks': [
+            ('Loopback10', '10.110.2.1', '255.255.255.255'),
+            ('Loopback11', '10.110.2.2', '255.255.255.255'),
+            ('Loopback101', '10.110.101.1', '255.255.255.255'),  # overlap with R9K-07
+        ],
+        'statics': [
+            ('10.111.2.0', 24, 100),
+            ('10.111.101.0', 24, 100),      # overlap with R9K-07 (tag=300)
+        ],
+    },
+    'R9K-03': {
+        'mgmt': '10.100.0.3',
+        'loopbacks': [
+            ('Loopback10', '10.110.3.1', '255.255.255.255'),
+            ('Loopback11', '10.110.3.2', '255.255.255.255'),
+            ('Loopback102', '10.110.102.1', '255.255.255.255'),  # overlap with R9K-05
+        ],
+        'statics': [
+            ('10.111.3.0', 24, 100),
+            ('10.111.102.0', 24, 100),      # overlap with R9K-04 (tag=200)
+        ],
+    },
+    'R9K-04': {
+        'mgmt': '10.100.0.4',
+        'loopbacks': [
+            ('Loopback10', '10.110.4.1', '255.255.255.255'),
+            ('Loopback11', '10.110.4.2', '255.255.255.255'),
+            ('Loopback103', '10.110.103.1', '255.255.255.255'),  # overlap with R9K-01, R9K-07
+        ],
+        'statics': [
+            ('10.111.4.0', 24, 200),        # tag=200 → LP=150
+            ('10.111.102.0', 24, 200),       # overlap with R9K-03 (tag=100)
+            ('10.111.103.0', 24, 100),       # overlap with R9K-01, R9K-07 (same tag)
+        ],
+    },
+    'R9K-05': {
+        'mgmt': '10.100.0.5',
+        'loopbacks': [
+            ('Loopback10', '10.110.5.1', '255.255.255.255'),
+            ('Loopback11', '10.110.5.2', '255.255.255.255'),
+            ('Loopback102', '10.110.102.1', '255.255.255.255'),  # overlap with R9K-03
+        ],
+        'statics': [
+            ('10.111.5.0', 24, 100),
+        ],
+    },
+    'R9K-06': {
+        'mgmt': '10.100.0.6',
+        'loopbacks': [
+            ('Loopback10', '10.110.6.1', '255.255.255.255'),
+            ('Loopback11', '10.110.6.2', '255.255.255.255'),
+            ('Loopback100', '10.110.100.1', '255.255.255.255'),  # overlap with R9K-01
+        ],
+        'statics': [
+            ('10.111.6.0', 24, 200),        # tag=200 → LP=150
+            ('10.111.100.0', 24, 200),       # overlap with R9K-01 (tag=100)
+        ],
+    },
+    'R9K-07': {
+        'mgmt': '10.100.0.7',
+        'loopbacks': [
+            ('Loopback10', '10.110.7.1', '255.255.255.255'),
+            ('Loopback11', '10.110.7.2', '255.255.255.255'),
+            ('Loopback101', '10.110.101.1', '255.255.255.255'),  # overlap with R9K-02
+            ('Loopback103', '10.110.103.1', '255.255.255.255'),  # overlap with R9K-01, R9K-04
+        ],
+        'statics': [
+            ('10.111.7.0', 24, 300),         # tag=300 → LP=100
+            ('10.111.101.0', 24, 300),        # overlap with R9K-02 (tag=100)
+            ('10.111.103.0', 24, 100),        # overlap with R9K-01, R9K-04 (same tag)
+        ],
+    },
+}
+
+# ──────────────────────────────────────────────────────────────────────
+# Route-policy (RPL text blob)
+# ──────────────────────────────────────────────────────────────────────
+
+ROUTE_POLICY_NAME = 'REDIST-TO-BGP'
+ROUTE_POLICY_BODY = """\
+route-policy REDIST-TO-BGP
+  if tag is 100 then
+    set local-preference 200
+    set med 50
+    set community (65020:100) additive
+    pass
+  elseif tag is 200 then
+    set local-preference 150
+    set med 100
+    set community (65020:200) additive
+    pass
+  elseif tag is 300 then
+    set local-preference 100
+    set med 200
+    set community (65020:300) additive
+    pass
+  else
+    set local-preference 100
+    pass
+  endif
+end-policy
+"""
+
+# ──────────────────────────────────────────────────────────────────────
+# XML builders
+# ──────────────────────────────────────────────────────────────────────
+
+def loopback_xml(name, addr, mask):
+    """Create a loopback interface with an IPv4 address."""
+    return f"""
+<config>
+  <interface-configurations xmlns="{IFMGR_NS}">
+    <interface-configuration>
+      <active>act</active>
+      <interface-name>{name}</interface-name>
+      <interface-virtual/>
+      <ipv4-network xmlns="{IPV4IO_NS}">
+        <addresses>
+          <primary>
+            <address>{addr}</address>
+            <netmask>{mask}</netmask>
+          </primary>
+        </addresses>
+      </ipv4-network>
+    </interface-configuration>
+  </interface-configurations>
+</config>
+"""
+
+
+def static_route_xml(prefix, prefix_len, tag):
+    """Create a static route to Null0 with a tag."""
+    return f"""
+<config>
+  <router-static xmlns="{STATIC_NS}">
+    <default-vrf>
+      <address-family>
+        <vrfipv4>
+          <vrf-unicast>
+            <vrf-prefixes>
+              <vrf-prefix>
+                <prefix>{prefix}</prefix>
+                <prefix-length>{prefix_len}</prefix-length>
+                <vrf-route>
+                  <vrf-next-hop-table>
+                    <vrf-next-hop-interface-name>
+                      <interface-name>Null0</interface-name>
+                      <tag>{tag}</tag>
+                    </vrf-next-hop-interface-name>
+                  </vrf-next-hop-table>
+                </vrf-route>
+              </vrf-prefix>
+            </vrf-prefixes>
+          </vrf-unicast>
+        </vrfipv4>
+      </address-family>
+    </default-vrf>
+  </router-static>
+</config>
+"""
+
+
+def route_policy_xml(name, body):
+    """Create/replace a route-policy (RPL text blob)."""
+    return f"""
+<config>
+  <routing-policy xmlns="{RPL_NS}">
+    <route-policies>
+      <route-policy>
+        <route-policy-name>{name}</route-policy-name>
+        <rpl-route-policy>{body}</rpl-route-policy>
+      </route-policy>
+    </route-policies>
+  </routing-policy>
+</config>
+"""
+
+
+def isis_passive_xml(intf_name):
+    """Add a loopback to IS-IS instance 1 (passive by default for loopbacks)."""
+    return f"""
+<config>
+  <isis xmlns="{ISIS_NS}">
+    <instances>
+      <instance>
+        <instance-name>1</instance-name>
+        <interfaces>
+          <interface>
+            <interface-name>{intf_name}</interface-name>
+            <running/>
+            <interface-afs>
+              <interface-af>
+                <af-name>ipv4</af-name>
+                <saf-name>unicast</saf-name>
+                <interface-af-data/>
+              </interface-af>
+            </interface-afs>
+          </interface>
+        </interfaces>
+      </instance>
+    </instances>
+  </isis>
+</config>
+"""
+
+
+def bgp_redistribute_xml():
+    """Configure redistribute connected + static with REDIST-TO-BGP policy."""
+    return f"""
+<config>
+  <bgp xmlns="{BGP_NS}">
+    <instance>
+      <instance-name>default</instance-name>
+      <instance-as>
+        <as>0</as>
+        <four-byte-as>
+          <as>65020</as>
+          <bgp-running/>
+          <default-vrf>
+            <global>
+              <global-afs>
+                <global-af>
+                  <af-name>ipv4-unicast</af-name>
+                  <enable/>
+                  <connected-routes>
+                    <route-policy-name>{ROUTE_POLICY_NAME}</route-policy-name>
+                  </connected-routes>
+                  <static-routes>
+                    <route-policy-name>{ROUTE_POLICY_NAME}</route-policy-name>
+                  </static-routes>
+                </global-af>
+              </global-afs>
+            </global>
+          </default-vrf>
+        </four-byte-as>
+      </instance-as>
+    </instance>
+  </bgp>
+</config>
+"""
+
+
+# ──────────────────────────────────────────────────────────────────────
+# Rollback XML builders (delete operations)
+# ──────────────────────────────────────────────────────────────────────
+
+NC_NS = 'urn:ietf:params:xml:ns:netconf:base:1.0'
+
+
+def delete_loopback_xml(name):
+    return f"""
+<config>
+  <interface-configurations xmlns="{IFMGR_NS}">
+    <interface-configuration xmlns:nc="{NC_NS}" nc:operation="delete">
+      <active>act</active>
+      <interface-name>{name}</interface-name>
+    </interface-configuration>
+  </interface-configurations>
+</config>
+"""
+
+
+def delete_static_route_xml(prefix, prefix_len):
+    return f"""
+<config>
+  <router-static xmlns="{STATIC_NS}">
+    <default-vrf>
+      <address-family>
+        <vrfipv4>
+          <vrf-unicast>
+            <vrf-prefixes>
+              <vrf-prefix xmlns:nc="{NC_NS}" nc:operation="delete">
+                <prefix>{prefix}</prefix>
+                <prefix-length>{prefix_len}</prefix-length>
+              </vrf-prefix>
+            </vrf-prefixes>
+          </vrf-unicast>
+        </vrfipv4>
+      </address-family>
+    </default-vrf>
+  </router-static>
+</config>
+"""
+
+
+def delete_bgp_redistribute_xml():
+    return f"""
+<config>
+  <bgp xmlns="{BGP_NS}">
+    <instance>
+      <instance-name>default</instance-name>
+      <instance-as>
+        <as>0</as>
+        <four-byte-as>
+          <as>65020</as>
+          <bgp-running/>
+          <default-vrf>
+            <global>
+              <global-afs>
+                <global-af>
+                  <af-name>ipv4-unicast</af-name>
+                  <enable/>
+                  <connected-routes xmlns:nc="{NC_NS}" nc:operation="delete"/>
+                  <static-routes xmlns:nc="{NC_NS}" nc:operation="delete"/>
+                </global-af>
+              </global-afs>
+            </global>
+          </default-vrf>
+        </four-byte-as>
+      </instance-as>
+    </instance>
+  </bgp>
+</config>
+"""
+
+
+def delete_isis_interface_xml(intf_name):
+    return f"""
+<config>
+  <isis xmlns="{ISIS_NS}">
+    <instances>
+      <instance>
+        <instance-name>1</instance-name>
+        <interfaces>
+          <interface xmlns:nc="{NC_NS}" nc:operation="delete">
+            <interface-name>{intf_name}</interface-name>
+          </interface>
+        </interfaces>
+      </instance>
+    </instances>
+  </isis>
+</config>
+"""
+
+
+def delete_route_policy_xml(name):
+    return f"""
+<config>
+  <routing-policy xmlns="{RPL_NS}">
+    <route-policies>
+      <route-policy xmlns:nc="{NC_NS}" nc:operation="delete">
+        <route-policy-name>{name}</route-policy-name>
+      </route-policy>
+    </route-policies>
+  </routing-policy>
+</config>
+"""
+
+
+# ──────────────────────────────────────────────────────────────────────
+# Configuration functions
+# ──────────────────────────────────────────────────────────────────────
+
+def nc_connect(mgmt_ip):
+    """Open NETCONF session."""
+    return manager.connect(
+        host=mgmt_ip,
+        port=830,
+        username='webui',
+        password='cisco',
+        hostkey_verify=False,
+        device_params={'name': 'iosxr'},
+        timeout=30,
+    )
+
+
+def configure_router(label, cfg):
+    """Apply full route-diversity config to a single router."""
+    mgmt_ip = cfg['mgmt']
+    print(f"\n{'─'*60}")
+    print(f"  Configuring {label} ({mgmt_ip})")
+    print(f"{'─'*60}")
+
+    lb_names = [lb[0] for lb in cfg['loopbacks']]
+    static_prefixes = [f"{s[0]}/{s[1]}" for s in cfg['statics']]
+    print(f"  Loopbacks: {', '.join(lb_names)}")
+    print(f"  Statics:   {', '.join(static_prefixes)}")
+
+    try:
+        with nc_connect(mgmt_ip) as m:
+            # Phase 1: Route-policy (must exist before BGP references it)
+            print(f"  → Creating route-policy {ROUTE_POLICY_NAME}...")
+            m.edit_config(target='candidate', config=route_policy_xml(ROUTE_POLICY_NAME, ROUTE_POLICY_BODY))
+
+            # Phase 2: Loopback interfaces
+            for name, addr, mask in cfg['loopbacks']:
+                print(f"  → Creating {name} ({addr})...")
+                m.edit_config(target='candidate', config=loopback_xml(name, addr, mask))
+
+            # Phase 3: Static routes
+            for prefix, plen, tag in cfg['statics']:
+                print(f"  → Static {prefix}/{plen} → Null0 tag={tag}...")
+                m.edit_config(target='candidate', config=static_route_xml(prefix, plen, tag))
+
+            # Phase 4: IS-IS passive on new loopbacks
+            for name, _, _ in cfg['loopbacks']:
+                print(f"  → IS-IS passive: {name}...")
+                m.edit_config(target='candidate', config=isis_passive_xml(name))
+
+            # Phase 5: BGP redistribution
+            print(f"  → BGP redistribute connected + static...")
+            m.edit_config(target='candidate', config=bgp_redistribute_xml())
+
+            # Phase 6: Commit
+            print(f"  → Committing...")
+            m.commit()
+            print(f"  ✓ {label} done.")
+            return True
+
+    except Exception as e:
+        print(f"  ✗ ERROR on {label}: {e}")
+        return False
+
+
+def rollback_router(label, cfg):
+    """Remove all route-diversity config from a single router."""
+    mgmt_ip = cfg['mgmt']
+    print(f"\n{'─'*60}")
+    print(f"  Rolling back {label} ({mgmt_ip})")
+    print(f"{'─'*60}")
+
+    try:
+        with nc_connect(mgmt_ip) as m:
+            # Remove BGP redistribution first (references the policy)
+            print(f"  → Removing BGP redistribute...")
+            try:
+                m.edit_config(target='candidate', config=delete_bgp_redistribute_xml())
+            except Exception as e:
+                print(f"    (skip — may not exist: {e})")
+
+            # Remove IS-IS interfaces
+            for name, _, _ in cfg['loopbacks']:
+                print(f"  → Removing IS-IS interface {name}...")
+                try:
+                    m.edit_config(target='candidate', config=delete_isis_interface_xml(name))
+                except Exception as e:
+                    print(f"    (skip: {e})")
+
+            # Remove static routes
+            for prefix, plen, _ in cfg['statics']:
+                print(f"  → Removing static {prefix}/{plen}...")
+                try:
+                    m.edit_config(target='candidate', config=delete_static_route_xml(prefix, plen))
+                except Exception as e:
+                    print(f"    (skip: {e})")
+
+            # Remove loopbacks
+            for name, _, _ in cfg['loopbacks']:
+                print(f"  → Removing {name}...")
+                try:
+                    m.edit_config(target='candidate', config=delete_loopback_xml(name))
+                except Exception as e:
+                    print(f"    (skip: {e})")
+
+            # Remove route-policy
+            print(f"  → Removing route-policy {ROUTE_POLICY_NAME}...")
+            try:
+                m.edit_config(target='candidate', config=delete_route_policy_xml(ROUTE_POLICY_NAME))
+            except Exception as e:
+                print(f"    (skip: {e})")
+
+            print(f"  → Committing rollback...")
+            m.commit()
+            print(f"  ✓ {label} rolled back.")
+            return True
+
+    except Exception as e:
+        print(f"  ✗ ERROR rolling back {label}: {e}")
+        return False
+
+
+def verify_router(label, cfg):
+    """Check if route-diversity config is present on a router."""
+    mgmt_ip = cfg['mgmt']
+
+    try:
+        with nc_connect(mgmt_ip) as m:
+            # Check loopbacks
+            filt_intf = f"""<filter>
+              <interface-configurations xmlns="{IFMGR_NS}"/>
+            </filter>"""
+            r_intf = str(m.get_config(source='running', filter=filt_intf))
+
+            found_lbs = []
+            for name, _, _ in cfg['loopbacks']:
+                if name in r_intf:
+                    found_lbs.append(name)
+
+            # Check route-policy
+            filt_rpl = f"""<filter>
+              <routing-policy xmlns="{RPL_NS}"/>
+            </filter>"""
+            r_rpl = str(m.get_config(source='running', filter=filt_rpl))
+            has_policy = ROUTE_POLICY_NAME in r_rpl
+
+            # Check BGP redistribute
+            filt_bgp = f"""<filter>
+              <bgp xmlns="{BGP_NS}">
+                <instance><instance-name>default</instance-name></instance>
+              </bgp>
+            </filter>"""
+            r_bgp = str(m.get_config(source='running', filter=filt_bgp))
+            has_redist_connected = 'connected-routes' in r_bgp and ROUTE_POLICY_NAME in r_bgp
+            has_redist_static = 'static-routes' in r_bgp and ROUTE_POLICY_NAME in r_bgp
+
+            total_lbs = len(cfg['loopbacks'])
+            lb_str = f"{len(found_lbs)}/{total_lbs}"
+            pol = '✓' if has_policy else '✗'
+            rc = '✓' if has_redist_connected else '✗'
+            rs = '✓' if has_redist_static else '✗'
+            ok = len(found_lbs) == total_lbs and has_policy and has_redist_connected and has_redist_static
+            status = 'OK' if ok else 'INCOMPLETE'
+
+            print(f"  {label:8s}  LBs={lb_str:5s}  Policy={pol}  Redist-C={rc}  Redist-S={rs}  [{status}]")
+
+    except Exception as e:
+        print(f"  {label:8s}  verify error: {e}")
+
+
+# ──────────────────────────────────────────────────────────────────────
+# Main
+# ──────────────────────────────────────────────────────────────────────
+
+def main():
+    parser = argparse.ArgumentParser(description='Route Diversity Configuration for RR Diff Analysis')
+    parser.add_argument('--verify-only', action='store_true', help='Only verify current state')
+    parser.add_argument('--rollback', action='store_true', help='Remove all added config')
+    args = parser.parse_args()
+
+    print("Route Diversity Configuration Script")
+    print("=" * 60)
+    print(f"Targets: {len(ROUTERS)} routers ({', '.join(ROUTERS.keys())})")
+    print()
+
+    if args.verify_only:
+        print("Verify-only mode")
+        print('-' * 60)
+        print(f"  {'Router':8s}  {'LBs':5s}  {'Policy':6s}  {'Redist-C':8s}  {'Redist-S':8s}  Status")
+        for label, cfg in ROUTERS.items():
+            verify_router(label, cfg)
+        return
+
+    if args.rollback:
+        print("ROLLBACK mode — removing all route-diversity config")
+        print('-' * 60)
+        results = []
+        for label, cfg in ROUTERS.items():
+            ok = rollback_router(label, cfg)
+            results.append((label, ok))
+
+        failed = [l for l, ok in results if not ok]
+        print()
+        if failed:
+            print(f"FAILED rollback: {', '.join(failed)}")
+            sys.exit(1)
+        else:
+            print("All routers rolled back successfully.")
+        return
+
+    # Apply mode
+    results = []
+    for label, cfg in ROUTERS.items():
+        ok = configure_router(label, cfg)
+        results.append((label, ok))
+
+    # Post-apply verification
+    print(f"\n{'='*60}")
+    print("Post-apply verification")
+    print('=' * 60)
+    print(f"  {'Router':8s}  {'LBs':5s}  {'Policy':6s}  {'Redist-C':8s}  {'Redist-S':8s}  Status")
+    for label, cfg in ROUTERS.items():
+        verify_router(label, cfg)
+
+    failed = [l for l, ok in results if not ok]
+    print()
+    if failed:
+        print(f"FAILED: {', '.join(failed)}")
+        sys.exit(1)
+    else:
+        total_lbs = sum(len(c['loopbacks']) for c in ROUTERS.values())
+        total_statics = sum(len(c['statics']) for c in ROUTERS.values())
+        print(f"All routers configured successfully.")
+        print(f"  {total_lbs} loopbacks + {total_statics} static routes created")
+        print()
+        print("Wait ~60s for BGP convergence and BMP collection, then verify:")
+        print()
+        print("  # Check new prefixes in OpenBMP")
+        print("  docker exec -i obmp-psql psql -U openbmp -d openbmp -c \\")
+        print("    \"SELECT prefix::text, COUNT(*) FROM ip_rib")
+        print("     WHERE (prefix::text LIKE '10.110.%' OR prefix::text LIKE '10.111.%')")
+        print("     AND iswithdrawn = false GROUP BY prefix ORDER BY prefix;\"")
+
+
+if __name__ == '__main__':
+    main()

exabgp/scenarios/__init__.py

@@ -441,6 +441,100 @@ _PATH_DIVERSITY_ROUTES = [
 # Registry
 # ---------------------------------------------------------------------------
 
+# ---------------------------------------------------------------------------
+# Full Internet Table Generator
+# Generates realistic-looking IPv4 prefixes across the routable address space
+# with varied AS paths, prefix lengths, origins, and communities.
+# Configurable count: 10K (quick test) to 900K+ (full table stress test).
+# ---------------------------------------------------------------------------
+
+# Well-known transit ASNs for realistic path construction
+_TRANSIT_ASNS = [174, 701, 1299, 2914, 3257, 3356, 6461, 6762, 7018, 3491, 5400, 1239]
+
+# Realistic origin ASNs (mix of large providers and small networks)
+_ORIGIN_POOL = [
+    13335, 15169, 16509, 8075, 20940, 32934, 714, 54113, 13414, 7922,
+    36459, 46489, 14618, 16276, 24940, 47541, 35916, 49981, 9808, 4134,
+    4837, 9121, 12322, 3320, 6830, 5511, 1273, 6939, 4766, 9318,
+    23693, 38001, 45102, 58453, 10026, 18881, 28573, 7738, 26599, 8151,
+    11888, 17676, 4713, 7545, 9299, 50304, 51167, 60068, 41095, 34984,
+]
+
+# IANA-allocated first octets for routable IPv4 (subset for realism)
+_ROUTABLE_FIRST_OCTETS = list(range(1, 56)) + list(range(57, 127)) + list(range(128, 224))
+
+
+def generate_full_internet(count=900000):
+    """Generate a realistic full IPv4 routing table.
+
+    Distributes prefixes across the IPv4 address space with realistic
+    prefix lengths (/8 through /24) and varied AS paths.
+
+    Args:
+        count: Number of prefixes to generate (default 900K).
+
+    Returns:
+        List of route dicts.
+    """
+    import random
+    rng = random.Random(42)  # deterministic for reproducibility
+
+    routes = []
+    generated = set()
+
+    # Prefix length distribution (approximates real DFZ):
+    # /24: ~55%, /23: ~8%, /22: ~7%, /21: ~5%, /20: ~5%,
+    # /19: ~4%, /18: ~3%, /17: ~2%, /16: ~5%, /15-/8: ~6%
+    prefix_len_weights = {
+        24: 55, 23: 8, 22: 7, 21: 5, 20: 5,
+        19: 4, 18: 3, 17: 2, 16: 5, 15: 2,
+        14: 1, 13: 1, 12: 1, 11: 0.5, 10: 0.3,
+        9: 0.1, 8: 0.1,
+    }
+    plen_choices = list(prefix_len_weights.keys())
+    plen_weights = list(prefix_len_weights.values())
+
+    # AS path length distribution: 1-hop: 5%, 2-hop: 30%, 3-hop: 40%, 4-hop: 20%, 5-hop: 5%
+    path_len_weights = [5, 30, 40, 20, 5]
+
+    while len(routes) < count:
+        # Pick a routable first octet weighted by allocation density
+        first = rng.choice(_ROUTABLE_FIRST_OCTETS)
+        plen = rng.choices(plen_choices, weights=plen_weights, k=1)[0]
+
+        # Generate random prefix within this /8
+        if plen <= 8:
+            prefix = f'{first}.0.0.0/{plen}'
+        elif plen <= 16:
+            second = rng.randint(0, 255) & (0xFF << (16 - plen))
+            prefix = f'{first}.{second}.0.0/{plen}'
+        elif plen <= 24:
+            second = rng.randint(0, 255)
+            third = rng.randint(0, 255) & (0xFF << (24 - plen))
+            prefix = f'{first}.{second}.{third}.0/{plen}'
+        else:
+            continue
+
+        if prefix in generated:
+            continue
+        generated.add(prefix)
+
+        # Build realistic AS path
+        path_len = rng.choices([1, 2, 3, 4, 5], weights=path_len_weights, k=1)[0]
+        origin = rng.choice(_ORIGIN_POOL) if rng.random() < 0.3 else (64512 + rng.randint(0, 65535 - 64512))
+        transits = rng.sample(_TRANSIT_ASNS, min(path_len - 1, len(_TRANSIT_ASNS)))
+        as_path = [65100] + transits[:path_len - 1] + [origin]
+
+        # Occasionally add communities (~20% of routes)
+        communities = []
+        if rng.random() < 0.2:
+            communities.append(f'65100:{rng.choice([100, 200, 300, 400, 500])}')
+
+        routes.append(_r(prefix, as_path, communities=communities or None))
+
+    return routes
+
+
 SCENARIOS = {
     'internet_sample': {
         'description': 'Partial internet table (~80 IPv4 + 14 IPv6 prefixes with realistic AS paths)',

obmp-grafana/dashboards/Learning/rr_locrib_diff.json

@@ -0,0 +1,418 @@
+{
+  "annotations": {"list": [{"builtIn": 1,"datasource": {"type": "datasource","uid": "grafana"},"enable": true,"hide": true,"iconColor": "rgba(0, 211, 255, 1)","name": "Annotations & Alerts","target": {"limit": 100,"matchAny": false,"tags": [],"type": "dashboard"},"type": "dashboard"}]},
+  "description": "Compare Route Reflector Loc-RIB tables via BMP Adj-RIB-In. Surfaces missing prefixes, attribute differences, and per-client consistency between two RRs.",
+  "editable": true,
+  "fiscalYearStartMonth": 0,
+  "graphTooltip": 1,
+  "id": null,
+  "links": [],
+  "panels": [
+    {
+      "datasource": {"type": "datasource","uid": "grafana"},
+      "gridPos": {"h": 5,"w": 24,"x": 0,"y": 0},
+      "id": 1,
+      "options": {
+        "content": "## Route Reflector Loc-RIB Diff\n\nCompares the RIB tables of two Route Reflectors via BMP. Select your two RRs using the **RR1** and **RR2** dropdowns above.\n\n**What this shows:**\n- **Summary stats** — prefix counts and diff totals\n- **Missing prefixes** — routes present on one RR but not the other (expected for each RR's own locally-originated routes)\n- **Attribute differences** — same prefix on both RRs but with different next-hop, AS path, or other attributes (indicates different best-path selection)\n- **Per-client consistency** — route counts each client sends to each RR (should be identical in a healthy RR cluster)",
+        "mode": "markdown"
+      },
+      "title": "RR Loc-RIB Diff — Overview",
+      "type": "text"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Total active (non-withdrawn) prefixes on RR1",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "blue","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "blue","value": null}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 0,"y": 5},
+      "id": 10,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT COUNT(DISTINCT r.prefix || '/' || r.prefix_len) AS \"prefixes\"\nFROM ip_rib r\nJOIN bgp_peers p ON p.hash_id = r.peer_hash_id\nJOIN routers rt ON rt.hash_id = p.router_hash_id\nWHERE rt.name = '$rr1' AND r.iswithdrawn = false",
+          "refId": "A"
+        }
+      ],
+      "title": "$rr1 Prefixes",
+      "type": "stat"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Total active (non-withdrawn) prefixes on RR2",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "green","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 4,"y": 5},
+      "id": 11,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT COUNT(DISTINCT r.prefix || '/' || r.prefix_len) AS \"prefixes\"\nFROM ip_rib r\nJOIN bgp_peers p ON p.hash_id = r.peer_hash_id\nJOIN routers rt ON rt.hash_id = p.router_hash_id\nWHERE rt.name = '$rr2' AND r.iswithdrawn = false",
+          "refId": "A"
+        }
+      ],
+      "title": "$rr2 Prefixes",
+      "type": "stat"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on RR1 but missing from RR2",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "orange","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "orange","value": null},{"color": "red","value": 1}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 8,"y": 5},
+      "id": 12,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT COUNT(*) AS \"only_rr1\"\nFROM (\n  SELECT DISTINCT r1.prefix, r1.prefix_len\n  FROM ip_rib r1\n  JOIN bgp_peers p1 ON p1.hash_id = r1.peer_hash_id\n  JOIN routers rt1 ON rt1.hash_id = p1.router_hash_id\n  WHERE rt1.name = '$rr1' AND r1.iswithdrawn = false\n  EXCEPT\n  SELECT DISTINCT r2.prefix, r2.prefix_len\n  FROM ip_rib r2\n  JOIN bgp_peers p2 ON p2.hash_id = r2.peer_hash_id\n  JOIN routers rt2 ON rt2.hash_id = p2.router_hash_id\n  WHERE rt2.name = '$rr2' AND r2.iswithdrawn = false\n) sub",
+          "refId": "A"
+        }
+      ],
+      "title": "Only on $rr1",
+      "type": "stat"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on RR2 but missing from RR1",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "orange","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "orange","value": null},{"color": "red","value": 1}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 12,"y": 5},
+      "id": 13,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT COUNT(*) AS \"only_rr2\"\nFROM (\n  SELECT DISTINCT r2.prefix, r2.prefix_len\n  FROM ip_rib r2\n  JOIN bgp_peers p2 ON p2.hash_id = r2.peer_hash_id\n  JOIN routers rt2 ON rt2.hash_id = p2.router_hash_id\n  WHERE rt2.name = '$rr2' AND r2.iswithdrawn = false\n  EXCEPT\n  SELECT DISTINCT r1.prefix, r1.prefix_len\n  FROM ip_rib r1\n  JOIN bgp_peers p1 ON p1.hash_id = r1.peer_hash_id\n  JOIN routers rt1 ON rt1.hash_id = p1.router_hash_id\n  WHERE rt1.name = '$rr1' AND r1.iswithdrawn = false\n) sub",
+          "refId": "A"
+        }
+      ],
+      "title": "Only on $rr2",
+      "type": "stat"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on both RRs but with different best-path attributes (next-hop, AS path, MED, or local-pref)",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "yellow","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null},{"color": "yellow","value": 1},{"color": "red","value": 10}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 16,"y": 5},
+      "id": 14,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "WITH c1 AS (\n  SELECT DISTINCT ON (r.prefix, r.prefix_len)\n    r.prefix, r.prefix_len, ba.next_hop, ba.as_path,\n    COALESCE(ba.local_pref, 0) AS lp, COALESCE(ba.med, 0) AS med\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  JOIN base_attrs ba ON ba.hash_id = r.base_attr_hash_id\n  WHERE rt.name = '$rr1' AND r.iswithdrawn = false\n  ORDER BY r.prefix, r.prefix_len, ba.local_pref DESC NULLS LAST\n),\nc2 AS (\n  SELECT DISTINCT ON (r.prefix, r.prefix_len)\n    r.prefix, r.prefix_len, ba.next_hop, ba.as_path,\n    COALESCE(ba.local_pref, 0) AS lp, COALESCE(ba.med, 0) AS med\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  JOIN base_attrs ba ON ba.hash_id = r.base_attr_hash_id\n  WHERE rt.name = '$rr2' AND r.iswithdrawn = false\n  ORDER BY r.prefix, r.prefix_len, ba.local_pref DESC NULLS LAST\n)\nSELECT COUNT(*) AS \"attr_diffs\"\nFROM c1 JOIN c2 ON c1.prefix = c2.prefix AND c1.prefix_len = c2.prefix_len\nWHERE c1.next_hop != c2.next_hop OR c1.as_path != c2.as_path\n   OR c1.lp != c2.lp OR c1.med != c2.med",
+          "refId": "A"
+        }
+      ],
+      "title": "Attribute Diffs",
+      "type": "stat"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Total RIB entries (including multiple paths per prefix) on each RR",
+      "fieldConfig": {
+        "defaults": {"color": {"fixedColor": "purple","mode": "fixed"},"thresholds": {"mode": "absolute","steps": [{"color": "purple","value": null}]}},
+        "overrides": []
+      },
+      "gridPos": {"h": 4,"w": 4,"x": 20,"y": 5},
+      "id": 15,
+      "options": {"colorMode": "value","graphMode": "none","justifyMode": "auto","orientation": "auto","reduceOptions": {"calcs": ["lastNotNull"],"fields": "","values": false},"textMode": "auto"},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT\n  SUM(CASE WHEN rt.name = '$rr1' THEN 1 ELSE 0 END) AS \"RR1 Total Paths\",\n  SUM(CASE WHEN rt.name = '$rr2' THEN 1 ELSE 0 END) AS \"RR2 Total Paths\"\nFROM ip_rib r\nJOIN bgp_peers p ON p.hash_id = r.peer_hash_id\nJOIN routers rt ON rt.hash_id = p.router_hash_id\nWHERE rt.name IN ('$rr1', '$rr2') AND r.iswithdrawn = false",
+          "refId": "A"
+        }
+      ],
+      "title": "Total RIB Paths",
+      "type": "stat"
+    },
+    {
+      "collapsed": false,
+      "gridPos": {"h": 1,"w": 24,"x": 0,"y": 9},
+      "id": 20,
+      "title": "Missing Prefixes",
+      "type": "row"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on RR1 but NOT on RR2. Expected: RR2's own locally-originated routes won't appear here (they're on RR2 only). Unexpected entries indicate a convergence or session issue.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "Learned From"},"properties": [{"id": "custom.width","value": 130}]},
+          {"matcher": {"id": "byName","options": "Origin AS"},"properties": [{"id": "custom.width","value": 90}]},
+          {"matcher": {"id": "byName","options": "Next Hop"},"properties": [{"id": "custom.width","value": 140}]}
+        ]
+      },
+      "gridPos": {"h": 10,"w": 12,"x": 0,"y": 10},
+      "id": 21,
+      "options": {"footer": {"fields": "","reducer": ["count"],"show": true},"showHeader": true,"sortBy": [{"desc": false,"displayName": "Prefix"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT DISTINCT ON (r1.prefix, r1.prefix_len)\n  r1.prefix::text AS \"Prefix\",\n  ba.origin_as AS \"Origin AS\",\n  ba.next_hop::text AS \"Next Hop\",\n  ba.as_path::text AS \"AS Path\",\n  p1.peer_addr::text AS \"Learned From\"\nFROM ip_rib r1\nJOIN bgp_peers p1 ON p1.hash_id = r1.peer_hash_id\nJOIN routers rt1 ON rt1.hash_id = p1.router_hash_id\nJOIN base_attrs ba ON ba.hash_id = r1.base_attr_hash_id\nWHERE rt1.name = '$rr1' AND r1.iswithdrawn = false\n  AND NOT EXISTS (\n    SELECT 1 FROM ip_rib r2\n    JOIN bgp_peers p2 ON p2.hash_id = r2.peer_hash_id\n    JOIN routers rt2 ON rt2.hash_id = p2.router_hash_id\n    WHERE rt2.name = '$rr2'\n      AND r2.prefix = r1.prefix AND r2.prefix_len = r1.prefix_len\n      AND r2.iswithdrawn = false\n  )\nORDER BY r1.prefix, r1.prefix_len, ba.local_pref DESC NULLS LAST",
+          "refId": "A"
+        }
+      ],
+      "title": "Prefixes Only on $rr1",
+      "type": "table"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on RR2 but NOT on RR1. Expected: RR1's own locally-originated routes won't appear here. Unexpected entries indicate a convergence or session issue.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "Learned From"},"properties": [{"id": "custom.width","value": 130}]},
+          {"matcher": {"id": "byName","options": "Origin AS"},"properties": [{"id": "custom.width","value": 90}]},
+          {"matcher": {"id": "byName","options": "Next Hop"},"properties": [{"id": "custom.width","value": 140}]}
+        ]
+      },
+      "gridPos": {"h": 10,"w": 12,"x": 12,"y": 10},
+      "id": 22,
+      "options": {"footer": {"fields": "","reducer": ["count"],"show": true},"showHeader": true,"sortBy": [{"desc": false,"displayName": "Prefix"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT DISTINCT ON (r2.prefix, r2.prefix_len)\n  r2.prefix::text AS \"Prefix\",\n  ba.origin_as AS \"Origin AS\",\n  ba.next_hop::text AS \"Next Hop\",\n  ba.as_path::text AS \"AS Path\",\n  p2.peer_addr::text AS \"Learned From\"\nFROM ip_rib r2\nJOIN bgp_peers p2 ON p2.hash_id = r2.peer_hash_id\nJOIN routers rt2 ON rt2.hash_id = p2.router_hash_id\nJOIN base_attrs ba ON ba.hash_id = r2.base_attr_hash_id\nWHERE rt2.name = '$rr2' AND r2.iswithdrawn = false\n  AND NOT EXISTS (\n    SELECT 1 FROM ip_rib r1\n    JOIN bgp_peers p1 ON p1.hash_id = r1.peer_hash_id\n    JOIN routers rt1 ON rt1.hash_id = p1.router_hash_id\n    WHERE rt1.name = '$rr1'\n      AND r1.prefix = r2.prefix AND r1.prefix_len = r2.prefix_len\n      AND r1.iswithdrawn = false\n  )\nORDER BY r2.prefix, r2.prefix_len, ba.local_pref DESC NULLS LAST",
+          "refId": "A"
+        }
+      ],
+      "title": "Prefixes Only on $rr2",
+      "type": "table"
+    },
+    {
+      "collapsed": false,
+      "gridPos": {"h": 1,"w": 24,"x": 0,"y": 20},
+      "id": 30,
+      "title": "Attribute Differences (Same Prefix, Different Best Path)",
+      "type": "row"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Prefixes present on both RRs but where the selected best path differs in next-hop, AS path, local-pref, or MED. This is normal for multi-homed link subnets where each RR selects a different best path based on router-id tiebreaker. Filter by address family using the AFI dropdown.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "RR1 Next Hop"},"properties": [{"id": "custom.displayMode","value": "color-text"},{"id": "color","value": {"mode": "fixed","fixedColor": "blue"}}]},
+          {"matcher": {"id": "byName","options": "RR2 Next Hop"},"properties": [{"id": "custom.displayMode","value": "color-text"},{"id": "color","value": {"mode": "fixed","fixedColor": "green"}}]},
+          {"matcher": {"id": "byName","options": "Diff Type"},"properties": [{"id": "custom.displayMode","value": "color-background"},{"id": "color","value": {"mode": "thresholds"}},{"id": "thresholds","value": {"mode": "absolute","steps": [{"color": "yellow","value": null}]}}]}
+        ]
+      },
+      "gridPos": {"h": 12,"w": 24,"x": 0,"y": 21},
+      "id": 31,
+      "options": {"footer": {"fields": "","reducer": ["count"],"show": true},"showHeader": true,"sortBy": [{"desc": false,"displayName": "Prefix"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "WITH c1 AS (\n  SELECT DISTINCT ON (r.prefix, r.prefix_len)\n    r.prefix, r.prefix_len, r.isipv4,\n    ba.next_hop, ba.as_path, ba.origin_as,\n    COALESCE(ba.local_pref, 0) AS lp, COALESCE(ba.med, 0) AS med,\n    ba.community_list, p.peer_addr AS learned_from\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  JOIN base_attrs ba ON ba.hash_id = r.base_attr_hash_id\n  WHERE rt.name = '$rr1' AND r.iswithdrawn = false\n  ORDER BY r.prefix, r.prefix_len, ba.local_pref DESC NULLS LAST\n),\nc2 AS (\n  SELECT DISTINCT ON (r.prefix, r.prefix_len)\n    r.prefix, r.prefix_len, r.isipv4,\n    ba.next_hop, ba.as_path, ba.origin_as,\n    COALESCE(ba.local_pref, 0) AS lp, COALESCE(ba.med, 0) AS med,\n    ba.community_list, p.peer_addr AS learned_from\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  JOIN base_attrs ba ON ba.hash_id = r.base_attr_hash_id\n  WHERE rt.name = '$rr2' AND r.iswithdrawn = false\n  ORDER BY r.prefix, r.prefix_len, ba.local_pref DESC NULLS LAST\n)\nSELECT\n  c1.prefix::text AS \"Prefix\",\n  CASE WHEN c1.isipv4 THEN 'IPv4' ELSE 'IPv6' END AS \"AFI\",\n  c1.next_hop::text AS \"RR1 Next Hop\",\n  c2.next_hop::text AS \"RR2 Next Hop\",\n  c1.as_path::text AS \"RR1 AS Path\",\n  c2.as_path::text AS \"RR2 AS Path\",\n  c1.lp AS \"RR1 LP\",\n  c2.lp AS \"RR2 LP\",\n  c1.med AS \"RR1 MED\",\n  c2.med AS \"RR2 MED\",\n  CASE\n    WHEN c1.next_hop != c2.next_hop AND c1.as_path != c2.as_path THEN 'NH+ASPath'\n    WHEN c1.next_hop != c2.next_hop THEN 'Next-Hop'\n    WHEN c1.as_path != c2.as_path THEN 'AS Path'\n    WHEN c1.lp != c2.lp THEN 'Local-Pref'\n    WHEN c1.med != c2.med THEN 'MED'\n    ELSE 'Other'\n  END AS \"Diff Type\"\nFROM c1 JOIN c2 ON c1.prefix = c2.prefix AND c1.prefix_len = c2.prefix_len\nWHERE (c1.next_hop != c2.next_hop OR c1.as_path != c2.as_path\n   OR c1.lp != c2.lp OR c1.med != c2.med)\n  AND ('$afi' = 'All' OR ('$afi' = 'IPv4' AND c1.isipv4 = true) OR ('$afi' = 'IPv6' AND c1.isipv4 = false))\nORDER BY c1.prefix",
+          "refId": "A"
+        }
+      ],
+      "title": "Attribute Differences — $rr1 vs $rr2",
+      "type": "table"
+    },
+    {
+      "collapsed": false,
+      "gridPos": {"h": 1,"w": 24,"x": 0,"y": 33},
+      "id": 40,
+      "title": "Per-Client Consistency",
+      "type": "row"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Route counts each RR client sends to each RR. In a healthy cluster, every client should send the same number of routes to both RRs. Mismatches indicate session issues, policy differences, or BMP reporting gaps.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "Delta"},"properties": [{"id": "custom.displayMode","value": "color-background"},{"id": "color","value": {"mode": "thresholds"}},{"id": "thresholds","value": {"mode": "absolute","steps": [{"color": "green","value": null},{"color": "yellow","value": 1},{"color": "red","value": 5}]}}]},
+          {"matcher": {"id": "byName","options": "Status"},"properties": [{"id": "custom.displayMode","value": "color-background"},{"id": "mappings","value": [{"options": {"Match": {"color": "green","index": 0,"text": "Match"},"MISMATCH": {"color": "red","index": 1,"text": "MISMATCH"}},"type": "value"}]}]}
+        ]
+      },
+      "gridPos": {"h": 10,"w": 12,"x": 0,"y": 34},
+      "id": 41,
+      "options": {"footer": {"fields": "","reducer": ["sum"],"show": false},"showHeader": true,"sortBy": [{"desc": true,"displayName": "Delta"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "WITH rr1_peers AS (\n  SELECT p.peer_addr, p.hash_id\n  FROM bgp_peers p\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  WHERE rt.name = '$rr1'\n),\nrr2_peers AS (\n  SELECT p.peer_addr, p.hash_id\n  FROM bgp_peers p\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  WHERE rt.name = '$rr2'\n),\nrr1_counts AS (\n  SELECT pp.peer_addr, COUNT(*) AS cnt\n  FROM rr1_peers pp\n  JOIN ip_rib r ON r.peer_hash_id = pp.hash_id AND r.iswithdrawn = false\n  GROUP BY pp.peer_addr\n),\nrr2_counts AS (\n  SELECT pp.peer_addr, COUNT(*) AS cnt\n  FROM rr2_peers pp\n  JOIN ip_rib r ON r.peer_hash_id = pp.hash_id AND r.iswithdrawn = false\n  GROUP BY pp.peer_addr\n)\nSELECT\n  COALESCE(c1.peer_addr, c2.peer_addr)::text AS \"Client\",\n  COALESCE(c1.cnt, 0) AS \"Routes to RR1\",\n  COALESCE(c2.cnt, 0) AS \"Routes to RR2\",\n  ABS(COALESCE(c1.cnt, 0) - COALESCE(c2.cnt, 0)) AS \"Delta\",\n  CASE WHEN COALESCE(c1.cnt, 0) = COALESCE(c2.cnt, 0) THEN 'Match' ELSE 'MISMATCH' END AS \"Status\"\nFROM rr1_counts c1\nFULL OUTER JOIN rr2_counts c2 ON c1.peer_addr = c2.peer_addr\nWHERE c1.peer_addr IS NOT NULL AND c2.peer_addr IS NOT NULL\nORDER BY ABS(COALESCE(c1.cnt, 0) - COALESCE(c2.cnt, 0)) DESC, COALESCE(c1.peer_addr, c2.peer_addr)",
+          "refId": "A"
+        }
+      ],
+      "title": "Per-Client Route Consistency",
+      "type": "table"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "For clients with route count mismatches, shows which specific prefixes differ. Select a client from the Client dropdown to drill down.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "Present On"},"properties": [{"id": "custom.displayMode","value": "color-text"},{"id": "color","value": {"mode": "thresholds"}},{"id": "thresholds","value": {"mode": "absolute","steps": [{"color": "blue","value": null}]}}]}
+        ]
+      },
+      "gridPos": {"h": 10,"w": 12,"x": 12,"y": 34},
+      "id": 42,
+      "options": {"footer": {"fields": "","reducer": ["count"],"show": true},"showHeader": true,"sortBy": [{"desc": false,"displayName": "Prefix"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "WITH rr1_client_routes AS (\n  SELECT r.prefix, r.prefix_len\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  WHERE rt.name = '$rr1' AND p.peer_addr::text = '$client'\n    AND r.iswithdrawn = false\n),\nrr2_client_routes AS (\n  SELECT r.prefix, r.prefix_len\n  FROM ip_rib r\n  JOIN bgp_peers p ON p.hash_id = r.peer_hash_id\n  JOIN routers rt ON rt.hash_id = p.router_hash_id\n  WHERE rt.name = '$rr2' AND p.peer_addr::text = '$client'\n    AND r.iswithdrawn = false\n)\nSELECT r1.prefix::text AS \"Prefix\", 'RR1 only' AS \"Present On\"\nFROM rr1_client_routes r1\nWHERE NOT EXISTS (\n  SELECT 1 FROM rr2_client_routes r2\n  WHERE r2.prefix = r1.prefix AND r2.prefix_len = r1.prefix_len\n)\nUNION ALL\nSELECT r2.prefix::text AS \"Prefix\", 'RR2 only' AS \"Present On\"\nFROM rr2_client_routes r2\nWHERE NOT EXISTS (\n  SELECT 1 FROM rr1_client_routes r1\n  WHERE r1.prefix = r2.prefix AND r1.prefix_len = r2.prefix_len\n)\nORDER BY \"Prefix\"",
+          "refId": "A"
+        }
+      ],
+      "title": "Client Prefix Diff — $client",
+      "type": "table"
+    },
+    {
+      "collapsed": false,
+      "gridPos": {"h": 1,"w": 24,"x": 0,"y": 44},
+      "id": 50,
+      "title": "Full RIB Comparison (All Paths)",
+      "type": "row"
+    },
+    {
+      "datasource": {"type": "postgres","uid": "obmp_postgres"},
+      "description": "Side-by-side view of all paths for a specific prefix on both RRs. Select a prefix from the Prefix dropdown to drill down.",
+      "fieldConfig": {
+        "defaults": {"color": {"mode": "thresholds"},"custom": {"align": "auto","displayMode": "auto","filterable": true},"thresholds": {"mode": "absolute","steps": [{"color": "green","value": null}]}},
+        "overrides": [
+          {"matcher": {"id": "byName","options": "Router"},"properties": [{"id": "custom.displayMode","value": "color-text"},{"id": "color","value": {"mode": "thresholds"}},{"id": "thresholds","value": {"mode": "absolute","steps": [{"color": "blue","value": null}]}}]}
+        ]
+      },
+      "gridPos": {"h": 10,"w": 24,"x": 0,"y": 45},
+      "id": 51,
+      "options": {"footer": {"fields": "","reducer": ["count"],"show": true},"showHeader": true,"sortBy": [{"desc": false,"displayName": "Router"}]},
+      "targets": [
+        {
+          "datasource": {"type": "postgres","uid": "obmp_postgres"},
+          "format": "table",
+          "rawSql": "SELECT\n  rt.name AS \"Router\",\n  p.peer_addr::text AS \"Learned From\",\n  ba.next_hop::text AS \"Next Hop\",\n  ba.as_path::text AS \"AS Path\",\n  ba.origin_as AS \"Origin AS\",\n  COALESCE(ba.local_pref, 0) AS \"Local Pref\",\n  COALESCE(ba.med, 0) AS \"MED\",\n  ba.community_list::text AS \"Communities\",\n  ba.cluster_list::text AS \"Cluster List\",\n  ba.originator_id::text AS \"Originator ID\",\n  r.labels AS \"Labels\",\n  r.timestamp AS \"Last Update\"\nFROM ip_rib r\nJOIN bgp_peers p ON p.hash_id = r.peer_hash_id\nJOIN routers rt ON rt.hash_id = p.router_hash_id\nJOIN base_attrs ba ON ba.hash_id = r.base_attr_hash_id\nWHERE rt.name IN ('$rr1', '$rr2')\n  AND r.iswithdrawn = false\n  AND r.prefix::text = '$prefix'\nORDER BY rt.name, p.peer_addr",
+          "refId": "A"
+        }
+      ],
+      "title": "All Paths for $prefix",
+      "type": "table"
+    }
+  ],
+  "refresh": "30s",
+  "schemaVersion": 38,
+  "tags": ["openbmp", "rr-diff", "bgp"],
+  "templating": {
+    "list": [
+      {
+        "current": {},
+        "datasource": {"type": "postgres","uid": "obmp_postgres"},
+        "definition": "SELECT name FROM routers WHERE state = 'up' ORDER BY name",
+        "hide": 0,
+        "includeAll": false,
+        "label": "RR1",
+        "multi": false,
+        "name": "rr1",
+        "options": [],
+        "query": "SELECT name FROM routers WHERE state = 'up' ORDER BY name",
+        "refresh": 1,
+        "regex": "",
+        "skipUrlSync": false,
+        "sort": 1,
+        "type": "query"
+      },
+      {
+        "current": {},
+        "datasource": {"type": "postgres","uid": "obmp_postgres"},
+        "definition": "SELECT name FROM routers WHERE state = 'up' ORDER BY name",
+        "hide": 0,
+        "includeAll": false,
+        "label": "RR2",
+        "multi": false,
+        "name": "rr2",
+        "options": [],
+        "query": "SELECT name FROM routers WHERE state = 'up' ORDER BY name",
+        "refresh": 1,
+        "regex": "",
+        "skipUrlSync": false,
+        "sort": 1,
+        "type": "query"
+      },
+      {
+        "current": {"selected": true,"text": "All","value": "All"},
+        "hide": 0,
+        "includeAll": false,
+        "label": "AFI",
+        "multi": false,
+        "name": "afi",
+        "options": [
+          {"selected": true,"text": "All","value": "All"},
+          {"selected": false,"text": "IPv4","value": "IPv4"},
+          {"selected": false,"text": "IPv6","value": "IPv6"}
+        ],
+        "query": "All,IPv4,IPv6",
+        "skipUrlSync": false,
+        "type": "custom"
+      },
+      {
+        "current": {},
+        "datasource": {"type": "postgres","uid": "obmp_postgres"},
+        "definition": "SELECT DISTINCT p.peer_addr::text FROM bgp_peers p JOIN routers rt ON rt.hash_id = p.router_hash_id WHERE rt.name IN ('$rr1', '$rr2') AND p.state = 'up' ORDER BY 1",
+        "hide": 0,
+        "includeAll": false,
+        "label": "Client",
+        "multi": false,
+        "name": "client",
+        "options": [],
+        "query": "SELECT DISTINCT p.peer_addr::text FROM bgp_peers p JOIN routers rt ON rt.hash_id = p.router_hash_id WHERE rt.name IN ('$rr1', '$rr2') AND p.state = 'up' ORDER BY 1",
+        "refresh": 1,
+        "regex": "",
+        "skipUrlSync": false,
+        "sort": 1,
+        "type": "query"
+      },
+      {
+        "current": {},
+        "datasource": {"type": "postgres","uid": "obmp_postgres"},
+        "definition": "SELECT DISTINCT r.prefix::text FROM ip_rib r JOIN bgp_peers p ON p.hash_id = r.peer_hash_id JOIN routers rt ON rt.hash_id = p.router_hash_id WHERE rt.name IN ('$rr1', '$rr2') AND r.iswithdrawn = false ORDER BY 1",
+        "hide": 0,
+        "includeAll": false,
+        "label": "Prefix",
+        "multi": false,
+        "name": "prefix",
+        "options": [],
+        "query": "SELECT DISTINCT r.prefix::text FROM ip_rib r JOIN bgp_peers p ON p.hash_id = r.peer_hash_id JOIN routers rt ON rt.hash_id = p.router_hash_id WHERE rt.name IN ('$rr1', '$rr2') AND r.iswithdrawn = false ORDER BY 1",
+        "refresh": 1,
+        "regex": "",
+        "skipUrlSync": false,
+        "sort": 1,
+        "type": "query"
+      }
+    ]
+  },
+  "time": {"from": "now-6h","to": "now"},
+  "timepicker": {},
+  "timezone": "",
+  "title": "RR Loc-RIB Diff",
+  "uid": "rr-locrib-diff",
+  "version": 1
+}

obmp-grafana/dashboards/Telemetry-3001/combined_bmp_telemetry.json

@@ -74,7 +74,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
           "refId": "A"
         }
       ],

obmp-grafana/dashboards/Telemetry-3001/interface_errors.json

@@ -59,7 +59,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-errors\" or r._field == \"out-errors\" or r._field == \"in-fcs-errors\")\n  |> derivative(unit: 1s, nonNegative: true)",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-errors\" or r._field == \"out-errors\" or r._field == \"in-fcs-errors\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)",
           "refId": "A"
         }
       ],
@@ -84,7 +84,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-discards\" or r._field == \"out-discards\")\n  |> derivative(unit: 1s, nonNegative: true)",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-discards\" or r._field == \"out-discards\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)",
           "refId": "A"
         }
       ],
@@ -114,7 +114,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-errors\" or r._field == \"out-errors\" or r._field == \"in-fcs-errors\" or r._field == \"in-discards\" or r._field == \"out-discards\")\n  |> last()\n  |> pivot(rowKey: [\"_time\"], columnKey: [\"_field\"], valueColumn: \"_value\")\n  |> keep(columns: [\"source\", \"name\", \"in-errors\", \"out-errors\", \"in-fcs-errors\", \"in-discards\", \"out-discards\"])\n  |> sort(columns: [\"in-errors\"], desc: true)",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-errors\" or r._field == \"out-errors\" or r._field == \"in-fcs-errors\" or r._field == \"in-discards\" or r._field == \"out-discards\")\n  |> toFloat()\n  |> last()\n  |> pivot(rowKey: [\"_time\"], columnKey: [\"_field\"], valueColumn: \"_value\")\n  |> keep(columns: [\"source\", \"name\", \"in-errors\", \"out-errors\", \"in-fcs-errors\", \"in-discards\", \"out-discards\"])\n  |> sort(columns: [\"in-errors\"], desc: true)",
           "refId": "A"
         }
       ],

obmp-grafana/dashboards/Telemetry-3001/interface_utilization.json

@@ -59,7 +59,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
           "refId": "A"
         }
       ],
@@ -84,7 +84,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-pkts\" or r._field == \"out-pkts\")\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-pkts\" or r._field == \"out-pkts\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)\n  |> map(fn: (r) => ({r with _value: if r._value < 0.0 then 0.0 else r._value}))",
           "refId": "A"
         }
       ],
@@ -109,7 +109,7 @@
       "targets": [
         {
           "datasource": {"type": "influxdb","uid": "obmp_influxdb"},
-          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> derivative(unit: 1s, nonNegative: true)\n  |> group(columns: [\"source\", \"name\", \"_field\"])\n  |> sum()\n  |> group(columns: [\"source\", \"name\"])\n  |> sum()\n  |> group()\n  |> sort(columns: [\"_value\"], desc: true)\n  |> limit(n: 20)",
+          "query": "from(bucket: \"telemetry\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"interface_counters\")\n  |> filter(fn: (r) => r.source =~ /${router:regex}/)\n  |> filter(fn: (r) => r.name =~ /${interface:regex}/)\n  |> filter(fn: (r) => r._field == \"in-octets\" or r._field == \"out-octets\")\n  |> toFloat()\n  |> derivative(unit: 1s, nonNegative: true)\n  |> group(columns: [\"source\", \"name\", \"_field\"])\n  |> sum()\n  |> group(columns: [\"source\", \"name\"])\n  |> sum()\n  |> group()\n  |> sort(columns: [\"_value\"], desc: true)\n  |> limit(n: 20)",
           "refId": "A"
         }
       ],

obmp-grafana/provisioning/datasources/influxdb-ds.yml

@@ -5,7 +5,7 @@ datasources:
     uid: obmp_influxdb
     type: influxdb
     access: proxy
-    url: http://obmp-influxdb:8086
+    url: http://10.40.40.202:8086
     jsonData:
       version: Flux
       organization: openbmp

portal/index.html

@@ -0,0 +1,106 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+  <meta charset="UTF-8">
+  <meta name="viewport" content="width=device-width, initial-scale=1.0">
+  <title>OpenBMP Lab Portal</title>
+  <style>
+    * { margin: 0; padding: 0; box-sizing: border-box; }
+    body {
+      font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
+      background: #111217;
+      color: #d8dee9;
+      min-height: 100vh;
+      display: flex;
+      flex-direction: column;
+      align-items: center;
+      padding: 2rem;
+    }
+    .header {
+      text-align: center;
+      margin-bottom: 2.5rem;
+    }
+    .header h1 {
+      font-size: 1.8rem;
+      color: #e2e8f0;
+      margin-bottom: 0.5rem;
+    }
+    .header p {
+      color: #7b8da0;
+      font-size: 0.95rem;
+    }
+    .grid {
+      display: grid;
+      grid-template-columns: repeat(auto-fit, minmax(260px, 1fr));
+      gap: 1.25rem;
+      max-width: 900px;
+      width: 100%;
+    }
+    .card {
+      background: #1a1d26;
+      border: 1px solid #2a2e3a;
+      border-radius: 10px;
+      padding: 1.5rem;
+      text-decoration: none;
+      color: inherit;
+      transition: border-color 0.2s, transform 0.15s;
+    }
+    .card:hover {
+      border-color: #3b82f6;
+      transform: translateY(-2px);
+    }
+    .card .icon {
+      font-size: 2rem;
+      margin-bottom: 0.75rem;
+      display: block;
+    }
+    .card h2 {
+      font-size: 1.1rem;
+      color: #e2e8f0;
+      margin-bottom: 0.4rem;
+    }
+    .card p {
+      font-size: 0.85rem;
+      color: #7b8da0;
+      line-height: 1.4;
+    }
+    .footer {
+      margin-top: 3rem;
+      color: #4a5568;
+      font-size: 0.8rem;
+      text-align: center;
+    }
+  </style>
+</head>
+<body>
+  <div class="header">
+    <h1>OpenBMP Lab</h1>
+    <p>BGP Monitoring Protocol &middot; Route Analysis &middot; Telemetry</p>
+  </div>
+
+  <div class="grid">
+    <a href="/grafana/" class="card">
+      <span class="icon">&#x1F4CA;</span>
+      <h2>Grafana Dashboards</h2>
+      <p>BGP analytics, RR Loc-RIB diff, IS-IS topology, telemetry, and 27+ dashboards.</p>
+    </a>
+
+    <a href="/exabgp/" class="card">
+      <span class="icon">&#x1F6E4;</span>
+      <h2>ExaBGP Route Injector</h2>
+      <p>Inject and withdraw BGP routes into the lab fabric via ExaBGP API.</p>
+    </a>
+
+    <a href="/traffic/" class="card">
+      <span class="icon">&#x1F680;</span>
+      <h2>Traffic Generator</h2>
+      <p>RFC 2544 throughput, latency, and loss testing across the network.</p>
+    </a>
+
+  </div>
+
+  <div class="footer">
+    OpenBMP Docker Stack &middot; 9 IOS-XR Routers &middot; CML Lab
+  </div>
+</body>
+</html>

traffic-gen-ui/nginx.conf

@@ -11,5 +11,11 @@ server {
 
     location / {
         try_files $uri $uri/ /index.html;
+        add_header Cache-Control "no-cache, no-store, must-revalidate";
+    }
+
+    location /assets/ {
+        expires 1y;
+        add_header Cache-Control "public, immutable";
     }
 }

traffic-gen-ui/src/App.vue

@@ -6,7 +6,7 @@
         <span class="logo-icon">&#9889;</span>
         <h1>Traffic Generator</h1>
       </div>
-      <StatusBar :health="health" :api-error="apiError" />
+      <StatusBar :health="health" :api-error="apiError" @modeChanged="fetchHealth(); fetchAll()" />
     </header>
 
     <!-- ERROR BANNER -->
@@ -19,7 +19,8 @@
     <div class="main-content">
       <!-- LEFT COLUMN: Flow Builder -->
       <aside class="left-col">
-        <FlowBuilder @created="fetchFlows" @updated="fetchFlows" />
+        <FlowBuilder :key="editFlow ? editFlow.id : 'new'" :editFlow="editFlow"
+          @created="onFlowSaved" @updated="onFlowSaved" @cancel="editFlow = null" />
       </aside>
 
       <!-- RIGHT COLUMN: Tabs -->
@@ -37,9 +38,16 @@
         </div>
 
         <div class="tab-content">
-          <FlowTable v-if="activeTab === 'flows'" :flows="flows" @refresh="fetchFlows" />
+          <div v-if="activeTab === 'flows'">
-          <TestBuilder v-else-if="activeTab === 'tests'" :flows="flows" @created="fetchTests" @refresh="fetchAll" />
+            <QuickPing />
-          <TestRunner v-else-if="activeTab === 'runner'" :tests="tests" @refresh="fetchTests" />
+            <FlowTable :flows="flows" @refresh="fetchFlows" @edit="startEdit" />
+          </div>
+          <div v-else-if="activeTab === 'tests'">
+            <TestBuilder @created="fetchTests" @refresh="fetchAll" />
+            <div style="margin-top: 20px;">
+              <TestRunner :tests="tests" @refresh="fetchTests" />
+            </div>
+          </div>
           <ResultsPanel v-else-if="activeTab === 'results'" :tests="tests" />
           <StatsMonitor v-else-if="activeTab === 'monitor'" :flows="flows" />
         </div>
@@ -50,19 +58,20 @@
     <footer class="app-footer">
       <span>Refreshing every 5s (health) / 3s (flows)</span>
       <span class="footer-sep">|</span>
-      <a href="http://localhost:3000" target="_blank" class="footer-link">Grafana: :3000</a>
+      <a :href="baseUrl + ':3000'" target="_blank" class="footer-link">Grafana: :3000</a>
       <span class="footer-sep">|</span>
-      <a href="http://localhost:5001" target="_blank" class="footer-link">Route Injector: :5001</a>
+      <a :href="baseUrl + ':5001'" target="_blank" class="footer-link">Route Injector: :5001</a>
     </footer>
   </div>
 </template>
 
 <script setup>
-import { ref, onMounted, onUnmounted } from 'vue'
+import { ref, computed, onMounted, onUnmounted } from 'vue'
 import { api } from './api.js'
 import StatusBar from './components/StatusBar.vue'
 import FlowBuilder from './components/FlowBuilder.vue'
 import FlowTable from './components/FlowTable.vue'
+import QuickPing from './components/QuickPing.vue'
 import TestBuilder from './components/TestBuilder.vue'
 import TestRunner from './components/TestRunner.vue'
 import ResultsPanel from './components/ResultsPanel.vue'
@@ -73,11 +82,15 @@ const flows = ref([])
 const tests = ref([])
 const apiError = ref(null)
 const activeTab = ref('flows')
+const editFlow = ref(null)
+const baseUrl = computed(() => `${window.location.protocol}//${window.location.hostname}`)
+
+function startEdit(flow) { editFlow.value = { ...flow } }
+function onFlowSaved() { editFlow.value = null; fetchFlows() }
 
 const tabs = [
   { id: 'flows', label: 'Flows' },
   { id: 'tests', label: 'Tests' },
-  { id: 'runner', label: 'Runner' },
   { id: 'results', label: 'Results' },
   { id: 'monitor', label: 'Monitor' },
 ]

traffic-gen-ui/src/api.js

@@ -1,4 +1,4 @@
-const BASE = '/api'
+const BASE = '/traffic/api'
 
 async function req(method, path, body) {
   const opts = { method, headers: { 'Content-Type': 'application/json' } }
@@ -12,6 +12,7 @@ export const api = {
   health:       () => req('GET', '/healthz'),
   interfaces:   () => req('GET', '/interfaces'),
   mode:         () => req('GET', '/mode'),
+  setMode:      (mode) => req('POST', '/mode', { mode }),
 
   // Flows
   flows:        () => req('GET', '/flows'),
@@ -38,6 +39,9 @@ export const api = {
   // Stats
   statsHistory: () => req('GET', '/stats/history'),
 
+  // Ping
+  ping:           (target, count) => req('POST', '/ping', { target, count: count || 5 }),
+
   // Responder
   responderStats: () => req('GET', '/responder/stats'),
   responderReset: () => req('POST', '/responder/reset'),

traffic-gen-ui/src/components/FlowBuilder.vue

@@ -42,14 +42,23 @@
           <input v-model.number="form.frame_size" type="number" min="64" max="9000" />
         </div>
         <div class="form-row">
-          <label>Rate (pps)</label>
+          <label>Rate</label>
-          <input v-model.number="form.rate_pps" type="number" min="1" max="100000" />
+          <input v-model.number="form.rate_val" type="number" min="1" step="any" />
+          <select v-model="form.rate_unit" class="rate-unit-standalone">
+            <option value="pps">pps</option>
+            <option value="kbps">Kbps</option>
+            <option value="mbps">Mbps</option>
+          </select>
         </div>
       </div>
       <div class="form-row-pair">
         <div class="form-row">
           <label>Duration (sec)</label>
-          <input v-model.number="form.duration" type="number" min="0" />
+          <input v-model.number="form.duration" type="number" min="0" :disabled="form.continuous" />
+          <label class="checkbox-inline">
+            <input type="checkbox" v-model="form.continuous" @change="onContinuousChange" />
+            Continuous
+          </label>
         </div>
         <div class="form-row">
           <label>DSCP</label>
@@ -71,32 +80,53 @@
 </template>
 
 <script setup>
-import { reactive, watch } from 'vue'
+import { reactive, computed } from 'vue'
 import { api } from '../api.js'
 
 const props = defineProps({ editFlow: Object })
 const emit = defineEmits(['created', 'updated', 'cancel'])
 
-const editing = !!props.editFlow
+const editing = computed(() => !!props.editFlow)
 
 const defaults = {
   name: '', dst_ip: '', src_ip: '', dst_mac: '',
   protocol: 'udp', src_port: 50000, dst_port: 5001,
-  frame_size: 512, rate_pps: 1000, duration: 30,
+  frame_size: 512, rate_val: 1000, rate_unit: 'pps', duration: 30,
-  dscp: 0, responder_url: '',
+  dscp: 0, responder_url: '', continuous: false,
 }
 
-const form = reactive({ ...defaults, ...(props.editFlow || {}) })
+function ppsToDisplay(pps, frameSize) {
+  // Convert stored PPS to a friendlier unit if it was originally set that way
+  return { rate_val: pps, rate_unit: 'pps' }
+}
+
+const initData = props.editFlow
+  ? { ...props.editFlow, continuous: props.editFlow.duration === 0, ...ppsToDisplay(props.editFlow.rate_pps, props.editFlow.frame_size) }
+  : {}
+const form = reactive({ ...defaults, ...initData })
+
+function onContinuousChange() { if (form.continuous) form.duration = 0 }
+
+function computePps(val, unit, frameSize) {
+  if (unit === 'kbps') return Math.max(1, Math.round((val * 1000) / (frameSize * 8)))
+  if (unit === 'mbps') return Math.max(1, Math.round((val * 1_000_000) / (frameSize * 8)))
+  return Math.round(val)
+}
 
 async function submit() {
   try {
     const payload = { ...form }
+    payload.rate_pps = computePps(form.rate_val, form.rate_unit, form.frame_size)
+    delete payload.rate_val
+    delete payload.rate_unit
+    delete payload.continuous
+    if (form.continuous) payload.duration = 0
     if (!payload.src_ip) delete payload.src_ip
     if (!payload.dst_mac) delete payload.dst_mac
     if (!payload.responder_url) delete payload.responder_url
     if (!payload.name) payload.name = `${payload.protocol.toUpperCase()} -> ${payload.dst_ip}`
 
-    if (editing) {
+    if (editing.value) {
       await api.updateFlow(props.editFlow.id, payload)
       emit('updated')
     } else {
@@ -123,4 +153,7 @@ h3 { font-size: 15px; margin-bottom: 12px; color: var(--accent); }
 .btn-accent:hover { opacity: 0.9; }
 .btn-accent:disabled { opacity: 0.4; }
 .btn-muted { background: var(--border); color: var(--text); }
+.rate-unit-standalone { width: 100%; margin-top: 4px; }
+.checkbox-inline { display: inline-flex !important; align-items: center; gap: 4px; margin-top: 4px; font-size: 12px; cursor: pointer; }
+.checkbox-inline input { width: auto; }
 </style>

traffic-gen-ui/src/components/FlowTable.vue

@@ -10,7 +10,9 @@
           <th>Size</th>
           <th>Rate</th>
           <th>State</th>
+          <th>TX Pkts</th>
           <th>TX pps</th>
+          <th>RX Pkts</th>
           <th>Actions</th>
         </tr>
       </thead>
@@ -20,14 +22,17 @@
           <td class="mono">{{ f.dst_ip }}</td>
           <td>{{ f.protocol.toUpperCase() }}</td>
           <td>{{ f.frame_size }}B</td>
-          <td>{{ f.rate_pps }} pps</td>
+          <td>{{ formatRate(f) }}</td>
           <td>
             <span class="state-badge" :class="'state-' + f.state">{{ f.state }}</span>
           </td>
-          <td class="mono">{{ stats[f.id]?.tx_pps || 0 }}</td>
+          <td class="mono">{{ formatNum(f.stats?.tx_packets || 0) }}</td>
+          <td class="mono">{{ pps[f.id] || 0 }}</td>
+          <td class="mono">{{ formatNum(f.stats?.rx_packets || 0) }}</td>
           <td class="actions">
             <button v-if="f.state !== 'running'" class="btn-sm btn-go" @click="start(f.id)">Start</button>
             <button v-else class="btn-sm btn-stop" @click="stop(f.id)">Stop</button>
+            <button class="btn-sm btn-edit" @click="emit('edit', f)" :disabled="f.state === 'running'">Edit</button>
             <button class="btn-sm btn-del" @click="del(f.id)" :disabled="f.state === 'running'">Del</button>
           </td>
         </tr>
@@ -41,23 +46,40 @@ import { ref, onMounted, onUnmounted } from 'vue'
 import { api } from '../api.js'
 
 const props = defineProps({ flows: Array })
-const emit = defineEmits(['refresh'])
+const emit = defineEmits(['refresh', 'edit'])
-const stats = ref({})
+const pps = ref({})
+const prevTx = ref({})
 
 let statsTimer = null
 
-async function fetchStats() {
+function computePps() {
   for (const f of (props.flows || [])) {
-    if (f.state === 'running') {
+    const txNow = f.stats?.tx_packets || 0
-      try {
+    const prev = prevTx.value[f.id] || 0
-        const s = await api.flowStats(f.id)
+    if (f.state === 'running' && prev > 0) {
-        stats.value[f.id] = s
+      pps.value[f.id] = Math.max(0, txNow - prev)
-      } catch (_) {}
+    } else if (f.state !== 'running') {
+      pps.value[f.id] = 0
     }
+    prevTx.value[f.id] = txNow
   }
 }
 
-onMounted(() => { statsTimer = setInterval(fetchStats, 1000) })
+function formatRate(f) {
+  const pps = f.rate_pps || 0
+  const mbps = (pps * (f.frame_size || 64) * 8) / 1_000_000
+  if (mbps >= 1) return mbps.toFixed(1) + ' Mbps'
+  if (mbps >= 0.001) return (mbps * 1000).toFixed(0) + ' Kbps'
+  return pps + ' pps'
+}
+
+function formatNum(n) {
+  if (n >= 1000000) return (n / 1000000).toFixed(1) + 'M'
+  if (n >= 1000) return (n / 1000).toFixed(1) + 'K'
+  return n
+}
+
+onMounted(() => { statsTimer = setInterval(computePps, 1000) })
 onUnmounted(() => { clearInterval(statsTimer) })
 
 async function start(id) {
@@ -88,5 +110,7 @@ tr.running { background: rgba(79,156,249,0.05); }
 .btn-go { background: var(--success); color: #fff; }
 .btn-stop { background: var(--warning); color: #000; }
 .btn-del { background: rgba(252,129,129,0.15); color: var(--danger); }
+.btn-edit { background: rgba(79,156,249,0.15); color: var(--accent); }
+.btn-edit:disabled { opacity: 0.3; }
 .btn-del:disabled { opacity: 0.3; }
 </style>

traffic-gen-ui/src/components/QuickPing.vue

@@ -0,0 +1,76 @@
+<template>
+  <div class="quick-ping">
+    <div class="ping-row">
+      <input
+        v-model="target"
+        placeholder="IP address to ping..."
+        @keyup.enter="runPing"
+        :disabled="pinging"
+      />
+      <button class="btn-ping" @click="runPing" :disabled="!target || pinging">
+        {{ pinging ? 'Pinging...' : 'Ping' }}
+      </button>
+    </div>
+    <div v-if="result" class="ping-result" :class="result.reachable ? 'reachable' : 'unreachable'">
+      <div class="ping-summary">
+        <span class="ping-target">{{ result.target }}</span>
+        <span v-if="result.reachable" class="ping-status ok">Reachable</span>
+        <span v-else class="ping-status fail">Unreachable</span>
+      </div>
+      <div v-if="result.reachable && result.stats" class="ping-stats">
+        <span>{{ result.received }}/{{ result.sent }} replies</span>
+        <span>Min: {{ result.stats.min_ms }}ms</span>
+        <span>Avg: {{ result.stats.avg_ms }}ms</span>
+        <span>Max: {{ result.stats.max_ms }}ms</span>
+        <span>Loss: {{ result.loss_pct }}%</span>
+      </div>
+      <div v-if="result.error" class="ping-error">{{ result.error }}</div>
+    </div>
+  </div>
+</template>
+
+<script setup>
+import { ref } from 'vue'
+import { api } from '../api.js'
+
+const target = ref('')
+const pinging = ref(false)
+const result = ref(null)
+
+async function runPing() {
+  if (!target.value || pinging.value) return
+  pinging.value = true
+  result.value = null
+  try {
+    result.value = await api.ping(target.value, 5)
+  } catch (e) {
+    result.value = { target: target.value, reachable: false, error: e.message }
+  } finally {
+    pinging.value = false
+  }
+}
+</script>
+
+<style scoped>
+.quick-ping { margin-bottom: 16px; }
+.ping-row { display: flex; gap: 6px; }
+.ping-row input { flex: 1; }
+.btn-ping {
+  padding: 6px 16px; font-weight: 600; font-size: 13px;
+  background: var(--accent); color: #fff; white-space: nowrap;
+}
+.btn-ping:disabled { opacity: 0.4; }
+.ping-result {
+  margin-top: 8px; padding: 8px 12px;
+  border-radius: var(--radius); font-size: 13px;
+}
+.ping-result.reachable { background: rgba(72,187,120,0.1); border: 1px solid rgba(72,187,120,0.3); }
+.ping-result.unreachable { background: rgba(252,129,129,0.1); border: 1px solid rgba(252,129,129,0.3); }
+.ping-summary { display: flex; align-items: center; gap: 10px; }
+.ping-target { font-weight: 600; font-family: monospace; }
+.ping-status { font-size: 11px; padding: 2px 8px; border-radius: 10px; font-weight: 600; }
+.ping-status.ok { background: rgba(72,187,120,0.2); color: var(--success); }
+.ping-status.fail { background: rgba(252,129,129,0.2); color: var(--danger); }
+.ping-stats { display: flex; gap: 12px; margin-top: 6px; font-size: 12px; color: var(--muted); font-family: monospace; }
+.ping-error { margin-top: 4px; color: var(--danger); font-size: 12px; }
+</style>

traffic-gen-ui/src/components/ResultsPanel.vue

@@ -13,8 +13,30 @@
         </div>
       </div>
 
-      <div v-if="t.results" class="result-table">
+      <div v-if="t.error" class="error-msg">Error: {{ t.error }}</div>
+      <div v-if="t.results && Object.keys(t.results).length" class="result-table">
+        <!-- Frame Loss: array of rate steps per frame size -->
+        <template v-if="t.type === 'frame_loss'">
+          <div v-for="(rates, size) in t.results" :key="size" class="fl-section">
+            <div class="fl-title">Frame Size: {{ size }} B</div>
             <table>
+              <thead>
+                <tr><th>Rate %</th><th>Rate (pps)</th><th>TX Packets</th><th>RX Packets</th><th>Loss %</th></tr>
+              </thead>
+              <tbody>
+                <tr v-for="r in rates" :key="r.rate_pct">
+                  <td class="mono">{{ r.rate_pct }}%</td>
+                  <td class="mono">{{ r.rate_pps }}</td>
+                  <td class="mono">{{ r.tx_packets }}</td>
+                  <td class="mono">{{ r.rx_packets }}</td>
+                  <td class="mono">{{ r.loss_pct }}%</td>
+                </tr>
+              </tbody>
+            </table>
+          </div>
+        </template>
+        <!-- Other test types -->
+        <table v-else>
           <thead>
             <tr>
               <th>Frame Size (B)</th>
@@ -32,7 +54,7 @@
         </table>
       </div>
 
-      <div v-if="t.results" class="result-chart">
+      <div v-if="t.results && t.type !== 'frame_loss'" class="result-chart">
         <div class="bar-chart">
           <div v-for="(val, size) in t.results" :key="size" class="bar-item">
             <div class="bar-fill" :style="{ height: barHeight(t, val) + '%' }"></div>
@@ -50,7 +72,7 @@ import { computed } from 'vue'
 const props = defineProps({ tests: Array })
 
 const completedTests = computed(() =>
-  (props.tests || []).filter(t => t.state === 'complete' && t.results)
+  (props.tests || []).filter(t => (t.state === 'complete' || t.state === 'error') && (t.results || t.error))
 )
 
 function resultColumns(t) {
@@ -63,22 +85,28 @@ function resultColumns(t) {
 
 function formatVal(val, col) {
   if (typeof val === 'object') {
-    if (col.includes('Rate')) return val.max_rate_pps ?? '-'
+    if (col.includes('Rate')) return val.max_throughput_pps ?? val.max_rate_pps ?? '-'
-    if (col.includes('Throughput')) return val.throughput_mbps ?? '-'
+    if (col.includes('Throughput')) {
-    if (col.includes('Min')) return val.min_ms ?? '-'
+      const pps = val.max_throughput_pps ?? val.max_rate_pps ?? 0
-    if (col.includes('Avg')) return val.avg_ms ?? '-'
+      const fs = val.frame_size ?? 64
-    if (col.includes('Max') && col.includes('ms')) return val.max_ms ?? '-'
+      return pps ? ((pps * fs * 8) / 1_000_000).toFixed(2) : '-'
-    if (col.includes('Jitter')) return val.jitter_ms ?? '-'
+    }
+    if (col.includes('Min')) return val.min_ms != null ? val.min_ms.toFixed(2) : '-'
+    if (col.includes('Avg')) return val.avg_ms != null ? val.avg_ms.toFixed(2) : '-'
+    if (col.includes('Max') && col.includes('ms')) return val.max_ms != null ? val.max_ms.toFixed(2) : '-'
+    if (col.includes('Jitter')) return val.jitter_ms != null ? val.jitter_ms.toFixed(2) : '-'
     if (col.includes('Loss')) return val.loss_pct ?? '-'
-    if (col.includes('Burst')) return val.max_burst ?? '-'
+    if (col.includes('Burst')) return val.max_burst_frames ?? val.max_burst ?? '-'
     return JSON.stringify(val)
   }
   return val
 }
 
 function barHeight(t, val) {
-  const v = typeof val === 'object' ? (val.max_rate_pps || val.avg_ms || val.loss_pct || val.max_burst || 0) : val
+  const v = typeof val === 'object' ? (val.max_throughput_pps || val.max_rate_pps || val.avg_ms || val.loss_pct || val.max_burst_frames || 0) : val
-  return Math.min(100, Math.max(5, v / 100))
+  const allVals = Object.values(t.results).map(r => typeof r === 'object' ? (r.max_throughput_pps || r.max_rate_pps || r.avg_ms || r.loss_pct || r.max_burst_frames || 0) : r)
+  const maxVal = Math.max(...allVals, 1)
+  return Math.min(100, Math.max(5, (v / maxVal) * 100))
 }
 
 function exportJSON(t) {
@@ -123,4 +151,7 @@ td { font-size: 13px; padding: 4px 8px; }
 .bar-item { flex: 1; display: flex; flex-direction: column; align-items: center; height: 100%; }
 .bar-fill { width: 100%; background: var(--accent); border-radius: 3px 3px 0 0; min-height: 4px; transition: height 0.3s; margin-top: auto; }
 .bar-label { font-size: 10px; color: var(--muted); margin-top: 4px; }
+.fl-section { margin-bottom: 12px; }
+.fl-title { font-size: 12px; font-weight: 600; color: var(--accent); margin-bottom: 4px; }
+.error-msg { color: var(--danger); font-size: 13px; padding: 8px 0; }
 </style>

traffic-gen-ui/src/components/StatsMonitor.vue

@@ -78,12 +78,79 @@ let timer = null
 async function fetchStats() {
   try {
     if (selectedFlow.value) {
+      // Single flow: /flows/<id>/stats returns {flow_id, counters, rates}
+      // rates contains: tx_pps, rx_pps, tx_mbps, rx_mbps, loss_pct, tx_packets, tx_bytes, etc.
       const s = await api.flowStats(selectedFlow.value)
-      current.value = s
+      const rates = s.rates || {}
+      const counters = s.counters || {}
+      current.value = {
+        tx_pps: Math.round(rates.tx_pps || 0),
+        rx_pps: Math.round(rates.rx_pps || 0),
+        tx_mbps: rates.tx_mbps || 0,
+        rx_mbps: rates.rx_mbps || 0,
+        loss_pct: rates.loss_pct || 0,
+        avg_latency_ms: rates.latency ? rates.latency.avg_ms : null,
+        tx_packets: counters.tx_packets || 0,
+        tx_bytes: counters.tx_bytes || 0,
+        rx_packets: counters.rx_packets || 0,
+        rx_bytes: counters.rx_bytes || 0,
+      }
+      // Append to history for sparkline
+      history.value.push({ tx_pps: current.value.tx_pps, rx_pps: current.value.rx_pps })
+      if (history.value.length > 60) history.value = history.value.slice(-60)
     } else {
+      // All flows: /stats/history returns {history: {flow_id: [samples]}}
       const h = await api.statsHistory()
-      if (h.current) current.value = h.current
+      const allHistory = h.history || {}
-      if (h.history) history.value = h.history.slice(-60)
+      // Aggregate latest sample across all flows
+      let txPps = 0, rxPps = 0, txMbps = 0, rxMbps = 0
+      let txPkts = 0, txBytes = 0, rxPkts = 0, rxBytes = 0
+      let lossPcts = [], latencies = []
+
+      for (const [, samples] of Object.entries(allHistory)) {
+        if (!samples.length) continue
+        const latest = samples[samples.length - 1]
+        txPps += latest.tx_pps || 0
+        rxPps += latest.rx_pps || 0
+        txMbps += latest.tx_mbps || 0
+        rxMbps += latest.rx_mbps || 0
+        txPkts += latest.tx_packets || 0
+        txBytes += latest.tx_bytes || 0
+        rxPkts += latest.rx_packets || 0
+        rxBytes += latest.rx_bytes || 0
+        if (latest.loss_pct > 0) lossPcts.push(latest.loss_pct)
+        if (latest.latency && latest.latency.avg_ms) latencies.push(latest.latency.avg_ms)
+      }
+
+      current.value = {
+        tx_pps: Math.round(txPps),
+        rx_pps: Math.round(rxPps),
+        tx_mbps: txMbps,
+        rx_mbps: rxMbps,
+        loss_pct: txPkts > 0 ? Math.max(0, ((txPkts - rxPkts) / txPkts) * 100) : 0,
+        avg_latency_ms: latencies.length ? latencies.reduce((a, b) => a + b, 0) / latencies.length : null,
+        tx_packets: txPkts,
+        tx_bytes: txBytes,
+        rx_packets: rxPkts,
+        rx_bytes: rxBytes,
+      }
+
+      // Build aggregated sparkline from history samples
+      // Find max sample count across all flows
+      const flowIds = Object.keys(allHistory)
+      if (flowIds.length) {
+        const maxLen = Math.max(...flowIds.map(id => allHistory[id].length))
+        const sparkData = []
+        for (let i = Math.max(0, maxLen - 60); i < maxLen; i++) {
+          let sTx = 0, sRx = 0
+          for (const fid of flowIds) {
+            const s = allHistory[fid][i]
+            if (s) { sTx += s.tx_pps || 0; sRx += s.rx_pps || 0 }
+          }
+          sparkData.push({ tx_pps: Math.round(sTx), rx_pps: Math.round(sRx) })
+        }
+        history.value = sparkData
+      }
     }
   } catch (_) {}
 }

traffic-gen-ui/src/components/StatusBar.vue

@@ -2,25 +2,45 @@
   <div class="status-bar">
     <div class="status-badges">
       <span class="badge" :class="connected ? 'badge-ok' : 'badge-err'">
-        {{ connected ? 'Connected' : 'Offline' }}
+        {{ connected ? 'API Connected' : 'API Offline' }}
+      </span>
+      <span v-if="health" class="badge badge-mode" :class="'mode-' + (health.mode || 'sender')" @click="toggleMode">
+        {{ (health.mode || 'sender').toUpperCase() }}
       </span>
       <span v-if="health" class="badge badge-info">
-        Mode: {{ health.mode || 'sender' }}
+        Active Flows: {{ health.active_flows || 0 }}
       </span>
       <span v-if="health" class="badge badge-info">
-        Flows: {{ health.active_flows || 0 }}
+        Active Tests: {{ health.active_tests || 0 }}
-      </span>
-      <span v-if="health" class="badge badge-info">
-        Tests: {{ health.active_tests || 0 }}
       </span>
     </div>
   </div>
 </template>
 
 <script setup>
-import { computed } from 'vue'
+import { computed, ref } from 'vue'
+import { api } from '../api.js'
+
 const props = defineProps({ health: Object, apiError: String })
+const emit = defineEmits(['modeChanged'])
 const connected = computed(() => !props.apiError && props.health)
+const switching = ref(false)
+
+async function toggleMode() {
+  if (switching.value || !props.health) return
+  const current = props.health.mode || 'sender'
+  const next = current === 'sender' ? 'responder' : 'sender'
+  if (!confirm(`Switch to ${next.toUpperCase()} mode? This will stop all active flows/tests.`)) return
+  switching.value = true
+  try {
+    await api.setMode(next)
+    emit('modeChanged')
+  } catch (e) {
+    alert('Failed to switch mode: ' + e.message)
+  } finally {
+    switching.value = false
+  }
+}
 </script>
 
 <style scoped>
@@ -33,5 +53,9 @@ const connected = computed(() => !props.apiError && props.health)
 .badge-ok { background: rgba(72,187,120,0.15); color: var(--success); }
 .badge-err { background: rgba(252,129,129,0.15); color: var(--danger); animation: pulse 1.5s infinite; }
 .badge-info { background: rgba(79,156,249,0.12); color: var(--accent); }
+.badge-mode { cursor: pointer; transition: background 0.2s; }
+.badge-mode:hover { opacity: 0.8; }
+.mode-sender { background: rgba(72,187,120,0.2); color: var(--success); }
+.mode-responder { background: rgba(246,173,85,0.2); color: var(--warning); }
 @keyframes pulse { 0%,100% { opacity: 1; } 50% { opacity: 0.5; } }
 </style>

traffic-gen-ui/src/components/TestBuilder.vue

@@ -13,14 +13,24 @@
     </div>
 
     <div class="form-row">
-      <label>Base Flow</label>
+      <label>Destination IP</label>
-      <select v-model="form.flow_id">
+      <input v-model="form.dst_ip" placeholder="10.100.0.1" />
-        <option value="" disabled>Select a flow...</option>
+    </div>
-        <option v-for="f in flows" :key="f.id" :value="f.id">
+
-          {{ f.name || f.dst_ip }} ({{ f.protocol }})
+    <div class="form-row-pair">
-        </option>
+      <div class="form-row">
+        <label>Protocol</label>
+        <select v-model="form.protocol">
+          <option value="udp">UDP</option>
+          <option value="icmp">ICMP</option>
+          <option value="tcp">TCP</option>
         </select>
       </div>
+      <div class="form-row">
+        <label>Source IP</label>
+        <input v-model="form.src_ip" placeholder="auto" />
+      </div>
+    </div>
 
     <div class="form-row">
       <label>Frame Sizes</label>
@@ -37,8 +47,13 @@
         <input v-model.number="form.trial_duration" type="number" min="5" max="300" />
       </div>
       <div class="form-row">
-        <label>Max Rate (pps)</label>
+        <label>Max Rate</label>
-        <input v-model.number="form.max_rate_pps" type="number" min="10" max="100000" />
+        <input v-model.number="form.max_rate_val" type="number" min="1" step="any" />
+        <select v-model="form.max_rate_unit" class="rate-unit-standalone">
+          <option value="pps">pps</option>
+          <option value="kbps">Kbps</option>
+          <option value="mbps">Mbps</option>
+        </select>
       </div>
     </div>
 
@@ -47,7 +62,7 @@
       <input v-model.number="form.acceptable_loss_pct" type="number" min="0" max="100" step="0.1" />
     </div>
 
-    <button class="btn btn-accent" @click="create" :disabled="!form.flow_id">
+    <button class="btn btn-accent" @click="create" :disabled="!form.dst_ip">
       Create & Run Test
     </button>
 
@@ -67,36 +82,60 @@
 import { reactive, ref, onMounted } from 'vue'
 import { api } from '../api.js'
 
-const props = defineProps({ flows: Array })
 const emit = defineEmits(['created', 'refresh'])
 
-const standardSizes = [64, 128, 256, 512, 1024, 1280, 1518]
+const standardSizes = [64, 128, 256, 512, 1024, 1280, 1518, 2048, 4096, 9000]
 const presets = ref({})
 
 const form = reactive({
   type: 'throughput',
-  flow_id: '',
+  dst_ip: '',
+  src_ip: '',
+  protocol: 'udp',
   frame_sizes: [64, 512, 1518],
   trial_duration: 30,
-  max_rate_pps: 10000,
+  max_rate_val: 10,
+  max_rate_unit: 'mbps',
   acceptable_loss_pct: 0.0,
 })
 
+function computePps(val, unit) {
+  if (unit === 'kbps') return Math.max(1, Math.round((val * 1000) / (512 * 8)))
+  if (unit === 'mbps') return Math.max(1, Math.round((val * 1_000_000) / (512 * 8)))
+  return Math.round(val)
+}
+
 onMounted(async () => {
   try { const r = await api.presets(); presets.value = r.presets || r } catch (_) {}
 })
 
 async function create() {
   try {
-    const test = await api.createTest({ ...form })
+    const payload = {
+      type: form.type,
+      flow_config: {
+        dst_ip: form.dst_ip,
+        src_ip: form.src_ip || 'auto',
+        protocol: form.protocol,
+        src_port: 50000,
+        dst_port: 5001,
+      },
+      frame_sizes: form.frame_sizes,
+      trial_duration: form.trial_duration,
+      max_rate_pps: computePps(form.max_rate_val, form.max_rate_unit),
+      acceptable_loss_pct: form.acceptable_loss_pct,
+    }
+    const test = await api.createTest(payload)
     await api.startTest(test.id)
     emit('created')
   } catch (e) { alert(e.message) }
 }
 
 async function loadPreset(name) {
+  const dstIp = prompt('Destination IP for this preset:', '10.100.0.100')
+  if (!dstIp) return
   try {
-    await api.loadPreset(name, {})
+    await api.loadPreset(name, { dst_ip: dstIp })
     emit('refresh')
   } catch (e) { alert(e.message) }
 }
@@ -109,6 +148,7 @@ h4 { font-size: 13px; margin: 16px 0 8px; color: var(--muted); }
 .form-row label { display: block; font-size: 11px; color: var(--muted); margin-bottom: 3px; text-transform: uppercase; letter-spacing: 0.05em; }
 .form-row input, .form-row select { width: 100%; }
 .form-row-pair { display: grid; grid-template-columns: 1fr 1fr; gap: 8px; }
+.rate-unit-standalone { width: 100%; margin-top: 4px; }
 .frame-sizes { display: flex; flex-wrap: wrap; gap: 8px; }
 .checkbox-label { font-size: 12px; display: flex; align-items: center; gap: 4px; color: var(--text); cursor: pointer; }
 .btn { padding: 8px 16px; font-weight: 600; font-size: 13px; width: 100%; margin-top: 8px; }

traffic-gen-ui/src/components/TestRunner.vue

@@ -1,67 +1,145 @@
 <template>
   <div class="test-runner">
-    <div v-if="!tests.length" class="empty">No tests created yet. Use the Test Builder tab.</div>
+    <h3>Running Tests</h3>
-    <div v-for="t in tests" :key="t.id" class="test-card" :class="'state-' + t.state">
+    <div v-if="!tests.length" class="empty">No tests yet. Create one above and click "Create & Run Test".</div>
+
+    <div v-for="t in sortedTests" :key="t.id" class="test-card" :class="'state-' + t.state">
       <div class="test-header">
-        <div>
+        <div class="test-title">
           <strong>{{ t.type }}</strong>
-          <span class="test-state">{{ t.state }}</span>
+          <span class="test-state" :class="'ts-' + t.state">{{ t.state }}</span>
+          <span v-if="t.frame_sizes" class="test-sizes">{{ t.frame_sizes.length }} frame sizes</span>
         </div>
         <div class="test-actions">
           <button v-if="t.state === 'idle'" class="btn-sm btn-go" @click="start(t.id)">Start</button>
           <button v-if="t.state === 'running'" class="btn-sm btn-stop" @click="stop(t.id)">Stop</button>
+          <button v-if="t.state === 'complete' || t.state === 'error'" class="btn-sm btn-del" @click="del(t.id)">Remove</button>
         </div>
       </div>
 
+      <!-- RUNNING: live progress -->
       <div v-if="t.state === 'running'" class="progress-section">
-        <div class="progress-label">Running {{ t.type }} test...</div>
+        <div class="progress-detail">
+          <span v-if="t.progress">{{ t.progress.message }}</span>
+          <span v-else>Starting...</span>
+          <span v-if="t.progress" class="progress-counter">
+            {{ (t.progress.completed_sizes || []).length }}/{{ t.progress.total_frames }} sizes done
+          </span>
+        </div>
         <div class="progress-bar">
           <div class="progress-fill" :style="{ width: progressPct(t) + '%' }"></div>
         </div>
+        <!-- Show partial results as they come in -->
+        <div v-if="t.results && Object.keys(t.results).length" class="partial-results">
+          <div v-for="(val, size) in t.results" :key="size" class="partial-item">
+            <span class="partial-size">{{ size }}B</span>
+            <span class="partial-val" v-if="!Array.isArray(val) && val.max_throughput_pps != null">{{ val.max_throughput_pps }} pps</span>
+            <span class="partial-val" v-else-if="!Array.isArray(val) && val.avg_ms != null">{{ val.avg_ms }}ms avg</span>
+            <span class="partial-val" v-else-if="!Array.isArray(val) && val.max_burst_frames != null">{{ val.max_burst_frames }} frames</span>
+            <span class="partial-val" v-else-if="Array.isArray(val)">{{ val.length }} rate steps</span>
+          </div>
+        </div>
       </div>
 
-      <div v-if="t.state === 'complete' && t.results" class="results-preview">
+      <!-- ERROR -->
+      <div v-if="t.state === 'error'" class="error-msg">{{ t.error || 'Test failed' }}</div>
+
+      <!-- COMPLETE: inline results summary -->
+      <div v-if="t.state === 'complete' && t.results && Object.keys(t.results).length" class="results-preview">
+        <!-- Frame Loss has a different structure (array per size) -->
+        <template v-if="t.type === 'frame_loss'">
+          <div v-for="(rates, size) in t.results" :key="size" class="fl-section">
+            <div class="fl-title">Frame Size: {{ size }} B</div>
             <table>
+              <thead>
+                <tr>
+                  <th>Rate %</th>
+                  <th>Rate (pps)</th>
+                  <th>TX Packets</th>
+                  <th>RX Packets</th>
+                  <th>Loss %</th>
+                </tr>
+              </thead>
+              <tbody>
+                <tr v-for="r in rates" :key="r.rate_pct">
+                  <td class="mono">{{ r.rate_pct }}%</td>
+                  <td class="mono">{{ r.rate_pps }}</td>
+                  <td class="mono">{{ r.tx_packets }}</td>
+                  <td class="mono">{{ r.rx_packets }}</td>
+                  <td class="mono">{{ r.loss_pct }}%</td>
+                </tr>
+              </tbody>
+            </table>
+          </div>
+        </template>
+        <!-- Other test types: single value per size -->
+        <table v-else>
           <thead>
             <tr>
               <th>Frame Size</th>
               <th v-if="t.type === 'throughput'">Max Rate (pps)</th>
-              <th v-if="t.type === 'latency'">Avg Latency (ms)</th>
+              <th v-if="t.type === 'throughput'">Throughput</th>
-              <th v-if="t.type === 'frame_loss'">Loss @ Max %</th>
+              <th v-if="t.type === 'latency'">Avg (ms)</th>
+              <th v-if="t.type === 'latency'">Min/Max (ms)</th>
               <th v-if="t.type === 'back_to_back'">Max Burst</th>
             </tr>
           </thead>
           <tbody>
             <tr v-for="(val, size) in t.results" :key="size">
               <td>{{ size }} B</td>
-              <td v-if="t.type === 'throughput'">{{ val.max_rate_pps || val }}</td>
+              <td v-if="t.type === 'throughput'" class="mono">{{ val.max_throughput_pps || '-' }}</td>
-              <td v-if="t.type === 'latency'">{{ val.avg_ms || val }}</td>
+              <td v-if="t.type === 'throughput'" class="mono">{{ formatMbps(val) }}</td>
-              <td v-if="t.type === 'frame_loss'">{{ val.loss_pct || val }}%</td>
+              <td v-if="t.type === 'latency'" class="mono">{{ val.avg_ms != null ? val.avg_ms.toFixed(2) : '-' }}</td>
-              <td v-if="t.type === 'back_to_back'">{{ val.max_burst || val }}</td>
+              <td v-if="t.type === 'latency'" class="mono">{{ val.min_ms != null ? val.min_ms.toFixed(2) + ' / ' + val.max_ms.toFixed(2) : '-' }}</td>
+              <td v-if="t.type === 'back_to_back'" class="mono">{{ val.max_burst_frames ?? '-' }}</td>
             </tr>
           </tbody>
         </table>
       </div>
 
       <div class="test-meta">
-        <span>Created: {{ t.created_at || '-' }}</span>
         <span v-if="t.started_at">Started: {{ t.started_at }}</span>
         <span v-if="t.completed_at">Completed: {{ t.completed_at }}</span>
+        <span v-if="t.state === 'running' && t.started_at">Elapsed: {{ elapsed(t) }}</span>
       </div>
     </div>
   </div>
 </template>
 
 <script setup>
+import { computed } from 'vue'
 import { api } from '../api.js'
 
 const props = defineProps({ tests: Array })
 const emit = defineEmits(['refresh'])
 
+const sortedTests = computed(() => {
+  const order = { running: 0, idle: 1, complete: 2, error: 3 }
+  return [...(props.tests || [])].sort((a, b) => (order[a.state] ?? 9) - (order[b.state] ?? 9))
+})
+
 function progressPct(t) {
-  if (!t.results || !t.frame_sizes) return 20
+  if (!t.progress || !t.progress.total_frames) return 10
-  const done = Object.keys(t.results).length
+  const done = (t.progress.completed_sizes || []).length
-  return Math.min(95, (done / t.frame_sizes.length) * 100)
+  const partial = t.progress.frame_idx > done ? 0.5 : 0
+  return Math.min(95, ((done + partial) / t.progress.total_frames) * 100)
+}
+
+function formatMbps(val) {
+  const pps = val.max_throughput_pps || 0
+  const fs = val.frame_size || 64
+  if (!pps) return '-'
+  const mbps = (pps * fs * 8) / 1_000_000
+  return mbps.toFixed(1) + ' Mbps'
+}
+
+function elapsed(t) {
+  if (!t.started_at) return ''
+  const start = new Date(t.started_at).getTime()
+  const secs = Math.round((Date.now() - start) / 1000)
+  const m = Math.floor(secs / 60)
+  const s = secs % 60
+  return m > 0 ? `${m}m ${s}s` : `${s}s`
 }
 
 async function start(id) {
@@ -70,28 +148,48 @@ async function start(id) {
 async function stop(id) {
   try { await api.stopTest(id); emit('refresh') } catch (e) { alert(e.message) }
 }
+async function del(id) {
+  emit('refresh')
+}
 </script>
 
 <style scoped>
-.empty { color: var(--muted); padding: 20px; text-align: center; }
+h3 { font-size: 15px; margin-bottom: 12px; color: var(--accent); }
+.empty { color: var(--muted); padding: 16px; text-align: center; font-size: 13px; }
 .test-card { background: var(--card-bg); border: 1px solid var(--border); border-radius: var(--radius); padding: 12px; margin-bottom: 10px; }
 .test-card.state-running { border-color: var(--accent); }
 .test-card.state-complete { border-color: var(--success); }
-.test-header { display: flex; justify-content: space-between; align-items: center; margin-bottom: 8px; }
+.test-card.state-error { border-color: var(--danger); }
-.test-header strong { font-size: 14px; text-transform: capitalize; }
+.test-header { display: flex; justify-content: space-between; align-items: center; }
-.test-state { font-size: 11px; padding: 2px 8px; border-radius: 10px; margin-left: 8px; font-weight: 600; background: rgba(113,128,150,0.2); color: var(--muted); }
+.test-title { display: flex; align-items: center; gap: 8px; }
-.state-running .test-state { background: rgba(79,156,249,0.15); color: var(--accent); }
+.test-title strong { font-size: 14px; text-transform: capitalize; }
-.state-complete .test-state { background: rgba(72,187,120,0.15); color: var(--success); }
+.test-state { font-size: 11px; padding: 2px 8px; border-radius: 10px; font-weight: 600; }
+.ts-idle { background: rgba(113,128,150,0.2); color: var(--muted); }
+.ts-running { background: rgba(79,156,249,0.15); color: var(--accent); }
+.ts-complete { background: rgba(72,187,120,0.15); color: var(--success); }
+.ts-error { background: rgba(252,129,129,0.15); color: var(--danger); }
+.test-sizes { font-size: 11px; color: var(--muted); }
 .test-actions { display: flex; gap: 4px; }
 .btn-sm { padding: 3px 10px; font-size: 11px; font-weight: 600; border-radius: 6px; }
 .btn-go { background: var(--success); color: #fff; }
 .btn-stop { background: var(--warning); color: #000; }
-.progress-section { margin: 8px 0; }
+.btn-del { background: rgba(252,129,129,0.15); color: var(--danger); }
-.progress-label { font-size: 12px; color: var(--muted); margin-bottom: 4px; }
+.progress-section { margin: 10px 0; }
+.progress-detail { display: flex; justify-content: space-between; font-size: 12px; color: var(--muted); margin-bottom: 6px; font-family: monospace; }
+.progress-counter { color: var(--accent); font-weight: 600; }
 .progress-bar { height: 6px; background: var(--border); border-radius: 3px; overflow: hidden; }
 .progress-fill { height: 100%; background: var(--accent); border-radius: 3px; transition: width 0.5s; }
-.results-preview table { width: 100%; border-collapse: collapse; margin-top: 8px; }
+.partial-results { display: flex; gap: 8px; flex-wrap: wrap; margin-top: 8px; }
+.partial-item { background: rgba(72,187,120,0.1); border: 1px solid rgba(72,187,120,0.2); padding: 2px 8px; border-radius: 6px; font-size: 11px; }
+.partial-size { font-weight: 600; color: var(--success); }
+.partial-val { color: var(--text); margin-left: 4px; font-family: monospace; }
+.error-msg { color: var(--danger); font-size: 13px; padding: 8px 0; }
+.results-preview { margin-top: 10px; }
+.results-preview table { width: 100%; border-collapse: collapse; }
 .results-preview th { font-size: 11px; color: var(--muted); text-align: left; padding: 4px 8px; border-bottom: 1px solid var(--border); }
-.results-preview td { font-size: 13px; padding: 4px 8px; font-family: monospace; }
+.results-preview td { font-size: 13px; padding: 4px 8px; }
+.mono { font-family: monospace; }
+.fl-section { margin-bottom: 12px; }
+.fl-title { font-size: 12px; font-weight: 600; color: var(--accent); margin-bottom: 4px; }
 .test-meta { display: flex; gap: 12px; margin-top: 8px; font-size: 11px; color: var(--muted); }
 </style>

traffic-gen-ui/vite.config.js

@@ -2,6 +2,7 @@ import { defineConfig } from 'vite'
 import vue from '@vitejs/plugin-vue'
 
 export default defineConfig({
+  base: '/traffic/',
   plugins: [vue()],
   server: {
     proxy: {

traffic-gen/Dockerfile

@@ -1,6 +1,6 @@
 FROM python:3.11-slim
 RUN apt-get update && apt-get install -y --no-install-recommends \
-    tcpreplay libpcap-dev procps && rm -rf /var/lib/apt/lists/*
+    tcpreplay libpcap-dev procps iputils-ping && rm -rf /var/lib/apt/lists/*
 RUN pip install --no-cache-dir flask scapy psutil
 COPY . /traffic-gen/
 WORKDIR /traffic-gen

traffic-gen/engine/packet_builder.py

@@ -87,7 +87,7 @@ def build_packet(flow_config: dict, seq: int = 0):
     # --- Layer 3 ---
     ip_kwargs = {'dst': flow_config['dst_ip']}
     src_ip = flow_config.get('src_ip')
-    if src_ip:
+    if src_ip and src_ip != 'auto':
         ip_kwargs['src'] = src_ip
 
     dscp = flow_config.get('dscp', 0)
@@ -99,13 +99,13 @@ def build_packet(flow_config: dict, seq: int = 0):
 
     # --- Layer 4 ---
     if protocol == 'udp':
-        src_port = flow_config.get('src_port', 12000)
+        src_port = flow_config.get('src_port') or 12000
-        dst_port = flow_config.get('dst_port', 5001)
+        dst_port = flow_config.get('dst_port') or 5001
         pkt = pkt / UDP(sport=int(src_port), dport=int(dst_port))
         header_overhead += 8
     elif protocol == 'tcp':
-        src_port = flow_config.get('src_port', 12000)
+        src_port = flow_config.get('src_port') or 12000
-        dst_port = flow_config.get('dst_port', 80)
+        dst_port = flow_config.get('dst_port') or 80
         pkt = pkt / TCP(sport=int(src_port), dport=int(dst_port), flags='S')
         header_overhead += 20
     elif protocol == 'icmp':

traffic-gen/engine/responder.py

@@ -1,185 +1,204 @@
 """
-Responder - listens for TGEN-tagged packets and collects receive statistics.
+Responder - high-performance UDP packet receiver for TGEN traffic.
 
-Two sub-modes:
+Uses multiple receiver threads on SO_REUSEPORT UDP sockets for parallel
-  - echo: swaps src/dst MAC and IP, sends packet back with receive timestamp
+packet processing. Each thread has its own socket and stats counters
-  - log:  records rx stats only, exposed via API
+to avoid contention.
 """
 
 import logging
+import os
+import socket
 import struct
 import threading
 import time
 
-from scapy.all import (
+from engine.packet_builder import MAGIC, HEADER_LEN
-    AsyncSniffer, Ether, IP, Raw, send, conf,
-)
-
-from engine.packet_builder import MAGIC, HEADER_LEN, parse_payload
 
 log = logging.getLogger(__name__)
-conf.verb = 0
+
+DEFAULT_LISTEN_PORT = 5001
+RECV_BUF_SIZE = 16 * 1024 * 1024
+NUM_WORKERS = int(os.environ.get('RESPONDER_WORKERS', 4))
+
+
+class _WorkerStats:
+    """Per-worker stats — no sharing, no locks."""
+    __slots__ = ('rx_packets', 'rx_bytes', 'out_of_order', 'duplicates',
+                 'last_seq', 'lat_buf', 'lat_idx', 'lat_count')
+
+    def __init__(self):
+        self.rx_packets = 0
+        self.rx_bytes = 0
+        self.out_of_order = 0
+        self.duplicates = 0
+        self.last_seq = -1
+        self.lat_buf = [0.0] * 4096
+        self.lat_idx = 0
+        self.lat_count = 0
 
 
 class Responder:
-    """Listens for TGEN packets on an interface and collects stats."""
+    def __init__(self, mode: str = 'log', listen_port: int = DEFAULT_LISTEN_PORT):
-
-    def __init__(self, mode: str = 'log'):
-        """
-        Args:
-            mode: 'echo' to reflect packets back, 'log' to only record stats.
-        """
         self._mode = mode
-        self._lock = threading.Lock()
+        self._listen_port = listen_port
-        self._sniffer = None
+        self._sockets = []
+        self._threads = []
+        self._workers = []
         self._running = False
-
+        self._stop_event = threading.Event()
-        # Stats
-        self._rx_packets = 0
-        self._rx_bytes = 0
-        self._latency_samples = []   # list of (latency_ms,)
-        self._seen_seqs = set()
-        self._last_seq = -1
-        self._out_of_order = 0
-        self._duplicates = 0
-
-    # ------------------------------------------------------------------
-    # Control
-    # ------------------------------------------------------------------
 
     def start(self, interface: str = None):
-        """Start sniffing for TGEN packets."""
         if self._running:
-            log.warning('Responder already running')
             return
 
-        bpf_filter = 'ip'  # broad filter; we check magic in callback
+        self._stop_event.clear()
-        kwargs = {
+        n = NUM_WORKERS
-            'prn': self._handle_packet,
-            'store': False,
-            'filter': bpf_filter,
-        }
-        if interface:
-            kwargs['iface'] = interface
 
-        self._sniffer = AsyncSniffer(**kwargs)
+        for i in range(n):
-        self._sniffer.start()
+            sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+            sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+            sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEPORT, 1)
+            try:
+                sock.setsockopt(socket.SOL_SOCKET, socket.SO_RCVBUF, RECV_BUF_SIZE)
+            except OSError:
+                pass
+            sock.settimeout(0.5)
+            sock.bind(('0.0.0.0', self._listen_port))
+            self._sockets.append(sock)
+
+            ws = _WorkerStats()
+            self._workers.append(ws)
+
+            t = threading.Thread(target=self._recv_loop, args=(sock, ws),
+                                 daemon=True, name=f'responder-rx-{i}')
+            self._threads.append(t)
+            t.start()
+
+        actual_buf = self._sockets[0].getsockopt(socket.SOL_SOCKET, socket.SO_RCVBUF)
         self._running = True
-        log.info('Responder started on interface=%s mode=%s', interface or 'all', self._mode)
+        log.info('Responder started on port=%d mode=%s workers=%d rcvbuf=%d',
+                 self._listen_port, self._mode, n, actual_buf)
 
     def stop(self):
-        """Stop sniffing."""
+        self._stop_event.set()
-        if self._sniffer and self._running:
+        for t in self._threads:
-            self._sniffer.stop()
+            if t.is_alive():
+                t.join(timeout=3)
+        for s in self._sockets:
+            try:
+                s.close()
+            except Exception:
+                pass
+        self._sockets.clear()
+        self._threads.clear()
+        self._workers.clear()
         self._running = False
         log.info('Responder stopped')
 
     def is_running(self) -> bool:
         return self._running
 
-    # ------------------------------------------------------------------
-    # Stats
-    # ------------------------------------------------------------------
-
     def get_stats(self) -> dict:
-        with self._lock:
+        rx_packets = 0
+        rx_bytes = 0
+        out_of_order = 0
+        duplicates = 0
+        all_lat = []
+
+        for ws in self._workers:
+            rx_packets += ws.rx_packets
+            rx_bytes += ws.rx_bytes
+            out_of_order += ws.out_of_order
+            duplicates += ws.duplicates
+            n = min(ws.lat_count, len(ws.lat_buf))
+            if n > 0:
+                all_lat.extend(ws.lat_buf[:n] if ws.lat_count <= len(ws.lat_buf) else ws.lat_buf[:])
+
         latency = {}
-            if self._latency_samples:
+        if all_lat:
-                vals = self._latency_samples
+            avg = sum(all_lat) / len(all_lat)
+            mn = min(all_lat)
+            mx = max(all_lat)
+            jitter = 0.0
+            if len(all_lat) > 1:
+                jitter = sum(abs(all_lat[i] - all_lat[i-1]) for i in range(1, len(all_lat))) / (len(all_lat) - 1)
             latency = {
-                    'min_ms': round(min(vals), 3),
+                'min_ms': round(mn, 3),
-                    'max_ms': round(max(vals), 3),
+                'max_ms': round(mx, 3),
-                    'avg_ms': round(sum(vals) / len(vals), 3),
+                'avg_ms': round(avg, 3),
-                    'jitter_ms': round(
+                'jitter_ms': round(jitter, 3),
-                        sum(abs(vals[i] - vals[i - 1]) for i in range(1, len(vals))) / max(1, len(vals) - 1),
+                'samples': len(all_lat),
-                        3
-                    ) if len(vals) > 1 else 0.0,
-                    'samples': len(vals),
             }
+
         return {
-                'rx_packets': self._rx_packets,
+            'rx_packets': rx_packets,
-                'rx_bytes': self._rx_bytes,
+            'rx_bytes': rx_bytes,
-                'out_of_order': self._out_of_order,
+            'out_of_order': out_of_order,
-                'duplicates': self._duplicates,
+            'duplicates': duplicates,
             'latency': latency,
             'running': self._running,
         }
 
     def reset_stats(self):
-        with self._lock:
+        for ws in self._workers:
-            self._rx_packets = 0
+            ws.rx_packets = 0
-            self._rx_bytes = 0
+            ws.rx_bytes = 0
-            self._latency_samples = []
+            ws.last_seq = -1
-            self._seen_seqs.clear()
+            ws.out_of_order = 0
-            self._last_seq = -1
+            ws.duplicates = 0
-            self._out_of_order = 0
+            ws.lat_idx = 0
-            self._duplicates = 0
+            ws.lat_count = 0
         log.info('Responder stats reset')
 
-    # ------------------------------------------------------------------
+    def _recv_loop(self, sock, ws):
-    # Packet handling
+        """Per-worker receive loop."""
-    # ------------------------------------------------------------------
+        echo = self._mode == 'echo'
+        recvfrom = sock.recvfrom
+        time_ns = time.time_ns
+        stop_is_set = self._stop_event.is_set
+        lat_buf = ws.lat_buf
+        lat_buf_len = len(lat_buf)
+        magic = MAGIC
 
-    def _handle_packet(self, pkt):
+        while not stop_is_set():
-        """Process a received packet, checking for TGEN magic bytes."""
-        if not pkt.haslayer(Raw):
-            return
-
-        payload = bytes(pkt[Raw].load)
-        parsed = parse_payload(payload)
-        if parsed is None:
-            return
-
-        seq, sender_ts_ns = parsed
-        rx_time_ns = time.time_ns()
-        pkt_len = len(bytes(pkt))
-
-        # Compute one-way latency (only meaningful if clocks are synced)
-        latency_ms = (rx_time_ns - sender_ts_ns) / 1_000_000
-
-        with self._lock:
-            self._rx_packets += 1
-            self._rx_bytes += pkt_len
-
-            # Duplicate detection
-            if seq in self._seen_seqs:
-                self._duplicates += 1
-            else:
-                self._seen_seqs.add(seq)
-                # Keep set bounded
-                if len(self._seen_seqs) > 100000:
-                    # Remove oldest entries (approximate)
-                    to_remove = sorted(self._seen_seqs)[:50000]
-                    self._seen_seqs -= set(to_remove)
-
-            # Out-of-order detection
-            if seq < self._last_seq and seq not in self._seen_seqs:
-                self._out_of_order += 1
-            self._last_seq = seq
-
-            # Record latency (only if plausible: 0 < latency < 60s)
-            if 0 < latency_ms < 60000:
-                self._latency_samples.append(latency_ms)
-                if len(self._latency_samples) > 10000:
-                    self._latency_samples = self._latency_samples[-5000:]
-
-        # Echo mode: swap and send back
-        if self._mode == 'echo' and pkt.haslayer(Ether) and pkt.haslayer(IP):
             try:
-                echo_pkt = pkt.copy()
+                data, addr = recvfrom(65535)
-                # Swap MACs
+            except socket.timeout:
-                echo_pkt[Ether].src, echo_pkt[Ether].dst = pkt[Ether].dst, pkt[Ether].src
+                continue
-                # Swap IPs
+            except OSError:
-                echo_pkt[IP].src, echo_pkt[IP].dst = pkt[IP].dst, pkt[IP].src
+                if stop_is_set():
-                # Append receive timestamp to payload
+                    break
-                rx_ts_bytes = struct.pack('!Q', rx_time_ns)
+                raise
-                echo_pkt[Raw].load = payload + rx_ts_bytes
+
-                # Clear checksums so Scapy recalculates
+            rx_ns = time_ns()
-                del echo_pkt[IP].chksum
+            dlen = len(data)
-                if echo_pkt.haslayer('UDP'):
+
-                    del echo_pkt['UDP'].chksum
+            if dlen < HEADER_LEN or data[:4] != magic:
-                elif echo_pkt.haslayer('TCP'):
+                continue
-                    del echo_pkt['TCP'].chksum
+
-                send(echo_pkt[IP], verbose=0)
+            seq = int.from_bytes(data[4:8], 'big')
-            except Exception as e:
+            sender_ns = int.from_bytes(data[8:16], 'big')
-                log.debug('Echo send error: %s', e)
+
+            ws.rx_packets += 1
+            ws.rx_bytes += dlen
+
+            last = ws.last_seq
+            if seq == last:
+                ws.duplicates += 1
+            elif seq < last:
+                ws.out_of_order += 1
+            ws.last_seq = seq
+
+            lat_ms = (rx_ns - sender_ns) / 1_000_000
+            if 0 < lat_ms < 60000:
+                idx = ws.lat_idx
+                lat_buf[idx] = lat_ms
+                ws.lat_idx = (idx + 1) % lat_buf_len
+                ws.lat_count += 1
+
+            if echo:
+                try:
+                    sock.sendto(data + struct.pack('!Q', rx_ns), addr)
+                except Exception:
+                    pass

traffic-gen/engine/rfc2544.py

@@ -7,13 +7,17 @@ RFC 2544 test implementations:
 """
 
 import logging
+import socket
+import struct
 import threading
 import time
+import urllib.request
+import json
 from typing import Dict, List, Optional
 
 from scapy.all import send, sr, conf, IP, ICMP
 
-from engine.packet_builder import build_packet, parse_payload
+from engine.packet_builder import build_packet, parse_payload, MAGIC
 
 log = logging.getLogger(__name__)
 conf.verb = 0
@@ -24,13 +28,14 @@ class _BaseTest:
 
     def __init__(self, test_id: str, flow_config: dict, frame_sizes: List[int],
                  trial_duration: float = 60, max_rate_pps: int = 10000,
-                 acceptable_loss_pct: float = 0.0):
+                 acceptable_loss_pct: float = 0.0, responder_url: str = None):
         self.test_id = test_id
         self.flow_config = dict(flow_config)
         self.frame_sizes = frame_sizes
         self.trial_duration = trial_duration
         self.max_rate_pps = max_rate_pps
         self.acceptable_loss_pct = acceptable_loss_pct
+        self.responder_url = responder_url  # e.g. "http://172.30.0.10:5053"
 
         self.state = 'idle'      # idle -> running -> complete/error
         self.results = {}
@@ -38,6 +43,11 @@ class _BaseTest:
         self.started_at = None
         self.completed_at = None
 
+        # Progress tracking
+        self._progress_msg = ''
+        self._current_frame_idx = 0
+        self._current_trial_tx = 0
+
         self._thread: Optional[threading.Thread] = None
         self._stop_event = threading.Event()
         self._lock = threading.Lock()
@@ -80,72 +90,155 @@ class _BaseTest:
     def _is_stopped(self) -> bool:
         return self._stop_event.is_set()
 
+    def _responder_reset(self):
+        """Reset responder stats before a trial."""
+        if not self.responder_url:
+            return
+        try:
+            req = urllib.request.Request(
+                f'{self.responder_url}/responder/reset', method='POST',
+                data=b'{}', headers={'Content-Type': 'application/json'})
+            urllib.request.urlopen(req, timeout=3)
+        except Exception as e:
+            log.warning('Responder reset failed: %s', e)
+
+    def _responder_stats(self) -> Optional[dict]:
+        """Query responder for rx stats after a trial."""
+        if not self.responder_url:
+            return None
+        try:
+            req = urllib.request.Request(f'{self.responder_url}/responder/stats')
+            resp = urllib.request.urlopen(req, timeout=5)
+            return json.loads(resp.read())
+        except Exception as e:
+            log.warning('Responder stats query failed: %s', e)
+            return None
+
     def _send_trial(self, frame_size: int, rate_pps: int, duration: float):
         """Send packets at a given rate for a duration. Returns (tx_count, rx_count, latencies)."""
         flow = dict(self.flow_config)
         flow['frame_size'] = frame_size
+        protocol = flow.get('protocol', 'udp').lower()
+
+        # Reset responder counters before trial
+        self._responder_reset()
 
-        interval = 1.0 / rate_pps if rate_pps > 0 else 1.0
         tx_count = 0
         rx_count = 0
         latencies = []
-        protocol = flow.get('protocol', 'udp').lower()
-
         start = time.time()
-        seq = 0
 
+        if protocol == 'icmp':
+            # ICMP: use sr() to measure latency from responses
+            seq = 0
             while (time.time() - start) < duration and not self._is_stopped():
                 pkt = build_packet(flow, seq=seq)
                 seq += 1
-
-            if protocol == 'icmp':
-                # Use sr() for ICMP to get responses
                 answered, _ = sr(pkt[IP], timeout=1, verbose=0)
                 tx_count += 1
                 for sent_pkt, recv_pkt in answered:
                     rx_count += 1
                     rtt_ms = (recv_pkt.time - sent_pkt.sent_time) * 1000
                     latencies.append(rtt_ms)
-            else:
-                send(pkt[IP], verbose=0)
-                tx_count += 1
-
-            # Rate limiting
                 elapsed = time.time() - start
-            expected_sent = elapsed * rate_pps
+                expected = elapsed * rate_pps
-            if tx_count > expected_sent:
+                if tx_count > expected:
-                sleep_time = (tx_count - expected_sent) / rate_pps
+                    sleep_time = (tx_count - expected) / rate_pps
                     if sleep_time > 0:
                         self._stop_event.wait(min(sleep_time, 0.1))
+        else:
+            # UDP/TCP: high-performance raw socket path
+            dst_ip = flow['dst_ip']
+            pkt_template = build_packet(flow, seq=0)
+            ip_template = bytes(pkt_template[pkt_template.firstlayer().payload.__class__])
+            magic_offset = ip_template.find(MAGIC)
+
+            # Find and zero UDP checksum in template so receivers accept packets
+            # IP header length from IHL field (byte 0, low nibble) * 4
+            ip_ihl = (ip_template[0] & 0x0F) * 4
+            ip_proto = ip_template[9]  # protocol field
+            udp_csum_offset = ip_ihl + 6 if ip_proto == 17 else -1  # 17 = UDP
+
+            raw_sock = socket.socket(socket.AF_INET, socket.SOCK_RAW, socket.IPPROTO_RAW)
+            raw_sock.setsockopt(socket.IPPROTO_IP, socket.IP_HDRINCL, 1)
+
+            batch_size = max(1, min(rate_pps // 5, 500))
+            interval = batch_size / rate_pps if rate_pps > 0 else 1.0
+            seq = 0
+
+            try:
+                while (time.time() - start) < duration and not self._is_stopped():
+                    batch_start = time.time()
+                    for _ in range(batch_size):
+                        pkt_bytes = bytearray(ip_template)
+                        if magic_offset >= 0:
+                            struct.pack_into('!I', pkt_bytes, magic_offset + 4, seq)
+                            struct.pack_into('!Q', pkt_bytes, magic_offset + 8, time.time_ns())
+                        pkt_bytes[10:12] = b'\x00\x00'  # zero IP checksum
+                        if udp_csum_offset > 0:
+                            pkt_bytes[udp_csum_offset:udp_csum_offset + 2] = b'\x00\x00'
+                        try:
+                            raw_sock.sendto(bytes(pkt_bytes), (dst_ip, 0))
+                            tx_count += 1
+                        except Exception:
+                            pass
+                        seq += 1
+                    batch_elapsed = time.time() - batch_start
+                    sleep_time = interval - batch_elapsed
+                    if sleep_time > 0:
+                        self._stop_event.wait(sleep_time)
+            finally:
+                raw_sock.close()
+
+        # Query responder for actual rx stats (UDP/TCP path)
+        if protocol != 'icmp':
+            resp_stats = self._responder_stats()
+            if resp_stats and resp_stats.get('rx_packets', 0) > 0:
+                rx_count = resp_stats['rx_packets']
+                lat = resp_stats.get('latency', {})
+                if lat.get('samples', 0) > 0:
+                    latencies = [lat['avg_ms']]  # Use avg as representative
 
-        # For non-ICMP, we can't easily measure rx without a responder.
-        # rx_count stays 0 for UDP/TCP unless a responder is configured.
         return tx_count, rx_count, latencies
 
     def get_info(self) -> dict:
-        return {
+        # Reverse-lookup the slug for this test class
+        type_slug = next((k for k, v in TEST_TYPES.items() if v is self.__class__), self.__class__.__name__)
+        info = {
+            'id': self.test_id,
             'test_id': self.test_id,
-            'type': self.__class__.__name__,
+            'type': type_slug,
             'state': self.state,
             'results': self.results,
             'error': self.error,
+            'frame_sizes': self.frame_sizes,
             'started_at': self.started_at,
             'completed_at': self.completed_at,
         }
+        if self.state == 'running':
+            info['progress'] = {
+                'frame_idx': self._current_frame_idx,
+                'total_frames': len(self.frame_sizes),
+                'message': self._progress_msg,
+                'completed_sizes': list(self.results.keys()),
+            }
+        return info
 
 
 class ThroughputTest(_BaseTest):
     """Binary search for maximum throughput with acceptable loss."""
 
     def _run(self):
-        for fs in self.frame_sizes:
+        for idx, fs in enumerate(self.frame_sizes):
             if self._is_stopped():
                 break
 
+            self._current_frame_idx = idx
             low = 0
             high = self.max_rate_pps
             best_rate = 0
             convergence_threshold = max(1, int(self.max_rate_pps * 0.01))
+            step = 0
 
             log.info('Throughput test: frame_size=%d, searching [%d, %d] pps', fs, low, high)
 
@@ -153,19 +246,17 @@ class ThroughputTest(_BaseTest):
                 mid = (low + high) // 2
                 if mid == 0:
                     break
+                step += 1
 
+                self._progress_msg = f'Frame {fs}B: trial {step}, testing {mid} pps [{low}-{high}]'
                 tx, rx, _ = self._send_trial(fs, mid, self.trial_duration)
 
                 if tx == 0:
                     loss_pct = 100.0
-                else:
+                elif rx > 0:
-                    # For ICMP we have rx; for UDP assume zero loss if no responder
-                    protocol = self.flow_config.get('protocol', 'udp').lower()
-                    if protocol == 'icmp':
                     loss_pct = ((tx - rx) / tx) * 100
                 else:
-                        # Without responder, assume success (user should use responder for accurate test)
+                    loss_pct = 0.0  # No responder/ICMP — assume success
-                        loss_pct = 0.0
 
                 log.info('  frame=%d rate=%d tx=%d rx=%d loss=%.2f%%',
                          fs, mid, tx, rx, loss_pct)
@@ -189,10 +280,12 @@ class LatencyTest(_BaseTest):
     def _run(self):
         rate = self.flow_config.get('rate_pps', 100)
 
-        for fs in self.frame_sizes:
+        for idx, fs in enumerate(self.frame_sizes):
             if self._is_stopped():
                 break
 
+            self._current_frame_idx = idx
+            self._progress_msg = f'Frame {fs}B: sending at {rate} pps for {self.trial_duration}s'
             log.info('Latency test: frame_size=%d at %d pps for %ds', fs, rate, self.trial_duration)
             _, _, latencies = self._send_trial(fs, rate, self.trial_duration)
 
@@ -227,10 +320,11 @@ class FrameLossTest(_BaseTest):
     """Measure frame loss at decreasing rates (100%, 90%, 80%, ...)."""
 
     def _run(self):
-        for fs in self.frame_sizes:
+        for idx, fs in enumerate(self.frame_sizes):
             if self._is_stopped():
                 break
 
+            self._current_frame_idx = idx
             results_for_size = []
             for pct in range(100, 0, -10):
                 if self._is_stopped():
@@ -240,14 +334,16 @@ class FrameLossTest(_BaseTest):
                 if rate == 0:
                     continue
 
+                self._progress_msg = f'Frame {fs}B: testing at {pct}% rate ({rate} pps)'
                 log.info('FrameLoss test: frame_size=%d rate=%d (%d%%)', fs, rate, pct)
                 tx, rx, _ = self._send_trial(fs, rate, self.trial_duration)
 
-                protocol = self.flow_config.get('protocol', 'udp').lower()
+                if tx > 0 and rx > 0:
-                if tx > 0 and protocol == 'icmp':
                     loss_pct = ((tx - rx) / tx) * 100
+                elif tx > 0 and rx == 0:
+                    loss_pct = 0.0  # No responder — cannot measure
                 else:
-                    loss_pct = 0.0  # Cannot measure without responder for non-ICMP
+                    loss_pct = 100.0
 
                 results_for_size.append({
                     'rate_pct': pct,
@@ -264,7 +360,7 @@ class BackToBackTest(_BaseTest):
     """Find maximum burst length with zero loss."""
 
     def _run(self):
-        for fs in self.frame_sizes:
+        for idx, fs in enumerate(self.frame_sizes):
             if self._is_stopped():
                 break
 
@@ -273,6 +369,7 @@ class BackToBackTest(_BaseTest):
             best_burst = 0
             convergence = max(1, high // 100)
 
+            self._current_frame_idx = idx
             log.info('BackToBack test: frame_size=%d searching burst [%d, %d]', fs, low, high)
 
             while (high - low) > convergence and not self._is_stopped():
@@ -299,10 +396,12 @@ class BackToBackTest(_BaseTest):
                         send(pkt[IP], verbose=0)
                         tx_count += 1
 
-                if protocol == 'icmp' and tx_count > 0:
+                if tx_count > 0 and rx_count > 0:
                     loss_pct = ((tx_count - rx_count) / tx_count) * 100
+                elif tx_count > 0:
+                    loss_pct = 0.0  # No responder — cannot measure
                 else:
-                    loss_pct = 0.0  # Can't measure without responder
+                    loss_pct = 100.0
 
                 log.info('  burst=%d tx=%d rx=%d loss=%.2f%%', mid, tx_count, rx_count, loss_pct)
 

traffic-gen/engine/sender.py

@@ -4,6 +4,8 @@ FlowSender - manages traffic generation with background threads per flow.
 
 import logging
 import shutil
+import socket
+import struct
 import threading
 import time
 import urllib.request
@@ -11,7 +13,7 @@ import json
 
 from scapy.all import send, sendpfast, sr, conf
 
-from engine.packet_builder import build_packet, stamp_payload
+from engine.packet_builder import build_packet, stamp_payload, MAGIC, HEADER_LEN
 
 log = logging.getLogger(__name__)
 
@@ -139,14 +141,113 @@ class FlowSender:
         pkt_template = build_packet(flow, seq=0)
         pkt_bytes_len = len(bytes(pkt_template))
 
-        # Calculate sleep interval: send in batches for efficiency
-        batch_size = max(1, min(rate_pps // 10, 100))
-        interval = batch_size / rate_pps if rate_pps > 0 else 1.0
-
         seq = 0
         start_time = time.time()
         last_responder_poll = 0
-        log.info('Flow %s: starting send loop at %d pps for %ds', flow_id[:8], rate_pps, duration)
+        log.info('Flow %s: starting send loop at %d pps for %ds',
+                 flow_id[:8], rate_pps, duration)
+
+        # Capture responder baseline so we report deltas, not cumulative totals
+        responder_baseline_rx = 0
+        responder_baseline_bytes = 0
+        if responder_url:
+            try:
+                base = self._fetch_responder(responder_url)
+                responder_baseline_rx = base.get('rx_packets', 0)
+                responder_baseline_bytes = base.get('rx_bytes', 0)
+                # Also reset responder so baseline is clean
+                self._reset_responder(responder_url)
+                responder_baseline_rx = 0
+                responder_baseline_bytes = 0
+            except Exception:
+                pass
+
+        raw_sock = None
+        try:
+            if use_icmp_sr:
+                self._send_loop_icmp(flow_id, flow, stop_event, rate_pps, duration)
+                return
+
+            # --- High-performance path: raw socket ---
+            dst_ip = flow['dst_ip']
+            # Build template as raw IP bytes (strip Ethernet layer)
+            ip_template = bytes(pkt_template[pkt_template.firstlayer().payload.__class__])
+            # Find where TGEN magic starts in the IP-layer bytes
+            magic_offset = ip_template.find(MAGIC)
+
+            # Find and zero UDP checksum offset in template
+            ip_ihl = (ip_template[0] & 0x0F) * 4
+            ip_proto = ip_template[9]
+            udp_csum_offset = ip_ihl + 6 if ip_proto == 17 else -1  # 17 = UDP
+
+            raw_sock = socket.socket(socket.AF_INET, socket.SOCK_RAW, socket.IPPROTO_RAW)
+            raw_sock.setsockopt(socket.IPPROTO_IP, socket.IP_HDRINCL, 1)
+
+            # Adaptive batching: send bursts then sleep to hit target rate
+            batch_size = max(1, min(rate_pps // 5, 500))
+            interval = batch_size / rate_pps if rate_pps > 0 else 1.0
+
+            while not stop_event.is_set():
+                elapsed = time.time() - start_time
+                if duration and elapsed >= duration:
+                    break
+
+                batch_start = time.time()
+                sent_this_batch = 0
+
+                for _ in range(batch_size):
+                    pkt_bytes = bytearray(ip_template)
+                    if magic_offset >= 0:
+                        struct.pack_into('!I', pkt_bytes, magic_offset + 4, seq)
+                        struct.pack_into('!Q', pkt_bytes, magic_offset + 8, time.time_ns())
+                    pkt_bytes[10:12] = b'\x00\x00'  # zero IP checksum
+                    if udp_csum_offset > 0:
+                        pkt_bytes[udp_csum_offset:udp_csum_offset + 2] = b'\x00\x00'
+                    try:
+                        raw_sock.sendto(bytes(pkt_bytes), (dst_ip, 0))
+                        sent_this_batch += 1
+                    except Exception:
+                        pass
+                    seq += 1
+
+                with self._lock:
+                    stats = self._stats.get(flow_id)
+                    if stats:
+                        stats['tx_packets'] += sent_this_batch
+                        stats['tx_bytes'] += pkt_bytes_len * sent_this_batch
+
+                # Poll responder for rx stats periodically
+                if responder_url and (time.time() - last_responder_poll) >= 1.0:
+                    self._poll_responder(flow_id, responder_url,
+                                         responder_baseline_rx, responder_baseline_bytes)
+                    last_responder_poll = time.time()
+
+                # Precise rate limiting: sleep remaining time for this batch
+                batch_elapsed = time.time() - batch_start
+                sleep_time = interval - batch_elapsed
+                if sleep_time > 0:
+                    stop_event.wait(sleep_time)
+
+        except Exception as e:
+            log.error('Flow %s: send loop error: %s', flow_id[:8], e)
+        finally:
+            if raw_sock:
+                raw_sock.close()
+            with self._lock:
+                if flow_id in self._flows:
+                    self._flows[flow_id]['state'] = 'stopped'
+            # Final responder poll
+            if responder_url:
+                self._poll_responder(flow_id, responder_url,
+                                     responder_baseline_rx, responder_baseline_bytes)
+            log.info('Flow %s: send loop finished. seq=%d', flow_id[:8], seq)
+
+    def _send_loop_icmp(self, flow_id, flow, stop_event, rate_pps, duration):
+        """ICMP mode: use sr() to measure latency from router responses."""
+        pkt_template = build_packet(flow, seq=0)
+        pkt_bytes_len = len(bytes(pkt_template))
+        seq = 0
+        start_time = time.time()
 
         try:
             while not stop_event.is_set():
@@ -154,8 +255,6 @@ class FlowSender:
                 if duration and elapsed >= duration:
                     break
 
-                if use_icmp_sr:
-                    # ICMP mode: use sr() to measure latency from responses
                 pkt = build_packet(flow, seq=seq)
                 answered, _ = sr(pkt[pkt.firstlayer().payload.__class__],
                                  timeout=1, verbose=0)
@@ -169,75 +268,47 @@ class FlowSender:
                             stats['rx_packets'] += 1
                             stats['rx_bytes'] += len(bytes(recv_pkt))
                             stats['latency_samples'].append(rtt_ms)
-                                # Keep only last 1000 samples
                             if len(stats['latency_samples']) > 1000:
                                 stats['latency_samples'] = stats['latency_samples'][-1000:]
                 seq += 1
-                    # Rate limit for ICMP
                 sleep_time = (1.0 / rate_pps) - (time.time() - start_time - elapsed)
                 if sleep_time > 0:
                     stop_event.wait(sleep_time)
-                else:
-                    # UDP/TCP mode: send batches
-                    packets = []
-                    for _ in range(batch_size):
-                        pkt = build_packet(flow, seq=seq)
-                        packets.append(pkt)
-                        seq += 1
-
-                    try:
-                        if HAS_TCPREPLAY and rate_pps >= 1000:
-                            sendpfast(packets, pps=rate_pps, loop=0)
-                        else:
-                            for p in packets:
-                                send(p[p.firstlayer().payload.__class__], verbose=0)
         except Exception as e:
-                        # Fallback: basic send
+            log.error('Flow %s: ICMP send error: %s', flow_id[:8], e)
-                        log.debug('Send error (falling back): %s', e)
-                        for p in packets:
-                            try:
-                                send(p[p.firstlayer().payload.__class__], verbose=0)
-                            except Exception:
-                                pass
-
-                    with self._lock:
-                        stats = self._stats.get(flow_id)
-                        if stats:
-                            stats['tx_packets'] += len(packets)
-                            stats['tx_bytes'] += pkt_bytes_len * len(packets)
-
-                    # Poll responder for rx stats periodically
-                    if responder_url and (time.time() - last_responder_poll) >= 2.0:
-                        self._poll_responder(flow_id, responder_url)
-                        last_responder_poll = time.time()
-
-                    # Rate limit
-                    stop_event.wait(interval)
-
-        except Exception as e:
-            log.error('Flow %s: send loop error: %s', flow_id[:8], e)
         finally:
             with self._lock:
                 if flow_id in self._flows:
                     self._flows[flow_id]['state'] = 'stopped'
-            # Final responder poll
-            if responder_url:
-                self._poll_responder(flow_id, responder_url)
-            log.info('Flow %s: send loop finished. seq=%d', flow_id[:8], seq)
 
-    def _poll_responder(self, flow_id: str, responder_url: str):
+    def _fetch_responder(self, responder_url: str) -> dict:
-        """Poll a responder's /responder/stats endpoint for rx metrics."""
+        """Fetch raw stats from the responder."""
-        try:
         url = responder_url.rstrip('/') + '/responder/stats'
         req = urllib.request.Request(url, method='GET')
         req.add_header('Accept', 'application/json')
         with urllib.request.urlopen(req, timeout=2) as resp:
-                data = json.loads(resp.read().decode())
+            return json.loads(resp.read().decode())
+
+    def _reset_responder(self, responder_url: str):
+        """Reset responder counters."""
+        url = responder_url.rstrip('/') + '/responder/reset'
+        req = urllib.request.Request(url, method='POST')
+        req.add_header('Content-Type', 'application/json')
+        with urllib.request.urlopen(req, timeout=2) as resp:
+            resp.read()
+
+    def _poll_responder(self, flow_id: str, responder_url: str,
+                        baseline_rx: int = 0, baseline_bytes: int = 0):
+        """Poll a responder's /responder/stats endpoint for rx metrics."""
+        try:
+            data = self._fetch_responder(responder_url)
+            rx_pkts = data.get('rx_packets', 0) - baseline_rx
+            rx_bytes = data.get('rx_bytes', 0) - baseline_bytes
             with self._lock:
                 stats = self._stats.get(flow_id)
                 if stats:
-                    stats['rx_packets'] = data.get('rx_packets', 0)
+                    stats['rx_packets'] = max(0, rx_pkts)
-                    stats['rx_bytes'] = data.get('rx_bytes', 0)
+                    stats['rx_bytes'] = max(0, rx_bytes)
                     lat = data.get('latency', {})
                     if lat.get('avg_ms') is not None:
                         stats['latency_samples'].append(lat['avg_ms'])

traffic-gen/engine/stats.py

@@ -56,9 +56,14 @@ class StatsCollector:
             else:
                 tx_pps = rx_pps = tx_mbps = rx_mbps = 0.0
 
-            # Loss calculation
+            # Loss calculation: use rate-based when actively sending (avoids
+            # poll-lag artifacts), cumulative when flow has stopped
             loss_pct = 0.0
-            if tx_packets > 0 and rx_packets > 0:
+            if prev is not None and tx_pps > 0 and rx_pps > 0:
+                # Rate-based: compare instantaneous tx/rx rates
+                loss_pct = max(0.0, ((tx_pps - rx_pps) / tx_pps) * 100)
+            elif tx_packets > 0 and rx_packets > 0 and tx_pps == 0:
+                # Flow stopped: use cumulative counters (final accurate value)
                 loss_pct = max(0.0, ((tx_packets - rx_packets) / tx_packets) * 100)
 
             sample = {

traffic-gen/server.py

@@ -78,6 +78,7 @@ _tests_lock = threading.Lock()
 _responder = None     # Responder instance
 
 
+
 # ---------------------------------------------------------------------------
 # Helper
 # ---------------------------------------------------------------------------
@@ -86,6 +87,7 @@ def _now_iso():
     return time.strftime('%Y-%m-%dT%H:%M:%SZ', time.gmtime())
 
 
+
 def _flow_response(flow_id: str) -> dict:
     """Build a serializable flow dict."""
     with _flows_lock:
@@ -94,8 +96,15 @@ def _flow_response(flow_id: str) -> dict:
             return None
     result = dict(meta)
     if _sender:
-        result['is_running'] = _sender.is_running(flow_id)
+        running = _sender.is_running(flow_id)
+        result['is_running'] = running
         result['stats'] = _sender.get_stats(flow_id)
+        # Sync state: if sender thread finished but meta still says running
+        if not running and result.get('state') == 'running':
+            with _flows_lock:
+                if flow_id in _flows_meta:
+                    _flows_meta[flow_id]['state'] = 'stopped'
+            result['state'] = 'stopped'
     return result
 
 
@@ -148,6 +157,114 @@ def get_mode():
     return jsonify({'mode': MODE})
 
 
+@app.route('/mode', methods=['POST'])
+def set_mode():
+    global MODE, _sender, _stats_collector, _responder
+
+    data = request.get_json(force=True)
+    new_mode = data.get('mode', '').lower()
+    if new_mode not in ('sender', 'responder'):
+        return jsonify({'error': 'mode must be "sender" or "responder"'}), 400
+    if new_mode == MODE:
+        return jsonify({'mode': MODE, 'changed': False})
+
+    listen_iface = os.environ.get('TRAFFIC_GEN_INTERFACE', None)
+    responder_sub_mode = os.environ.get('TRAFFIC_GEN_RESPONDER_MODE', 'log')
+
+    # Tear down current mode
+    if MODE == 'sender':
+        # Stop all running flows
+        if _sender:
+            for fid in list(_sender.get_all_flows().keys()):
+                _sender.stop(fid)
+        with _flows_lock:
+            _flows_meta.clear()
+        with _tests_lock:
+            for t in _tests.values():
+                if t.state == 'running':
+                    t.stop()
+            _tests.clear()
+    elif MODE == 'responder':
+        if _responder:
+            _responder.stop()
+            _responder = None
+
+    # Start new mode
+    if new_mode == 'sender':
+        from engine.sender import FlowSender
+        from engine.stats import StatsCollector
+        _sender = FlowSender()
+        _stats_collector = StatsCollector()
+        _responder = None
+    elif new_mode == 'responder':
+        from engine.responder import Responder
+        _sender = None
+        _stats_collector = None
+        _responder = Responder(mode=responder_sub_mode)
+        _responder.start(interface=listen_iface)
+
+    MODE = new_mode
+    log.info('Mode switched to %s', MODE)
+    return jsonify({'mode': MODE, 'changed': True})
+
+
+# ---------------------------------------------------------------------------
+# Quick Ping (works in any mode)
+# ---------------------------------------------------------------------------
+
+@app.route('/ping', methods=['POST'])
+def quick_ping():
+    """Send ICMP pings to a target and return results."""
+    import subprocess
+
+    data = request.get_json(force=True)
+    target = data.get('target', '').strip()
+    count = min(int(data.get('count', 5)), 20)
+
+    if not target:
+        return jsonify({'error': 'target is required'}), 400
+
+    try:
+        result = subprocess.run(
+            ['ping', '-c', str(count), '-W', '2', target],
+            capture_output=True, text=True, timeout=count * 3 + 5
+        )
+        output = result.stdout + result.stderr
+
+        # Parse ping output
+        lines = output.strip().split('\n')
+        replies = []
+        for line in lines:
+            if 'time=' in line:
+                try:
+                    time_ms = float(line.split('time=')[1].split(' ')[0])
+                    replies.append(time_ms)
+                except (IndexError, ValueError):
+                    pass
+
+        stats = {}
+        if replies:
+            stats = {
+                'min_ms': round(min(replies), 2),
+                'avg_ms': round(sum(replies) / len(replies), 2),
+                'max_ms': round(max(replies), 2),
+            }
+
+        return jsonify({
+            'target': target,
+            'sent': count,
+            'received': len(replies),
+            'loss_pct': round(((count - len(replies)) / count) * 100, 1),
+            'replies': replies,
+            'stats': stats,
+            'reachable': len(replies) > 0,
+        })
+    except subprocess.TimeoutExpired:
+        return jsonify({'target': target, 'error': 'Ping timed out', 'reachable': False})
+    except Exception as e:
+        return jsonify({'target': target, 'error': str(e), 'reachable': False})
+
+
 # ---------------------------------------------------------------------------
 # Sender-mode: Flow endpoints
 # ---------------------------------------------------------------------------
@@ -193,7 +310,7 @@ def create_flow():
         'dscp': int(data.get('dscp', 0)),
         'vlan_id': data.get('vlan_id'),
         'state': 'idle',
-        'responder_url': data.get('responder_url'),
+        'responder_url': data.get('responder_url') or os.environ.get('RESPONDER_URL') or None,
         'created_at': _now_iso(),
     }
 
@@ -335,34 +452,51 @@ def create_test():
     from engine.rfc2544 import create_test as _create_test
 
     data = request.get_json(force=True)
-    flow_id = data.get('flow_id')
     test_type = data.get('type')
 
-    if not flow_id or not test_type:
+    if not test_type:
-        return jsonify({'error': 'flow_id and type are required'}), 400
+        return jsonify({'error': 'type is required'}), 400
 
+    # Accept either flow_config directly or flow_id to look up
+    flow_config = data.get('flow_config')
+    flow_id = data.get('flow_id')
+
+    if flow_config:
+        # Direct flow config provided — no flow_id needed
+        if not flow_config.get('dst_ip'):
+            return jsonify({'error': 'flow_config.dst_ip is required'}), 400
+        flow_config.setdefault('src_ip', 'auto')
+        flow_config.setdefault('protocol', 'udp')
+        flow_config.setdefault('src_port', 50000)
+        flow_config.setdefault('dst_port', 5001)
+    elif flow_id:
         with _flows_lock:
             flow_meta = _flows_meta.get(flow_id)
             if flow_meta is None:
                 return jsonify({'error': 'Flow not found'}), 404
+        flow_config = dict(flow_meta)
+    else:
+        return jsonify({'error': 'flow_config or flow_id is required'}), 400
 
     test_id = str(uuid.uuid4())
     kwargs = {
         'frame_sizes': data.get('frame_sizes', [64, 512, 1518]),
         'trial_duration': float(data.get('trial_duration', 60)),
-        'max_rate_pps': int(data.get('max_rate_pps', flow_meta.get('rate_pps', 10000))),
+        'max_rate_pps': int(data.get('max_rate_pps', flow_config.get('rate_pps', 10000))),
         'acceptable_loss_pct': float(data.get('acceptable_loss_pct', 0.0)),
+        'responder_url': data.get('responder_url') or os.environ.get('RESPONDER_URL') or None,
     }
 
     try:
-        test = _create_test(test_id, test_type, dict(flow_meta), **kwargs)
+        test = _create_test(test_id, test_type, flow_config, **kwargs)
     except ValueError as e:
         return jsonify({'error': str(e)}), 400
 
     with _tests_lock:
         _tests[test_id] = test
 
-    log.info('Created test %s (type=%s, flow=%s)', test_id[:8], test_type, flow_id[:8])
+    log.info('Created test %s (type=%s, dst=%s)', test_id[:8], test_type,
+             flow_config.get('dst_ip', '?'))
     return jsonify(test.get_info()), 201
 
 
@@ -483,7 +617,7 @@ def load_preset(name):
         'dscp': int(flow_data.get('dscp', 0)),
         'vlan_id': flow_data.get('vlan_id'),
         'state': 'idle',
-        'responder_url': flow_data.get('responder_url'),
+        'responder_url': flow_data.get('responder_url') or os.environ.get('RESPONDER_URL') or None,
         'created_at': _now_iso(),
     }