discobox

Syncs network device inventory from Netdisco into Netbox, enriching existing Netbox device records with live data discovered by Netdisco.

Devices are matched by management IP. Hostnames are verified and a warning is logged on mismatch, but the sync proceeds regardless.

Netbox is written to directly via pynetbox for simplicity. Migration to Netbox Diode is a future consideration once its entity coverage matures.

Architecture

flowchart LR
    DEV["Network Devices"]
    ND["Netdisco"]
    DB["discobox"]
    NB["Netbox"]
    VM["VictoriaMetrics"]

    DEV -- SNMP --> ND
    ND -- "webhook after discovery" --> DB
    DB -- "fetch device data" --> ND
    DB -- "write inventory" --> NB
    DB -. "reconcile: enqueue missing devices" .-> ND
    DB -. "reconcile: liveness check" .-> VM

Run Modes

discobox can be run in two modes:

Server mode (recommended)

A FastAPI server that receives webhook calls from Netdisco after each discovery and runs syncs in the background. Exposes Prometheus metrics and a Swagger UI.

# Docker
docker compose up -d discobox

# venv
python server.py

One-shot CLI

Sync a single device by IP directly from the command line. Useful for manual runs, testing, or scripting.

# Docker
docker compose run --rm cli --host 10.0.0.1

# venv
python cli.py --host 10.0.0.1

--host <IP>        Management IP of the device to sync (required)
--no-mac           Skip MAC address sync
--no-ip            Skip IP address sync (the management IP is always synced)
--no-modules       Skip module bay / module sync
--no-sfp           Skip SFP inventory item sync
--no-poe           Skip PoE mode sync
--housekeeping     Remove stale device bays and empty dummy interfaces
--debug            Enable debug logging

Features

Device fields — updates serial number, OS version, OS name, OS release (parsed from description)
Interfaces — creates and updates all physical and logical interfaces; maps interface type by name prefix; links subinterfaces (e.g. Gi0/0/1.100) to their parent interface
PoE — sets poe_mode = pse on ports reported as PoE-capable by Netdisco
MAC addresses — creates dcim.mac-addresses objects and links them as primary MAC per interface (Netbox 4.x model)
IP addresses — assigns interface IPs; fixes prefix mismatches (e.g. /32 → /26); moves IPs from dummy placeholder interfaces to the correct one
Module bays & modules — models physical chassis members as module bays with installed modules; assigns interfaces to their parent module
Blades — models linecards, supervisors, and fabric modules as module bays on the device they belong to (routed per VSS member for split chassis)
Device type auto-creation — creates manufacturer, device type (with part number and slug) if not present in Netbox
SFP / transceiver inventory — creates inventory items for transceivers with serial numbers, linked to their interface
PSU inventory — creates inventory items for power supplies
HA / VIP detection — detects cluster VIPs by hostname mismatch; redirects sync to the real active node; creates a Virtual Chassis linking both HA members; optionally deletes the VIP device (housekeeping)
Housekeeping — removes stale device bays auto-created from DeviceType templates (e.g. PS-A, Fan 1, Slot 1) and deletes empty dummy interfaces

Supported Topologies

Topology	Detection	Netbox model
Standalone	Single chassis, no stack root	Updates `device_type` directly on the device
Traditional stack (e.g. Cisco 3850)	`class=stack` root present	Module bay + module per stack member; interfaces linked to their member's module
Nexus FEX (e.g. N9K-C93180LC-EX)	Stack root type `cevContainerNexusLogicalFabric`	Primary N9K updates `device_type`; each FEX unit becomes a module bay + module; FEX interfaces linked to their FEX module
Cat9500 / Cat9600 StackWise Virtual	Stack root type contains `VirtualStack` or name contains `Virtual Stack`	Two separate Netbox devices linked via Virtual Chassis; each gets its own blades and device type; partner found by serial or hostname (`-2` suffix)
HA pair / VIP (e.g. Fortinet)	SNMP hostname differs from Netbox device found by IP, or hostname contains `p1h`/`p2h`, `node1`/`node2`, or `-1`/`-2`	Virtual Chassis created linking both physical nodes and the VIP device (p0h). Management VIP IP stays on the VIP device record; interface VIPs are assigned to whichever physical node syncs first. All shared IPs get `role=vip`. Behaviour of the VIP device is controlled by `DISCOBOX_VIP_MODE` (see Configuration).
Modular chassis (e.g. N9K-C9508, C9606R)	Standalone with `class=module` blade entries	Blades (linecards, supervisors, fabric modules) modelled as module bays

Netbox Mapping

Device

Netdisco field	Netbox field
`serial`	`serial`
`os_ver`	`custom_fields.os_version`
`os`	`custom_fields.os_name`
OS release (parsed from `description`)	`custom_fields.os_release`
Chassis `model`	`device_type.model`
Chassis `type` (vendor prefix)	`device_type.manufacturer`
Chassis count (traditional stack only)	`custom_fields.stack_members`

Interfaces

Netdisco field	Netbox field
`port` / `descr`	`name`
Interface name prefix	`type` (e.g. `10gbase-x-sfpp`, `lag`, `virtual`)
`speed`	`speed` (kbps)
`duplex`	`duplex`
`description`	`description`
`up`	`enabled`
`mac`	`dcim.mac-addresses` → `primary_mac_address`

IP Addresses

Netdisco field	Netbox field
`alias` / `ip` + `subnet`	`ipam.ip-addresses.address` (CIDR)
Interface port name	`assigned_object` (dcim.interface)

Module Bays & Modules

Netdisco field	Netbox field
Chassis `name`	Module bay `name`
Chassis `pos`	Module bay `position`
Chassis `model`	Module type `model` / `part_number`
Chassis `serial`	Module `serial`
Chassis `type` vendor prefix	Module type manufacturer

Blades (Linecards / Supervisors)

Netdisco field	Netbox field
Module `name`	Module bay `name`
Slot number (from `name`)	Module bay `position`
Module `model`	Module type `model` / `part_number`
Module `serial`	Module `serial`

SFP / Transceivers

Netdisco field	Netbox field
Module `name` (expanded)	Inventory item `name` + interface link
Module `model`	Inventory item `part_id`
Module `serial`	Inventory item `serial`

Configuration

Copy .env.example to .env and fill in the values:

NETDISCO_URL=https://netdisco.example.com
NETDISCO_TOKEN=your-long-lived-token    # preferred; takes priority over username/password (Bearer prefix added automatically)
NETDISCO_USERNAME=admin                 # fallback if NETDISCO_TOKEN is not set
NETDISCO_PASSWORD=secret
NETBOX_URL=https://netbox.example.com
NETBOX_TOKEN=your-token-here

# Server mode
DISCOBOX_LISTEN_PORT=8080   # default: 8080 (avoid DISCOBOX_PORT — Docker overrides it with a service URL)
DISCOBOX_WORKERS=4          # uvicorn worker count, default: 4. Metrics are plain in-memory (no
                            # cross-process aggregation): with >1 worker, /metrics only reflects
                            # whichever worker answers that particular scrape. Recommended: 1.
DISCOBOX_AUTH_TOKEN=secret  # bearer token for /sync; leave unset to disable auth
DISCOBOX_STATE_DIR=/tmp/discobox  # writable dir for pause flag, circuit breaker, reconcile
                                   # leadership/gap reports, in-flight sync claims; default: /tmp/discobox

# Sync concurrency
DISCOBOX_MAX_CONCURRENT_SYNCS=3   # max parallel syncs (semaphore), default: 3
DISCOBOX_MAX_QUEUE=10000          # max devices queued at once; excess hooks are dropped, default: 10000

# Sync feature defaults (server mode) — can be overridden per-request
DISCOBOX_NO_MAC=true        # disable MAC sync globally
DISCOBOX_NO_IP=true         # disable IP sync globally (e.g. if Netdisco is not VRF-aware)
DISCOBOX_NO_MODULES=true    # disable module bay sync globally
DISCOBOX_NO_SFP=true        # disable SFP inventory sync globally
DISCOBOX_NO_POE=true        # disable PoE sync globally
DISCOBOX_HOUSEKEEPING=true       # enable housekeeping globally
DISCOBOX_LLDP_CLEAR_STALE=true  # clear neighbor fields + delete stale discobox-owned cables when a port loses its neighbor

# HA / VIP device handling — controls what happens to the VIP placeholder device (e.g. p0h)
# threenode (default) — add VIP device as VC member at pos 0; management VIP IP stays on it (role=vip)
# soft               — unassign all IPs from VIP device so physical nodes can claim them; keep device
# hard               — delete VIP device entirely
# off                — do nothing
DISCOBOX_VIP_MODE=threenode

# Reconcile loop — daily job comparing Netbox against Netdisco; enqueues discovery for devices
# present in Netbox (SoT) but missing from Netdisco
DISCOBOX_RECONCILE_INTERVAL=86400  # seconds between runs, default: 86400 (24h)
DISCOBOX_RECONCILE_MAX_ENQUEUE=200 # max devices enqueued per run; unset = no limit
                                   # use this to throttle the initial backfill on large inventories
                                   # 0 = never enqueue, but still run the gap scan (report-only mode)
DISCOBOX_RECONCILE_MAX_QUEUED=500  # abort run if Netdisco queue has more than N jobs queued (default: 500)
DISCOBOX_RECONCILE_MAX_FAILED=500  # abort run if Netdisco reports more than N failures (default: 500)
# A few reconcile options are yaml-only (no env var): reconcile.roles, reconcile.statuses,
# reconcile.require_auth_tag — see discobox.yaml for defaults and comments.

# Liveness gate (optional) — before enqueueing, query a Prometheus-compatible API
# (VictoriaMetrics vmselect) and skip devices whose monitoring says they are down.
# Dead devices would only produce failed Netdisco jobs and burn the MAX_FAILED budget.
# Fail-open: devices absent from the query result (not yet monitored) are still
# enqueued, and a failed/unreachable query disables the gate for that run.
DISCOBOX_LIVENESS_URL=              # query API base incl. tenant prefix, e.g.
                                    # http://vmselect:8481/select/0/prometheus; unset = disabled
DISCOBOX_LIVENESS_QUERY='max_over_time(up{job="snmp"}[6h])'  # >0 alive, 0 down, absent unknown
DISCOBOX_LIVENESS_LABEL=netbox_primary_ip  # series label identifying the device
DISCOBOX_LIVENESS_KEY=ip            # device attribute compared to the label: ip | name
DISCOBOX_LIVENESS_TIMEOUT=15        # query timeout in seconds, default: 15
DISCOBOX_LIVENESS_TLS_VERIFY=true   # set to false for self-signed certs

# Overload protection — circuit breaker and per-device retry (see below)
DISCOBOX_CB_WINDOW=120      # look-back window in seconds for counting timeouts, default: 120
DISCOBOX_CB_THRESHOLD=3     # number of timeouts within window to trip the breaker, default: 3
DISCOBOX_CB_BACKOFF=120     # seconds to hold new syncs after breaker trips, default: 120
DISCOBOX_RETRY_DELAY=480    # seconds before first retry, default: 480 (8m); doubles each attempt
DISCOBOX_RETRY_MAX_DELAY=1800  # cap on backoff growth, default: 1800 (30m)
DISCOBOX_RETRY_MAX=6        # max per-device retries before dropping, default: 6

venv setup:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
set -a && source .env && set +a

Netbox Custom Fields

Written by discobox (Netdisco → Netbox)

These fields are updated on every sync. Create them on the Device object in Netbox.

Slug	Type	Description
`os_version`	Text	OS version string (e.g. `17.3.4`)
`os_name`	Text	OS platform name (e.g. `ios-xe`, `fortios`, `nx-os`)
`os_release`	Text	IOS release name parsed from device description (e.g. `Gibraltar`)
`stack_members`	Integer	Total physical units in a traditional stack with more than one member. Unset for standalone, a degraded 1-member stack (a count of `1` adds no signal), VSS (split across two Netbox devices), and FEX (satellites aren't stack members). By default (`custom_fields.stack_members_only_increase: true`, yaml-only) the recorded value never decreases — a dead member won't silently reduce it

Read by discobox (Netbox → Netdisco, reconcile loop)

These fields are optional. If set on a Device they are passed to the Netdisco discovery queue.

Slug	Type	Description
`snmp_auth_profile`	Select	SNMP credential tag in Netdisco (maps to `snmp_tag`)
`snmp_polling_timeout`	Select	Discovery timeout, e.g. `30s`, `3m` (default: `3s` if unset)

Written by discobox on Interfaces (Netdisco → Netbox)

CDP/LLDP neighbor data from Netdisco, written on every interface sync. Text fields — no dependency on the neighbor existing in Netbox.

Slug	Type	Object	Description
`lldp_neighbor`	Text	`dcim.interface`	Neighbor device hostname as reported by CDP/LLDP
`lldp_neighbor_port`	Text	`dcim.interface`	Neighbor port name as reported by CDP/LLDP

Server Endpoints

Endpoint	Method	Auth	Description
`/`	GET	no	Status page — endpoints, in-flight syncs, unknown devices
`/sync?host=<ip>`	GET, POST	yes	Trigger device sync
`/sync/all`	GET, POST	yes	Queue a sync for every device known to Netdisco; optional `?limit=N&force=true` (force bypasses the cooldown window). Any other query param is forwarded to the Netbox device filter, e.g. `?last_updated__lt=2026-10-01`, `?name__ic=foo`, `?site=zrh&role=switch` — selection is then intersected with Netdisco's device list
`/sync/pause`	GET, POST	yes	Pause queued syncs (also skips reconcile)
`/sync/resume`	GET, POST	yes	Resume queued syncs
`/reconcile`	GET, POST	yes	Trigger reconcile run immediately; optional `?max_enqueue=N&offset=N`
`/unknown-devices`	GET	no	Devices seen in Netdisco webhooks but not found in Netbox (JSON)
`/metrics`	GET	no	Prometheus metrics
`/health`	GET	no	Liveness check + in-flight hosts
`/docs`	GET	no	Swagger UI

All sync flags are available as query parameters or JSON body fields, defaulting to the env var values:

POST /sync?host=10.0.0.1&housekeeping=true&sync_mac=false
POST /sync  {"host": "10.0.0.1", "housekeeping": true, "sync_mac": false}

Syncs run in a background thread pool. Duplicate requests for the same host are dropped while a sync is in progress. Concurrent syncs log under discobox.<ip> so they are distinguishable in the log stream.

Overload Protection

discobox distinguishes between isolated device timeouts and global Netdisco overload, and handles each differently.

Errors covered: ReadTimeout (TCP read timeout) and HTTP 502/503/504 (gateway/proxy errors). Both indicate Netdisco is unable to serve the request.

Circuit breaker

When ≥ DISCOBOX_CB_THRESHOLD (default 3) timeout errors occur within DISCOBOX_CB_WINDOW seconds (default 120), the circuit breaker trips:

All new and queued syncs are held for DISCOBOX_CB_BACKOFF seconds (default 120)
After the backoff the breaker resets automatically — no manual intervention needed
The reconcile loop also skips while the breaker is active
Metric: discobox_circuit_breaker_trips_total

This is distinct from DISCOBOX_PAUSE_ON_ERROR / /sync/pause, which pause permanently until manually resumed.

Per-device retry

When a sync times out but the circuit breaker has not tripped (isolated failure), the device is automatically retried:

Retry is scheduled after DISCOBOX_RETRY_DELAY seconds (default 480 = 8m), doubling each attempt up to a cap of DISCOBOX_RETRY_MAX_DELAY (default 1800 = 30m)
Up to DISCOBOX_RETRY_MAX retries (default 6: 8m, 16m, 30m, 30m, 30m, 30m — then dropped)
If a retry fires while the circuit breaker is active, it waits until the breaker clears
If the queue is full when a retry becomes due, it is deferred 60s and re-attempted (the attempt is not consumed) rather than dropped
Metric: discobox_sync_retries_total

Netdisco Hook

Add to your Netdisco config.yml to trigger a sync automatically after each device discovery:

hooks:
  - event: discover
    action: HTTP
    url: "http://discobox:8080/sync?host=[% device.ip %]"
    method: POST
    headers:
      Authorization: "Bearer your-token"
    timeout: 30000

Traditional stack — Cisco C9300 (3-member)

INFO     discobox  ── device 10.10.1.1
INFO     discobox  Netdisco  hostname='SWCORE01-STK.corp.example.com'  ports=167
INFO     discobox  Netbox    device='swcore01-stk.corp.example.com'  id=443
INFO     discobox  Device fields updated — serial='FOC2404H5VQ' os_name='ios-xe' os_ver=None os_release='Amsterdam'
INFO     discobox    stack (root)  'c93xx Stack'  model=C9300-24UX  serial=FOC2404H5VQ
INFO     discobox    ├── chassis  'Switch 1'  model=C9300-24UX  serial=FOC2404H5VQ
INFO     discobox    ├── chassis  'Switch 2'  model=C9300-24T  serial=FCW2352M8JP
INFO     discobox    └── chassis  'Switch 3'  model=C9300-24T  serial=FCW2352R3TN
INFO     discobox  Modules   chassis=3  topology=stack
INFO     discobox    Switch 1  C9300-24UX  serial=FOC2404H5VQ  created
INFO     discobox    Switch 2  C9300-24T  serial=FCW2352M8JP  created
INFO     discobox    Switch 3  C9300-24T  serial=FCW2352R3TN  created
INFO     discobox  Modules — updated=3 unchanged=0 errors=0
INFO     discobox  PSUs      entries: 3
INFO     discobox    PSU Switch 2 - Power Supply A  model=PWR-C1-350WAC-P  serial=DCC2344W9KR  created
INFO     discobox    PSU Switch 3 - Power Supply A  model=PWR-C1-350WAC-P  serial=DCC2344B4QM  created
INFO     discobox    PSU Switch 1 - Power Supply A  model=PWR-C1-1100WAC-P  serial=DCC2344L7FN  created
INFO     discobox  PSUs — created=3 updated=0 unchanged=0 errors=0
INFO     discobox    GigabitEthernet2/0/4                     created
INFO     discobox    Vlan653                                  created
INFO     discobox    TenGigabitEthernet1/0/12                 updated
INFO     discobox    GigabitEthernet3/0/15                    created
...
INFO     discobox    TenGigabitEthernet1/0/3.10               created
WARNING  discobox    StackPort1/1                             in Netbox but not in Netdisco
WARNING  discobox    StackPort1/2                             in Netbox but not in Netdisco
INFO     discobox  Subinterfaces — linked 1 parent(s)
INFO     discobox  Interface→Module — updated=124 unchanged=0 skipped=44
WARNING  discobox    IP 192.168.63.11/24     already in Netbox (id=2688, assigned to unassigned) — skipping
INFO     discobox    IP 172.16.56.1/26       → created on Vlan720
...
INFO     discobox  IPs — created=22 fixed=1 moved=0 unchanged=0 skipped=12 errors=0
INFO     discobox  SFPs      entries: 12
INFO     discobox    SFP Te2/1/3              model=SFP-10G-SR-S         serial=ACW2402P6SM → created
INFO     discobox    SFP Te3/1/8              model=SFP-10G-SR           serial=AGA16034T3N → created
...
INFO     discobox  SFPs — created=12 updated=0 unchanged=0 errors=0
INFO     discobox  PoE — updated=0 unchanged=24 skipped=0 errors=0
INFO     discobox  ── done 10.10.1.1  created=140 updated=26 unchanged=0 errors=0
INFO     discobox.server  Sync success for 10.10.1.1 in 214.7s

StackWise Virtual — Cisco C9500 (2-member VSS)

INFO     discobox.server  hook from 10.10.0.5: 10.10.2.87  queued
INFO     discobox  ── device 10.10.2.87
INFO     discobox  Netdisco  hostname='SWDIST01-VSS1.corp.example.com'  ports=106
INFO     discobox  Netbox    device='swdist01-vss1.corp.example.com'  id=507
INFO     discobox  Device fields updated — serial='FDO25031N7Q' os_name='ios-xe' os_ver=None os_release='Amsterdam'
INFO     discobox    stack (root)  'Virtual Stack'  model=  serial=
INFO     discobox    ├── chassis  'Switch 2 Chassis'  model=C9500-24Y4C  serial=FDO25031B4R
INFO     discobox    └── chassis  'Switch 1 Chassis'  model=C9500-24Y4C  serial=FDO25031N7Q
INFO     discobox  Modules   chassis=2  topology=vss
INFO     discobox    DeviceType → Cisco / C9500-24Y4C  serial=FDO25031N7Q  updated
INFO     discobox    VSS partner found by hostname 'swdist01-vss2.corp.example.com'
INFO     discobox    VSS partner DeviceType → C9500-24Y4C  serial=FDO25031B4R  updated
INFO     discobox    VirtualChassis 'SWDIST01-VSS1' — created
INFO     discobox  Modules — updated=2 unchanged=0 errors=0
...
INFO     discobox  ── done 10.10.2.87  created=0 updated=0 unchanged=105 errors=0
INFO     discobox.server  Sync success for 10.10.2.87 in 20.3s

Modular chassis — Cisco Nexus 9508 (blades + PSUs)

INFO     discobox.server  hook from 10.10.0.6: 10.10.3.21  queued
INFO     discobox  ── device 10.10.3.21
INFO     discobox  Netdisco  hostname='NXCORE01-DC1'  ports=570
INFO     discobox  Netbox    device='nxcore01-dc1.corp.example.com'  id=534
WARNING  discobox  Hostname mismatch for 10.10.3.21 — Netdisco='NXCORE01-DC1'  Netbox='nxcore01-dc1.corp.example.com'
INFO     discobox  Device fields updated — serial='FGE22124W6T' os_name='nx-os' os_ver=None os_release=None
WARNING  discobox    Dummy interface 'mgmt0' still has 1 IP(s) — skipping deletion
INFO     discobox  Housekeeping — deleted 0 stale device bay(s), 0 empty dummy interface(s)
INFO     discobox    chassis (root)  'Nexus9000 C9508 (8 Slot) Chassis'  type=cevChassisN9Kc9508  model=N9K-C9508  serial=FGE22124W6T
INFO     discobox    └── chassis  'Nexus9000 C9508 (8 Slot) Chassis'  model=N9K-C9508  serial=FGE22124W6T
INFO     discobox  Modules   chassis=1  topology=standalone
INFO     discobox    DeviceType → Cisco / N9K-C9508  serial=FGE22124W6T  updated
INFO     discobox  Modules — updated=1 unchanged=0 errors=0
INFO     discobox  PSUs      entries: 6
INFO     discobox    PSU PSU 1  model=N9K-PAC-3000W-B  serial=ART2216B3QN  created
INFO     discobox    PSU PSU 2  model=N9K-PAC-3000W-B  serial=ART2216C8RM  created
INFO     discobox    PSU PSU 3  model=N9K-PAC-3000W-B  serial=ART2216D5WP  created
INFO     discobox    PSU PSU 4  model=N9K-PAC-3000W-B  serial=ART2216J4TS  created
INFO     discobox    PSU PSU 5  model=N9K-PAC-3000W-B  serial=ART2213K7VP  created
INFO     discobox    PSU PSU 6  model=N9K-PAC-3000W-B  serial=DTM214009XM  created
INFO     discobox  PSUs — created=6 updated=0 unchanged=0 errors=0
INFO     discobox  Blades    entries: 12
INFO     discobox    Blade Supervisor Module 1       model=N9K-SUP-B+        serial=FOC22203T8L  created
INFO     discobox    Blade Supervisor Module 2       model=N9K-SUP-B+        serial=FOC222030S4  created
INFO     discobox    Blade 32p 40/100G Ethernet Module 1  model=N9K-X9732C-EX  serial=FOC22191Q4T  created
INFO     discobox    Blade 32p 40/100G Ethernet Module 2  model=N9K-X9732C-EX  serial=FOC21186S7K  created
INFO     discobox    Blade 32p 40/100G Ethernet Module 3  model=N9K-X9732C-EX  serial=FOC22191P5N  created
INFO     discobox    Blade 32p 40/100G Ethernet Module 4  model=N9K-X9732C-EX  serial=FOC27028M3L  created
INFO     discobox    Blade Fabric Module 2  model=N9K-C9508-FM-E  serial=FOC22191R2M  created
INFO     discobox    Blade Fabric Module 3  model=N9K-C9508-FM-E  serial=FOC21410B9R  created
INFO     discobox    Blade Fabric Module 4  model=N9K-C9508-FM-E  serial=FOC21410C4P  created
INFO     discobox    Blade Fabric Module 6  model=N9K-C9508-FM-E  serial=FOC21410D8N  created
INFO     discobox    Blade System Controller 1  model=N9K-SC-A  serial=FOC22155G3Q  created
INFO     discobox    Blade System Controller 2  model=N9K-SC-A  serial=FOC22155H7X  created
INFO     discobox  Blades — created=12 updated=0 unchanged=0 errors=0
INFO     discobox    Ethernet4/20                             created
INFO     discobox    Tunnel26                                 created
INFO     discobox    Tunnel211                                created
...

HA pair / VIP — Fortinet (threenode mode)

INFO     discobox  Netbox    device='fwcluster01p1h.corp.example.com'  id=24761
INFO     discobox.10.10.4.98  HA VIP device 'fwcluster01p0h.corp.example.com' added as VC member pos=0
INFO     discobox    VC member fwcluster01p1h.corp.example.com          pos=1  unchanged
INFO     discobox    VC member fwcluster01p2h.corp.example.com          pos=2  unchanged
INFO     discobox    VC member fwcluster01p0h.corp.example.com          pos=0  unchanged
INFO     discobox.10.10.4.98  HA peer VirtualChassis 'fwcluster01' — unchanged  (topology=standalone per Netdisco; VC is Netbox-side)
INFO     discobox.10.10.4.98  VIP device 'fwcluster01p0h.corp.example.com' — device_type updated
INFO     discobox.10.10.4.98  VIP device 'fwcluster01p0h.corp.example.com' — primary IP 10.10.4.97/24 role set to vip
INFO     discobox  Device fields updated — serial='FG6H0FTB24913847' os_name='fortios' os_ver=None os_release=None

Name		Name	Last commit message	Last commit date
Latest commit History 135 Commits
.github/workflows		.github/workflows
tests		tests
.env.example		.env.example
.gitignore		.gitignore
Dockerfile		Dockerfile
Pipfile		Pipfile
Pipfile.lock		Pipfile.lock
README.md		README.md
banner.png		banner.png
cli.py		cli.py
discobox.py		discobox.py
discobox.yaml		discobox.yaml
docker-compose.yml		docker-compose.yml
grafana-dashboard.json		grafana-dashboard.json
pyproject.toml		pyproject.toml
requirements.txt		requirements.txt
server.py		server.py

Navigation Menu

Uh oh!

Repository files navigation

discobox

Architecture

Run Modes

Server mode (recommended)

One-shot CLI

Features

Supported Topologies

Netbox Mapping

Device

Interfaces

IP Addresses

Module Bays & Modules

Blades (Linecards / Supervisors)

SFP / Transceivers

Configuration

Netbox Custom Fields

Written by discobox (Netdisco → Netbox)

Read by discobox (Netbox → Netdisco, reconcile loop)

Written by discobox on Interfaces (Netdisco → Netbox)

Server Endpoints

Overload Protection

Circuit breaker

Per-device retry

Netdisco Hook

Sample Log Output

About

Releases

Packages

Contributors

Languages

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

discobox

Architecture

Run Modes

Server mode (recommended)

One-shot CLI

Features

Supported Topologies

Netbox Mapping

Device

Interfaces

IP Addresses

Module Bays & Modules

Blades (Linecards / Supervisors)

SFP / Transceivers

Configuration

Netbox Custom Fields

Written by discobox (Netdisco → Netbox)

Read by discobox (Netbox → Netdisco, reconcile loop)

Written by discobox on Interfaces (Netdisco → Netbox)

Server Endpoints

Overload Protection

Circuit breaker

Per-device retry

Netdisco Hook

Sample Log Output

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages