The first cut put the drive bays in a squat block across the top half and left
the name across the middle, which looked nothing like a UGREEN DXP or a
Synology DS: their front is two to five tall tray doors filling the height, over
a thin strip carrying the badge, the LED and the sockets.
The name band was the blocker — it was nailed to mid-height for every plate, so
anything tall wore its name across its own artwork. `labelBox.y` (default 0.5,
so rack gear is untouched) now places the band, and the status LED rides it.
The NAS plates put it at 0.86 with the trays above and the ports on the same
strip.
Bay corners also scaled badly: a flat 1.5px radius reads as a square hole once
the cells are as large as a 3U tray, so the radius now follows the cell size.
ha-relevant: yes
A UGREEN DXP or a Synology DS is not rack gear: it sits on a shelf, taking
about a third of the width and standing some 3U tall, with its drive trays side
by side rather than in a grid. The catalog had no shape for that — a homelab
NAS had to borrow the 2U rack-mount plate.
Trays fill the upper half, one column per disk. The name band and the ports
share the lower half side by side, since the name is always drawn at mid-height
whatever the U count. Ports follow the class: one RJ45 on the 2-bay, two on the
4-bay, two plus an SFP+ on the 5-bay.
`suggestFaceplate` still proposes the rack-mount `nas-2u` for a discovered NAS;
which of the two a scan should assume is a separate question.
ha-relevant: yes
Four defects found reviewing the last two commits.
Reverting a faceplate choice took every cable on the device down with it.
Browsing to another plate and back left `plateChanged` set with freshly seeded
port ids, while `commitEdit` skipped `applyFaceplate` because the id matched —
so `setPorts` received ids the device never had and dropped every cable
attached to the old ones, with no warning, since from the user's side the plate
had not changed. Coming back to the device's own plate now restores its ports,
ids included.
The inventory select preselects the first unracked entry, but only
`pickInventory` seeded the status from it, so submitting without touching the
select wrote `unknown` over the status the mount had just copied. Status now
seeds from that same entry.
Creating a device POSTs to the Device Inventory before the mount is attempted,
so a rack with no free slot stranded the row — and stranded another on every
retry. The rack is checked for a slot before anything is created.
And `_service_name` accepted a service with no port, where the inventory
requires one (`None not in _COMMON_PORTS` is True): a portless service named a
device in the rack picker but not in the Device Inventory, the exact mismatch
the naming fix set out to remove.
ha-relevant: maybe
The rack picker listed "192.168.1.63 · 192.168.1.63" where the inventory
showed "jellyfin", so a device the user picked out by one name turned up here
under another.
Two causes. The inventory route labelled an entry
`friendly_name or hostname or ip or id`, skipping the app-name step: `_device_label`
now mirrors `deviceLabel` in PendingDevicesModal — friendly name, host, the app
its ports say it runs (ssh/http/https ignored, generic web deprioritized), IP,
IEEE, id. And the option line appended the IP unconditionally, doubling it
whenever the label already was the IP; it now adds the IP only when it says
something new, plus the type as the tiebreaker between look-alikes.
Also fixes the port row: the name field and the type select both carried
`w-full` from the shared input class, so the select took the row and the name
field collapsed into an unlabelled box.
ha-relevant: maybe
A plate is a drawing, not a name — "Switch 24 ports + 2 SFP" in a <select>
tells you far less than seeing the plate. The Faceplate field now previews the
current plate and opens FaceplatePicker: every template drawn with the canvas
renderer, grouped by category, searchable by name or category, and filtered by
kind so accessories are offered only to accessories. Switching the device
source between device and accessory swaps the plate to a default of the other
kind.
Each tile scales the plate to the tile width and keeps its fraction of the rack
(a half-width plate draws half as wide), so sizes compare honestly; the meta
line states U height, width and port count.
Two layout notes worth keeping: DialogContent carries `sm:max-w-sm`, which is
media-scoped and beats a plain `max-w-*`, so the override has to be prefixed;
and an SVG is a flex item that shrinks but never grows, so the plate is drawn
at a nominal size and scaled up by CSS instead of sized in pixels.
ha-relevant: no
The rack canvas loses its right rail: a mount is now edited in its own dialog,
like a network node, and the canvas keeps the full width.
Devices:
- `RackDeviceModal` adds and edits a mount — label, faceplate, U position,
height, column, width, status, colour, ports, unmount. Adding offers three
sources: an existing Device Inventory entry, a new device created here, or a
rack-only accessory. Opened by `+ Device` in the left rail or a double-click
on a plate.
- `RackSettingsModal` takes over rack chrome (name, location, capacity, 19"/10",
numbering, colours, delete), on a double-click of the chassis.
- The left rail's `+ Rack` becomes `+ Device`; `+ Rack` stays in the header.
Inventory:
- Gear created from a rack canvas lands in the Device Inventory with
`discovery_source: "rack"`, so it inherits the existing search, filters, hide
and delete instead of a parallel list in the sidebar. New "Rack devices"
source filter and badge.
- It is never placed on a logical canvas: approve returns 409, bulk-approve
skips it with `match: "rack"`, and the UI drops the Approve button and reports
the skip.
- `InventoryTray` becomes `AccessoryTray` — blanks, shelves and cable managers
only, still a drag source.
Fixes:
- A taller faceplate (or a hand-typed height) no longer silently keeps the old
geometry when it collides: the device relocates to the nearest slot that takes
the new size, and only a rack with no such slot refuses the edit. This is why
a 1U and a 2U plate rendered at the same height and the field looked locked.
- The empty-state buttons sat under `.react-flow__renderer` (z-index 4), so the
pane swallowed their clicks as a canvas drag.
Tests: rack device/settings modals, empty-state stacking, relocate-vs-refuse in
the store, rack source bucketing, inventory approve gating (frontend + backend).
ha-relevant: maybe
The modal-based editors and the rack store are portable; the inventory tagging
rides on the standalone FastAPI scan routes. Discuss before porting.
The rack prototype lived behind /racklab with no backend. It now becomes a
third design_type alongside network and electrical, sharing the app shell:
same header, same left rail, same right panel, same design switcher. App
branches the renderer on design_type.
Backend
- racks / rack_devices / rack_cables, all design_id-scoped.
- rack_devices.device_id -> pending_devices is the primary link. Those rows
survive approval and node deletion, so unracking never removes the device
from the Device Inventory. node_id -> nodes is a second, optional link used
only for live status and for matching network-link imports. Both are
ON DELETE SET NULL and label is denormalized, so a rack keeps rendering
after an inventory purge.
- GET /racks, POST /racks/save (full-state upsert + prune, same contract as
canvas/save), GET /racks/inventory.
- POST /scan/pending adds an inventory entry by hand, for hardware no scan
can discover.
- design_type is now validated and actually branched on; the schema comment
calling it vestigial is gone.
- Design delete unwinds rack rows explicitly (SQLite does not always honour
the cascade); design copy duplicates racks, mounts and patches with fresh
ids and re-points the cables at the copies.
Frontend
- /racklab and RackLab are removed; no dual maintenance.
- Rack types lifted into @/types; rackSerializer maps the snake_case wire
shapes to them, narrowing every enum on the way in.
- The store gains load/save, dirty tracking and an editSeq, so racks get the
same explicit Save and the same opt-in autosave as the logical canvas.
- Header, sidebar and inspector adapt in rack mode: canvas-only actions are
hidden, Add Rack / Patch / cable visibility / type filter / Import links
take their place, the sidebar body becomes the inventory tray, and the
footer counts racks, mounts, cables and free U.
- rackTheme derives the rack palette from the active app theme rather than
declaring one per theme, so a new theme works here for free. Per-rack
chrome stays user-editable; only its default comes from the theme.
- New Canvas gains a Kind picker (hidden when copying, since a copy inherits
the source kind).
Tests: 24 backend, 5 frontend files (store, persistence, networkLinks,
rackTheme, rackSerializer) plus DesignModal coverage for the Kind picker.
ha-relevant: maybe
Frontend rack renderer is portable; the FastAPI routes and SQLAlchemy models
are standalone-only by the persistence/transport split. Discuss before porting.
Physical counterpart to the logical canvas: racks hold hardware at a U
position, with drawn front panels and port-to-port patching.
- 12-column grid so 2 or 3 machines can share one U; collision and
snap-to-free-slot handled in layout.ts
- 19 faceplate templates drawn as SVG in unit coordinates, so a plate
scales with U height and rack width
- RJ45 / SFP port artwork only; ports stay hidden on non-patch gear
until hover, selection or patch mode. Outlets are artwork, never cabled
- cables are their own relation, not React Flow edges, rendered in a
ViewportPortal so they cross racks and follow pan/zoom
- unmounting a device never removes its inventory entry
- rack settings: size, 19"/10", numbering direction, colours, enclosure
Isolated behind its own entry point (like /view) — no API, no
persistence, no design_type change until the UX is settled.
ha-relevant: yes
Lockfile only — every advisory is transitive and no direct dependency range
had to move:
- postcss 8.5.14 -> 8.5.25 (high, GHSA-r28c-9q8g-f849, path traversal in
source-map auto-loading), pulled in by vite and shadcn
- fast-uri 3.1.2 -> 3.1.5 (two high advisories), via shadcn's ajv
- hono 4.12.25 -> 4.12.33 and @hono/node-server 1.19.14 -> 2.0.12 (three
moderate advisories), via shadcn's @modelcontextprotocol/sdk, itself bumped
1.29.0 -> 1.30.0
`npm audit` now reports 0 vulnerabilities. Build, lint and the full frontend
suite (1542 tests) pass unchanged.
ha-relevant: no
Add a label substring filter to GET /api/v1/nodes, plus MCP tools
list_node_summaries (id/label/type/status only) and get_node
(by id -> single object, by label -> array) so an LLM client can look
up nodes without paying the token cost of the full node payload.
The node type fix left the same stale-enum bug in place for edges:
create_edge only advertised ethernet/wifi/iot/vlan/virtual, while the
frontend EdgeType union also defines cluster, fibre and electrical.
The backend's EdgeBase.type is a plain str, so those three were always
accepted server-side — only the MCP schema rejected them.
- Extract EDGE_TYPES in mcp/app/tools.py and use it for create_edge.
- Add mcp/tests/test_edge_types_sync.py, mirroring the NodeType guard:
parses EdgeType out of frontend/src/types/index.ts and asserts set
equality with EDGE_TYPES.
- Add create_edge cases for the three previously-rejected types plus a
schema test pinning the enum contents and the ethernet default.
- Add an `mcp` job to quality.yml. Both sync guards were unreachable
from CI before: quality.yml only ran frontend and backend, so the
enums could drift again with nothing to catch it. The job runs from a
full checkout because the guards read the frontend sources, which are
not in the mcp Docker build context.
- Pin mcp[cli]>=1.26,<2. The bound is load-bearing: SDK 2.0 dropped the
low-level Server.list_resources/list_tools decorator API that
app/resources.py and app/tools.py are built on, so an unbounded spec
breaks both the tests and the Docker image at import time.
- Move pytest/pytest-anyio/trio into mcp/requirements.txt, matching how
backend/requirements.txt carries its own test deps. trio is required
because pytest-anyio parametrises every async test over both backends.
Tested: 114 passed in mcp/, both in the existing venv and in a clean
Python 3.13 venv installed the way the new CI job does.
ha-relevant: no
Services can now carry a brand logo (homarr dashboard-icons) or a generic
lucide icon, the same way nodes already could — closes the request in
discussion #299 where services stand in for docker images.
The add/edit form moves out of the right-hand DetailPanel into a proper
ServiceModal, which gives the icon picker room to breathe and matches how
nodes are edited.
- ServiceModal: name, port, protocol, path, icon; validates the name and
clamps the port, and spreads over the original service on save so fields
it doesn't own (category) survive an edit
- IconPickerPanel: Generic/Brand picker extracted from NodeModal so both
surfaces share one implementation
- ServiceIcon: renders a lucide or brand:<slug> icon in the DetailPanel
badge and on the canvas node, and nothing when the key can't be resolved
- Legacy scanner signature keys (bar-chart-2) still don't resolve to an
icon, but the modal shows them verbatim and preserves them on save
ServiceInfo.icon already existed and the backend stores services as
list[Any], so no schema or migration change.
ha-relevant: yes
The hardcoded NODE_TYPES list in mcp/app/tools.py only had 11 entries,
causing create_node/update_node/approve_device to reject valid device
types (firewall, docker_container, ups, etc.) that the backend and
frontend already support. Expanded the list to match
frontend/src/types/index.ts's NodeType union, excluding group/groupRect/text
which are canvas annotations, not devices. Added a test that parses the
frontend union and fails if the two lists drift apart again.
The login page rendered lucide's generic Network glyph instead of the
brand mark. Use the Logo component there, and realign Logo with the
canonical assets in docs/logo:
- match icon.svg geometry exactly (bg glow gradient, node glow filter)
- fix inverted wordmark colors (Home #e6edf3 / lable #00d4ff)
- scope gradient/filter ids per instance so the Sidebar and Toolbar
logos don't share defs ids
ha-relevant: maybe
The dependency-audit job failed on vulnerabilities from `shadcn` and its
transitive tree (hono, fast-uri, body-parser, @modelcontextprotocol/sdk).
`shadcn` is a build-time scaffolding CLI plus a CSS `@import`, never shipped
in the browser bundle, so it belongs in devDependencies.
- move `shadcn` to devDependencies (regenerated lockfile: dev flags only,
no version changes)
- run `npm audit` with `--omit=dev`, matching the runtime-only intent of the
pip-audit step
npm audit --omit=dev --audit-level=high now reports 0 vulnerabilities; the
production build still resolves the shadcn CSS import.
ha-relevant: no
The Proxmox/Zigbee auto-sync Save tests clicked Save after only
waiting for an unrelated field, so under CI load getConfig could
resolve after the click and Save fired with the default
sync_enabled=false. Wait for the toggle to reflect the loaded
state first.
YAML export wrote per-side handle counts but never the
show_port_numbers toggle, so the option was lost on re-import
while connection points survived.
ha-relevant: maybe
Fixes#272
Setting --error-bg alone had no effect: sonner only honors per-type
surfaces when richColors is enabled, so error toasts kept using
--normal-bg and looked identical to a normal toast. Enable richColors
and pin success/info/warning back to the neutral popover surface so only
errors turn red.
Error toasts previously used the same neutral popover background as
success toasts, so a failure (e.g. a save while the backend is down)
read like a normal 'Canvas saved'. Set sonner's per-type --error-bg/
--error-text/--error-border so errors stand out in red.
Neutralize the app modal backdrop's dim/blur while the tour runs so a
ringed sidebar/toolbar control behind an open modal (steps 2, 10) stays
readable — the overlay's SVG spotlight is now the single source of
dimming. Add a walkthrough-running body class toggled by the overlay and
a CSS rule stripping [data-slot=dialog-overlay] background/backdrop-filter.
Also refine coach-card placement: dialogCardPosition drops the card into
the empty bottom-right corner for large modals (inventory, scan history)
and beside small modals, so it never covers modal content (steps 3, 4).
Bespoke spotlight walkthrough (no new deps) that guides new users through the
main features. A version-stamped invite bubble offers the tour to new users and
re-offers it once to existing users after an upgrade; restartable from Settings.
- Multi-hole SVG-mask overlay keeps a ringed control and any open modal lit
- Steps: scan, scan history, inventory, nodes, edit, text/zone, grouping,
style, imports, and a closing recap with a GitHub link
- Scan/history/inventory/import steps inject demo data (no backend calls) and
are filtered out in standalone, which ends on a full-mode recap
- Persisted invite state via localStorage (utils/walkthrough.ts), live tour
state via zustand (stores/walkthroughStore.ts)
ha-relevant: maybe
Cleared canvas re-showed the demo, and backend errors silently fell back to
demo — hiding real outages and forcing users to wipe demo nodes before bulk
edits. Now:
- backend load reports `initialized` (CanvasState row exists = ever saved)
- decideCanvasLoad() picks real | empty | demo; empty is kept empty
- backend down/error shows an error banner + toast, never the demo
- data-new-user flag reserved as the Getting Started walkthrough hook
ha-relevant: yes
Add native HTML5 drag-and-drop reorder to the property list in the
detail panel. A grip handle appears when more than one property exists;
dragging updates property order (persisted with canvas save) and is
undoable via history snapshot.
Closes#292
ha-relevant: yes
- add tests for notes, property key/value, hidden-property exclusion, matched-value display
- fix flatMap indentation and stray semicolon
- use stable match key instead of display_value (avoids dup-key collisions)
- keep label searchable but not repeated in matched-values column
nmap -sV sends plaintext probes to TLS ports (e.g. Proxmox 8006). They
stall until the host crosses --host-timeout, at which point nmap skips
the whole host and discards ports it already found — a reachable host is
returned with 0 open ports (issue #277).
Split Phase 2 into two passes: Pass A (-sS/-sT --open) discovers ports
fast with no version detection; its ports are authoritative. Pass B
(-sV --host-timeout, scoped to the found ports) enriches banners
best-effort — on timeout/failure the discovered ports are kept with
empty banners instead of losing the host.
Healthy hosts still get full-intensity -sV banners (no quality drop).
Version host-timeout is now configurable via SCANNER_VERSION_HOST_TIMEOUT
(default 60s).
ha-relevant: yes
Pouzor