Skip to content

Physical map

The physical map shows where a device stands — on a floor plan, rack layout, or photo. Unlike the logical map, it draws no dependencies: everything here is pure placement. Connections you draw here become visual arrows, never real dependencies with alerting effect.

A physical map is attached to exactly one site — that's how you find it again via the site pin on the world map. You set the link when creating the map or later via the toolbar.

Background image(s)

Add an image uploads PNG, JPG, or WebP (no SVG, 10 MB size limit). Vesana re-encodes the image server-side and corrects orientation from its EXIF data. You can place several images on one map — each as its own freely movable and resizable node (drag handles at the corners).

  • Images always sit below all other nodes. No matter how you click or select — host tiles and other elements always stay visible and clickable on top.
  • Several stacked images can be ordered via right-click "bring forward/send backward".
  • Lock image (right-click) fixes position and size — the floor plan then stops accidentally moving along when you place hosts next to it. A locked image also can't be deleted by accident with Delete.
  • A removed image is only really deleted from the map on save — if you cancel afterwards (discard), the file is kept.
  • Older maps with the earlier single background image keep working; the toolbar also lets you remove that legacy background.

Positions on physical maps are stored relative to the image (normalized 0–1) — changing the image therefore doesn't move placed elements.

Host tiles

Hosts appear as tiles of fixed shape. On the physical map they are resizable (drag handles on the edge) — resizing scales the whole tile proportionally, content stays legible. On the logical map, size is fixed.

Arrows instead of dependencies

Every connect drag on the physical map — even between two hosts — creates an arrow (solid line with an arrowhead), never a dependency. This lets you draw, say, network cabling or a signal path without accidentally triggering inhibition logic.

  • A flat drag produces a straight arrow, a diagonal drag a diagonal one.
  • Right-click an arrow: "arrowhead to other end" (reverse direction) or "mirror diagonal".
  • The toolbar also lets you place free-standing arrows that aren't attached to any node — freely placeable, resizable, and deletable as a single object.
  • Arrows and decorative edges can be labeled directly by clicking, or deleted by right-click.

Since all of this consists of decorative edges (see Maps & sites), they never trigger an alert, suppression, or status rollup. If you want a real dependency with alerting effect, that belongs on the logical map — trying to link two real objects (e.g. two hosts) as a dependency on the physical map is rejected, with a hint that it belongs on the logical map.

What's deliberately missing here

  • No dependency edges — only the logical map knows inhibition.
  • No aggregate node and no dependencies panel — both are logical constructs.

This avoids "two truths": dependencies live exclusively on the logical map.

Versions & trash

As with the logical map, every position or structure change creates a version on save; deleted maps land in the trash for 30 days (including uploaded image files).

Next

  • Maps & sites — overview, world map, sites
  • Logical map — dependencies with real alerting effect
  • Aggregates — K-of-N redundancy (appears only on the logical map)