NAS setups

How to scan a NAS folder into your 3D print library

By Petros, who builds Meshory · Updated Aug 27, 2026 · 8 min read

Scanning a NAS folder into a 3D print library takes two moves: mount the share so your computer sees it as an ordinary folder, then add that mount as a source in a library app that reads it in place. No software runs on the NAS, nothing is copied off it, and nothing is written to it. This guide walks through both moves with Meshory, then covers the parts specific to network storage: how a network source stays in sync without a live file watcher, what happens when the NAS goes offline, and why duplicate detection deliberately runs at a gentler pace against a share.

One decision comes first: who the library is for. If what you want is a private library on storage you own, browsed from your computers, the desktop pattern below is the short path. If several people need accounts on one shared library, you're in self-hosted server territory instead, and do you need a Docker or NAS server? weighs that fork in detail. This guide assumes you've picked the no-server route.

Before you start: mount the share

A library app doesn't talk to the NAS directly; it reads the share your operating system has mounted. So the one prerequisite is that the NAS folder shows up in your file manager like any other folder. Every NAS exposes its folders over SMB (and usually NFS), and every desktop OS can mount them:

1. 1Windows: open File Explorer, right-click This PC, choose Map network drive, and enter the share path (\\nas-name\3d-models). Tick Reconnect at sign-in so the mount survives a reboot.
2. 2macOS: in Finder, press Cmd+K (Go → Connect to Server) and enter smb://nas-name/3d-models. Add the server to your login items if you want it remounted automatically.
3. 3Linux: mount the share with your file manager's Network entry, or add an fstab/systemd mount for something permanent.

Add the mount as a source

In Meshory, add the mounted folder as a source the same way you'd add a local one. Sources are watched in place: the app reads from where the files already live and never relocates them, so the share stays exactly as other tools and other computers expect to find it. You can add several shares, or a share next to local folders and external drives, and they all land in one searchable library.

One rule to know up front: a folder can only belong to one source. If you try to add a folder that's already inside a watched source, Meshory refuses and names the clash, because an overlapping source would index nothing and sit permanently empty.

What the first scan does (and what it deliberately skips)

The first scan indexes: it lists every supported file in the source, archive contents included, and that's all. It does not render the whole library and it does not hash it, because both are heavy reads that shouldn't happen to a terabyte of network storage unasked. The index appears almost immediately and you can browse, search and open the 3D viewer while the scan is still running. What Meshory indexes has the exact format list; the short version is models (STL, 3MF, OBJ, FBX, STEP), images, PDFs, and ZIP, RAR and 7Z archives read in place, with everything else left alone.

For scale: the reference library Meshory was hardened against is a 3.5 TB NAS share holding 118,270 models. Indexing it takes about a minute and a half, plus roughly seven minutes to read inside its 7,500 archives. If a scan finishes and something you expected is missing, the scan-issues list names every file it couldn't read and why.

Thumbnails without hammering the share

Previews are where a naive scanner hurts a NAS, because rendering needs the file's actual bytes. Meshory renders on demand instead of up front: a pool of background workers follows whatever is onscreen, so the share is read for the models you browse, at the pace you browse them. Each render lands in a local cache on your computer, kept separate from your models, so browsing the same folder tomorrow reads the cache, not the network.

Nothing in any of this writes to the NAS. Scanning, rendering and organizing only ever read; the only commands that touch your files are the ones you run yourself, renaming a model or moving one to the Trash. The share can stay read-only mounted if you prefer it that way, and the library still works.

Keeping a network source in sync

Local folders get a live file watcher, but live watching doesn't hold up well over SMB, so network sources work on a different rhythm: they're refreshed on a short cycle, and again whenever the app window regains focus. New files another computer dropped onto the share show up on the next refresh without you doing anything.

When the NAS goes offline (a reboot, a sleeping laptop leaving the network), Meshory backs off gracefully and picks the source back up automatically once the mount is reachable again. If the share changed a lot while the app wasn't running, trigger a rescan of that source: only the differences are processed, so it's far quicker than the first scan.

Duplicate detection at a NAS-friendly pace

Duplicate analysis is the heaviest read a library app can do, since content hashing means reading files end to end. Against a network source Meshory runs it at a deliberately gentler pace, and it narrows the work first: only files whose size matches another file's get read at all, which skips the vast majority of the share. The result feeds the same duplicates view as local sources, so copies that exist both on the NAS and on a local drive group together, which is exactly where years of "I'll just keep a local copy" get cleaned up.

With the share indexed, the rest of the system takes over: the folder structure, tagging and cleanup passes from the complete guide to organizing an STL library apply to a NAS library unchanged, because the app treats the mount like any other folder.