NAS > local indexes and server-side databases

Dear forum members,

I am using a Synology NAS to manage my data securely. Before using DEVONthink, I moved all local files of my laptop with Hazel to the NAS server.

In the future, I want to use DEVONthink to manage all my data. To save space on my local computer I want to store all databases on the NAS and only the indexes locally. I don’t want to synchronize the NAS server directly with the laptop, because the same data woold be stored on both sides.

Is there a secure way to store the indexes locally and manage all DEVONthink databases centrally on the NAS?

I look forward to your advice. Greetings Phil

Enclosed further information to my setup:
1.) I create weekly backups of my NAS server.
2.) I am the only user of the DEVONthink databases and will not get any problems with direct database access without synchronization.

Welcome @Philcomputing

No. The databases contain the indices. They are inseparable.
Also, don’t think the indices would necessarily be small, just because you’re indexing data. This is a common misconception about indexing. The metadata contains the full text of text-based files. That text takes up space.


thank you so much for your answer. I have found the following German explanation [1]:

Da man ja lokal den Platz sparen möchte, kann man folgendes tun:
- Datenbank anlegen und einen Ordner auf dem NAS als Verknüpfung (Index) in die Datenbank legen (cmd+ctrl+DragDrop). Alle Dokumente/Unterordner in diesen Ordner packen
- Immer dran denken, über das Kontextmenü dieses Index-Ordners zu sagen “Dokumente in indexierten Ordner verschieben” (oder so ähnlich) - dann landen die Dokumente wirklich als reine verknüpfte Dateien auf dem NAS; die Metadaten bleiben auf dem lokalen Rechner
- Die Datenbank über das NAS mit anderen Clients syncen

So haben alle Clients die Metadaten lokal (=schneller Start; können aber auch ein paar hundert MB werden) und die Dokumente liegen auf dem NAS.

Is that a workaround that can save space on the local machine?

Grüße Phil

What it’s referring to - called indexing - is actually what I was describing, so my comments still relate. Indexing merely points to data outside a database, but again, the index is internal to the database and will still consume space with the full text. How much space is consumed depends on the amount of text.

Also, indexing should be thoughtfully considered.

  1. It is inadvisable to do indiscrimate data dumps into a database, even if indexing not importing. See:
  1. Indexing generally should be used on a fairly static location, ie. you’re not moving folders in and out of the location, renaming, or deleting them. This leads to missing files in the database.