1

We are a small team (10 p) working on spatial planning. We are using QGIS as a main tool. We have a lot of open source data, we daily use on multiple projects, and therefore some files are duplicated over and over on our shared drive. For each project, people download or copy some files (like administrative boundries, or population data, etc.).

I would like to compile those sources in a unique central database or file, which would be accessible by anyone in my team. I've looked for some information about each format (SpatiaLite, GeoPackage, PostGIS), but struggle to pick one. Given that SpatiaLite is slower, ive decided to remove it from this equation.

At a high level, what are the functional and non-functional differences between a postGIS database and a Geopackage database.

4
  • 6
    That's not how relational databases operate; you create them on an unshared volume, and connect to them over the network. – Vince Jul 30 '20 at 12:42
  • 1
    Of those three, only PostGIS is "a unique central database". – Dan C Jul 30 '20 at 13:06
  • 1
    Postgres has the ability to have roles (user, admin, editor, read-only) you might want that control at some point. Whatever you decide have a backup, disaster recovery plan for your data. – Mapperz Jul 30 '20 at 13:28
  • 2
    Advice questions are off-topic in GIS SE. Editing the question to be more opinion-based is not likely to get the question reopened. I suggest you do some research on relational databases in general and PostgreSQL in particular, then set up a PosgreSQL database, install PostGIS in it, and give it a try. Along the way you might have specific questions, for which many answers already exist here, and if one doesn't then, you can ask it. @Mapperz has made a very important suggestion, so don't neglect the backup/recovery aspect of the equation. – Vince Jul 31 '20 at 13:54
4

It sounds like you have two paths, but need to distinguish each option.

  1. Files on a shared drive - SQLite/Spatialite is a file-based database that may serve the purpose. GeoPackage are flat files much like shapefiles, but with more capabilities.

  2. Data stored on a database server - files moved from shared drive to the server, and users would connect in QGIS to the server to access/create/manage/edit the files there instead. User access is managed through roles as mentioned above.

If you have the resources to set up a database server on your network, I recommend doing so. It would require a server (Windows, Linux) to run the PostgreSQL server, the PostGIS extension is installed and run through that database server, and QGIS is a client of the database.

If, however, you can only use your existing shared drive structure due to technical limitations of your organization, it might make sense to dig into the SQLite/Spatialite 'database' (again, a flat-file database not a database server), but it will accomplish some of what you want in terms of central access and avoiding file redundancy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.