Skip to content

Qlever deployment documents #52

@fils

Description

@fils

Initial thoughts.

  • we create our own container... which means we have to trust that we sync up to a version.
  • second thoughts, we write a startup script, and add container.

Maybe making the qlever thing too hard...
Make a bash scripts to do this>
if configdone file does not exist, file with the org triples.
run command to build initial index, write a configdone file
Run command to start server
If config done exists, start server.

@valentinedwv @ylyangtw

OK, I pushed up the directory: https://github.com/earthcube/geocodes/tree/qlever/triplestore/qlever/configs/blank

Points:

  1. The directory rdf in this folder isn't needed, so if directories are annoying to load, it does not need to be included
  2. Qlever can't init with 0 triples, so I made just a simple RDF nq file (https://github.com/earthcube/geocodes/blob/qlever/triplestore/qlever/configs/blank/rdf/org.nq) to init with. Later we could modify this into an actual information set of triples about the graph. Likely using something like VoID. I'll work up an example for that, but we don't need to hold up for that.
  3. Just FYI, it looks like they might be looking into this, (run with empty graph) but it's on a TODO list.

Metadata

Metadata

Assignees

No one assigned

    Type

    Projects

    Status

    No status

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions