Edit this page | Blame

Synchronising the Different Environments

Tags

  • status: open
  • priority:
  • type: documentation
  • assigned: fredm
  • keywords: doc, docs, documentation

Introduction

We have different environments we run for various reasons, e.g.

  • Production: This is the user-facing environment. This is what GeneNetwork is about.
  • gn2-fred: production-adjacent. It is meant to test out changes before they get to production. It is **NOT** meant for users.
  • CI/CD: Used for development. The latest commits get auto-deployed here. It's the first place (outside of developer machines) where errors and breakages are caught and/or revealed. This will break a lot. Do not expose to users!
  • staging: Uploader environment. This is where Felix, Fred and Arthur flesh out the upload process, and tasks, and also test out the uploader.

These different environments demand synchronisation, in order to have mostly similar results and failure modes.

Synchronisation of the Environments

Main Database: MariaDB

  • [ ] TODO: Describe process
  • Automate? Will probably need some checks for data sanity.

Authorisation Database

  • [ ] TODO: Describe process
  • Copy backup from production
  • Update/replace GN2 client configs in database
  • What other things?

Virtuoso/RDF

  • [ ] TODO: Describe process
  • Copy TTL (Turtle) files from (where?). Production might not always be latest source of TTL files.
  • Can we automate this? What checks are necessary?

Genotype Files

  • [ ] TODO: Describe process
  • Copy from source-of-truth (currently Zach's tux01 and/or production).
  • Rsync?

gn-docs

  • [ ] TODO: Describe process
  • Not sure changes from other environments should ever take

AI Summaries (aka. gnqna)

  • [ ] TODO: Describe process
  • Update configs (should be once, during container setup)

Others?

  • [ ] TODO: Describe process
(made with skribilo)