Digital Läsesal

Contributing

Requirements

You need Docker, Node.js and the following tools to work with this project:

brew install jq kubectl kubectx skaffold

# Optionally, install these to make life easier:
brew install direnv nvm
# ...remember to setup shell hooks for them :) +1

Quick start

Assuming you are using direnv, create a .envrc file in this dir. Run docker compose up to run all of the microservices as well as elasticsearch and kibana. Run docker compose up kibana postgres elasticsearch & to run postgres, elasticsearch and kibana.

# Template .envrc
export COMPRIMA_SERVICE_URL=
export COMPRIMA_USER=
export COMPRIMA_PASSWORD=
export COMPRIMA_USER_TEST=
export COMPRIMA_PASSWORD_TEST=
export POSTGRES_USER=
export POSTGRES_PASSWORD=
export COMPRIMA_ADAPTER__URL=
export TESSDATA_PREFIX=
export ELASTIC_PASSWORD=
export KIBANA_PASSWORD=

Don't forget to run direnv allow after changing exports in .envrc!

Configuration

The following environment variables are needed when running locally:

reading-room-search

COOKIE_DOMAIN=localhost

reading-room-frontend

VITE_COOKIE_DOMAIN=localhost

Operations

All microservices are hosted with Kubernetes, see the Operations repo for more details.

External dependencies

We use GitHub Actions for ci/cd. Docker images are stored in the GitHub Container Registry. If builds/deploys are failing, check githubstatus.com.

Any server node that hosts the Comprima Adapter microservice needs to be connected to a VPN to be able to reach Comprima.

Troubleshooting

OCR Processing

There is a kibana setup for ci and production respecively. You can run a query to quickly check the status of the OCR processing.

# Open a tunnel to kibana.
kubectl -n production port-forward services/kibana-kibana 5601 # or -n ci

Open http://localhost:5601 and create the index pattern comprima if it does not already exist. Then go to the Discover tab, click Add filter and Edit as Query DSL to be able to enter JSON queries.

Find all documents that has the ocrText field set.

{
  "query": {
    "bool": {
      "must": {
        "exists": {
          "field": "ocrText"
        }
      }
    }
  }
}