Full-text Search with ParadeDB

Vikunja supports using ParadeDB for improved full-text search capabilities. ParadeDB is a PostgreSQL extension that provides fast, fuzzy matching search similar to Typesense, but without requiring a separate service.

Requirements #

  • PostgreSQL database: ParadeDB only works with PostgreSQL (not MySQL or SQLite)
  • pg_search extension: The ParadeDB pg_search extension must be installed in your PostgreSQL instance

Setup #

Using the ParadeDB Docker Image #

The easiest way to get started is using the official ParadeDB Docker image, which comes with all extensions pre-installed and can be used as a drop-in replacement for a standard PostgreSQL container:

services:
  db:
    image: paradedb/paradedb:latest
    environment:
      POSTGRES_PASSWORD: changeme
      POSTGRES_USER: vikunja
      POSTGRES_DB: vikunja
    volumes:
      - ./db:/var/lib/postgresql/data
    restart: unless-stopped
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -h localhost -U $$POSTGRES_USER"]
      interval: 2s
      start_period: 30s

That's it! Vikunja will automatically detect ParadeDB and enable full-text search.

Using an Existing PostgreSQL Installation #

If you're running your own PostgreSQL server:

  1. Install the pg_search extension following the ParadeDB installation guide
  2. Ensure the extension is available in your Vikunja database
  3. Restart Vikunja - it will automatically detect and use ParadeDB

Migration from Typesense #

If you're currently using Typesense and want to migrate to ParadeDB:

  1. Set up PostgreSQL with ParadeDB (see setup instructions above)
  2. If you're not already using PostgreSQL, migrate your database first
  3. Remove the Typesense configuration from your Vikunja config
  4. Restart Vikunja

Vikunja will automatically detect ParadeDB and begin using it for search. Your existing task data will be indexed automatically.

How It Works #

When ParadeDB is available, Vikunja creates a BM25 search index on your tasks table. This index enables:

  • Fast full-text search across task titles and descriptions
  • Fuzzy matching for typo tolerance
  • Relevance-based result ranking

The index is maintained automatically as tasks are created, updated, or deleted.

On this page