61 lines
2.0 KiB
Markdown
61 lines
2.0 KiB
Markdown
# chromadb-admin
|
|
|
|
Local deployment of [chromadb-admin](https://github.com/flanker/chromadb-admin) — a Next.js admin UI for [ChromaDB](https://docs.trychroma.com), the vector database used by **ScraperControl** for semantic search and deduplication.
|
|
|
|
## Services
|
|
|
|
| Service | Port | Description |
|
|
|---------|------|-------------|
|
|
| `chromadb` | `8000` | ChromaDB vector database |
|
|
| `chromadb-admin` | `3002` | Next.js admin UI (not currently running) |
|
|
|
|
## ChromaDB Configuration
|
|
|
|
- **Config file:** `/opt/chromadb/config.yaml`
|
|
- **Data directory:** `/opt/chromadb/data` (bind-mounted)
|
|
- **Persistence:** Enabled (`IS_PERSISTENT=TRUE`)
|
|
- **Telemetry:** Disabled (`ANONYMIZED_TELEMETRY=FALSE`)
|
|
|
|
## ScraperControl Collections
|
|
|
|
ScraperControl uses 5 ChromaDB collections:
|
|
|
|
| Collection | Purpose | Document format |
|
|
|------------|---------|-----------------|
|
|
| `church_identity` | Church deduplication | `{name} {address} {city} {country}` |
|
|
| `search_results` | FreeSearch result matching | `{title} {snippet} {url}` |
|
|
| `page_classification` | Content classification | Page text (first 2000 chars) |
|
|
| `schedule_sections` | Mass schedule detection | Text blocks with mass times |
|
|
| `page_snapshots` | Change detection | Full page text |
|
|
|
|
Embeddings are generated by **Ollama** (`nomic-embed-text`, 274 MB) running at `http://192.168.0.241:11434`.
|
|
|
|
## Running
|
|
|
|
```bash
|
|
# Start ChromaDB only
|
|
docker compose up -d chromadb
|
|
|
|
# Start everything including admin UI
|
|
docker compose up -d
|
|
```
|
|
|
|
Access the admin UI at `http://192.168.0.241:3002` (when running).
|
|
|
|
Connect using: `http://192.168.0.241:8000`
|
|
|
|
## Deployment
|
|
|
|
Runs on `192.168.0.241` (albert-MacBookPro Linux server).
|
|
|
|
**Docker Compose:** `/opt/docker/chromadb/docker-compose.yml`
|
|
|
|
| Endpoint | URL |
|
|
|----------|-----|
|
|
| ChromaDB API | `http://192.168.0.241:8000` |
|
|
| Admin UI | `http://192.168.0.241:3002` (when running) |
|
|
|
|
## Upstream
|
|
|
|
This is a local deployment of the open-source [flanker/chromadb-admin](https://github.com/flanker/chromadb-admin) project (MIT license), with a custom `docker-compose.yml` for this server.
|