# Selbstgehostete Vektordatenbanken mit Qdrant: Ein tiefer Tauchgang

<datetime class="hidden">2025-11-23T13:00</datetime>

<!-- category -- ASP.NET, Semantic Search, Vector Databases, Qdrant, RAG, AI-Article -->
# Einleitung

**Im Zusammenhang mit der RAG-Serie:** Dieser Artikel bietet einen tiefen Einblick in Qdrant, die Vektordatenbank, die verwendet wird in:

- [Teil 4: Umsetzung von ONNX und Qdrant](/blog/semantic-search-with-onnx-and-qdrant) - Gebäude semantische Suche
- [Teil 5: Hybride Suche & Auto-Indexierung](/blog/rag-hybrid-search-and-indexing) - Produktionsintegration

[Qdrant](https://qdrant.tech/) (ausgesprochen "Quadrant") ist eine Open-Source-Vektor-Datenbank in Rust. Dieser Artikel umfasst Kernkonzepte, den C#-Client, Performance-Tuning und Produktionsmuster.

[TOC]

# Was ist Qdrant?

A [Vektordatenbank](https://qdrant.tech/documentation/overview/) speichert hochdimensionale Vektoren (Embeddings) und ermöglicht eine schnelle Ähnlichkeitssuche. Im Gegensatz zu traditionellen Datenbanken, die exakte Übereinstimmungen finden, findet Qdrant *semantisch ähnlich* Artikel.

```mermaid
flowchart LR
    A[Text: 'Docker deployment'] --> B[Embedding Model]
    B --> C["Vector: [0.12, -0.34, 0.56, ...]"]
    C --> D[Qdrant]
    E[Query: 'container setup'] --> F[Embedding Model]
    F --> G["Vector: [0.11, -0.32, 0.58, ...]"]
    G --> H[Similarity Search]
    D --> H
    H --> I[Similar Results]

    style B stroke:#6366f1,stroke-width:3px
    style D stroke:#ef4444,stroke-width:3px
    style F stroke:#6366f1,stroke-width:3px
    style H stroke:#10b981,stroke-width:2px
```

**Hauptmerkmale von Qdrant:**

- [HNSW-Indizierung](https://qdrant.tech/documentation/concepts/indexing/) - Sublineare Suchzeiten
- [Filtern](https://qdrant.tech/documentation/concepts/filtering/) - Ähnlichkeitssuche mit Metadatenfiltern kombinieren
- [gRPC & REST APIs](https://qdrant.tech/documentation/interfaces/) - High-Performance-Zugang
- [Verteilte Bereitstellung](https://qdrant.tech/documentation/guides/distributed_deployment/) - Horizontal skalieren
- [Schnappschüsse](https://qdrant.tech/documentation/concepts/snapshots/) - Sicherung und Wiederherstellung

# Kernkonzepte

## Sammlungen

A [Sammlung](https://qdrant.tech/documentation/concepts/collections/) ist wie eine Tabelle - es hält Vektoren mit einer festen Dimensionalität und Distanzmetrik.

```mermaid
flowchart TB
    subgraph Collection["Collection: blog_posts"]
        A[Vector Size: 384]
        B[Distance: Cosine]
        C[HNSW Index]
    end

    subgraph Points
        D[Point 1: slug=docker-intro]
        E[Point 2: slug=kubernetes-basics]
        F[Point N...]
    end

    Collection --> Points

    style A stroke:#6366f1,stroke-width:2px
    style B stroke:#6366f1,stroke-width:2px
    style C stroke:#f59e0b,stroke-width:2px
    style D stroke:#10b981,stroke-width:2px
    style E stroke:#10b981,stroke-width:2px
```

```csharp
// Create collection - see https://qdrant.tech/documentation/concepts/collections/#create-a-collection
await client.CreateCollectionAsync(
    collectionName: "blog_posts",
    vectorsConfig: new VectorParams
    {
        Size = 384,              // Must match your embedding model
        Distance = Distance.Cosine  // Best for text embeddings
    }
);
```

**Abstandsmetriken** ([n s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s](https://qdrant.tech/documentation/concepts/collections/#distance-metrics)):

- **Cosin** - misst Winkel zwischen Vektoren (am besten für Text)
- **Punkt** - Rohes Innenprodukt (für vornormierte Vektoren)
- **Euklid** - Geometrische Entfernung (für Geodaten)

## Punkte

A [Ziffer](https://qdrant.tech/documentation/concepts/points/) Es handelt sich um einen einzigen Datensatz, der Folgendes enthält:

```mermaid
flowchart LR
    subgraph Point
        A[ID: uuid/int]
        B["Vector: float[384]"]
        C[Payload: JSON metadata]
    end

    style A stroke:#8b5cf6,stroke-width:2px
    style B stroke:#f59e0b,stroke-width:2px
    style C stroke:#10b981,stroke-width:2px
```

```csharp
// Upsert points - see https://qdrant.tech/documentation/concepts/points/#upload-points
var point = new PointStruct
{
    Id = new PointId { Uuid = Guid.NewGuid().ToString() },
    Vectors = embedding,  // float[384]
    Payload =
    {
        ["slug"] = "my-post",
        ["title"] = "Vector Databases",
        ["language"] = "en",
        ["categories"] = new[] { "AI", "Databases" },
        ["published"] = DateTimeOffset.UtcNow.ToUnixTimeSeconds()
    }
};

await client.UpsertAsync("blog_posts", points: new[] { point });
```

## Filtern

[Filtern](https://qdrant.tech/documentation/concepts/filtering/) läuft *Vorher* Ähnlichkeitssuche - extrem effizient.

```mermaid
flowchart TB
    A[Search Query] --> B{Apply Filters First}
    B --> C[Language = 'en']
    B --> D[Year >= 2024]
    C --> E[Filtered Subset]
    D --> E
    E --> F[Vector Similarity Search]
    F --> G[Ranked Results]

    style B stroke:#ec4899,stroke-width:3px
    style E stroke:#f59e0b,stroke-width:2px
    style F stroke:#6366f1,stroke-width:2px
    style G stroke:#10b981,stroke-width:2px
```

```csharp
// Filter conditions - see https://qdrant.tech/documentation/concepts/filtering/#filtering-conditions
var filter = new Filter
{
    Must =  // AND conditions
    {
        new Condition { Field = new FieldCondition
        {
            Key = "language",
            Match = new Match { Keyword = "en" }
        }},
        new Condition { Field = new FieldCondition
        {
            Key = "published",
            Range = new Range { Gte = 1704067200 }  // 2024-01-01
        }}
    },
    MustNot =  // Exclude conditions
    {
        new Condition { Field = new FieldCondition
        {
            Key = "slug",
            Match = new Match { Keyword = "draft-post" }
        }}
    }
};
```

**Filtertypen** ([n s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s s](https://qdrant.tech/documentation/concepts/filtering/#match)):

- `Match.Keyword` - Genauer String-Match
- `Match.Text` - Volltext-Match
- `Match.Any` - Passen Sie alle in Array
- `Range` - Zahlenbereiche (Gte, Lte, Gt, Lt)
- `GeoBoundingBox` / `GeoRadius` - Geofilterung

# Der C#-Client

Installieren Sie den offiziellen [Qdrant.Client](https://www.nuget.org/packages/Qdrant.Client) Verpackung ([GitHub](https://github.com/qdrant/qdrant-dotnet)):

```bash
dotnet add package Qdrant.Client
```

## Verbindungsaufbau

```csharp
using Qdrant.Client;
using Qdrant.Client.Grpc;

// gRPC client (recommended) - see https://qdrant.tech/documentation/interfaces/#grpc-interface
var client = new QdrantClient(
    host: "localhost",
    port: 6334,  // gRPC port (6333 is REST)
    https: false
);

// With API key - see https://qdrant.tech/documentation/guides/security/
var secureClient = new QdrantClient(
    host: "your-qdrant.cloud",
    port: 6334,
    https: true,
    apiKey: "your-api-key"
);
```

> **Verwenden Sie gRPC immer** (Port 6334) für die Produktion - 3-5x schneller als REST.

## Windows HTTP/2 Fixieren

Aktivieren Sie unter Windows unverschlüsselt HTTP/2 **Vorher** Erstellen des Clients:

```csharp
AppContext.SetSwitch("System.Net.Http.SocketsHttpHandler.Http2UnencryptedSupport", true);
```

## Schlüsseloperationen

### Suchen

```csharp
// Vector search - see https://qdrant.tech/documentation/concepts/search/
var results = await client.SearchAsync(
    collectionName: "blog_posts",
    vector: queryEmbedding,
    limit: 10,
    filter: filter,
    scoreThreshold: 0.5f,  // Minimum similarity
    searchParams: new SearchParams
    {
        HnswEf = 128,  // Search accuracy (higher = better recall)
        Exact = false  // Use approximate search
    },
    withPayload: true
);

foreach (var result in results)
{
    Console.WriteLine($"{result.Payload["title"].StringValue}: {result.Score}");
}
```

### Batch Upsert

```csharp
// Batch operations - see https://qdrant.tech/documentation/concepts/points/#batch-update
var points = documents.Select(doc => new PointStruct
{
    Id = new PointId { Uuid = doc.Id },
    Vectors = doc.Embedding,
    Payload = { ["slug"] = doc.Slug, ["title"] = doc.Title }
}).ToList();

await client.UpsertAsync(
    collectionName: "blog_posts",
    points: points,
    wait: true  // Wait for indexing
);
```

### Löschen

```csharp
// Delete by filter - see https://qdrant.tech/documentation/concepts/points/#delete-points
await client.DeleteAsync(
    collectionName: "blog_posts",
    filter: new Filter
    {
        Must = { new Condition { Field = new FieldCondition
        {
            Key = "slug",
            Match = new Match { Keyword = "old-post" }
        }}}
    }
);
```

# HNSW-Index-Tuning

[HNSW](https://qdrant.tech/documentation/concepts/indexing/#vector-index) (Hierarchical Navigable Small World) ist Qdrants Indexalgorithmus.

```mermaid
flowchart TB
    subgraph "HNSW Graph Layers"
        L2[Layer 2 - Sparse]
        L1[Layer 1 - Medium]
        L0[Layer 0 - Dense]
    end

    Q[Query] --> L2
    L2 --> L1
    L1 --> L0
    L0 --> R[Nearest Neighbors]

    style L2 stroke:#8b5cf6,stroke-width:2px
    style L1 stroke:#6366f1,stroke-width:2px
    style L0 stroke:#3b82f6,stroke-width:2px
    style Q stroke:#10b981,stroke-width:2px
    style R stroke:#ef4444,stroke-width:2px
```

## Index-Parameter

```csharp
// HNSW config - see https://qdrant.tech/documentation/concepts/indexing/#hnsw-index
var hnswConfig = new HnswConfigDiff
{
    M = 16,              // Edges per node (16-32 recommended)
    EfConstruct = 100,   // Build-time accuracy (100-200)
    FullScanThreshold = 10000  // Brute force threshold
};

await client.UpdateCollectionAsync(
    collectionName: "blog_posts",
    hnswConfig: hnswConfig
);
```

**Suchzeitgenauigkeit:**

```csharp
var searchParams = new SearchParams
{
    HnswEf = 128  // Higher = better recall, slower (64-256)
};
```

**Tuning-Leitlinien:**
Verwenden Sie Fall , M , EfConstruct , HnswEf , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,
|----------|---|-------------|--------|
Schnelles, niedriges Rückrufen 8, 64, 32,
Ausgeglichen 16, 100, 128,
Hoch zurückrufen 32, 200, 256,

# Nutzlastindizes

Erstellen [Nutzlastindizes](https://qdrant.tech/documentation/concepts/indexing/#payload-index) für häufig gefilterte Felder:

```csharp
// Keyword index - see https://qdrant.tech/documentation/concepts/indexing/#payload-index
await client.CreatePayloadIndexAsync(
    collectionName: "blog_posts",
    fieldName: "language",
    schemaType: PayloadSchemaType.Keyword
);

// Integer index for ranges
await client.CreatePayloadIndexAsync(
    collectionName: "blog_posts",
    fieldName: "published",
    schemaType: PayloadSchemaType.Integer
);
```

**Auswirkungen:** 10-100x schnellere Filterung auf großen Sammlungen.

# Quantifizierung

[Quantifizierung](https://qdrant.tech/documentation/guides/quantization/) reduziert den Speicherverbrauch:

```csharp
// Scalar quantization - see https://qdrant.tech/documentation/guides/quantization/#scalar-quantization
await client.UpdateCollectionAsync(
    collectionName: "blog_posts",
    quantizationConfig: new ScalarQuantization
    {
        Scalar = new ScalarQuantizationConfig
        {
            Type = ScalarType.Int8,  // float32 -> int8
            Quantile = 0.99f,
            AlwaysRam = true
        }
    }
);
```

**Abtretung:** 4x weniger Speicher, ~2% Rückrufverlust, 1.5x schnellere Suche.

# Docker-Einsatz

```yaml
# docker-compose.yml - see https://qdrant.tech/documentation/guides/installation/
services:
  qdrant:
    image: qdrant/qdrant:v1.12.1  # Pin version!
    ports:
      - "6333:6333"  # REST
      - "6334:6334"  # gRPC
    volumes:
      - qdrant_data:/qdrant/storage
    environment:
      - QDRANT__SERVICE__GRPC_PORT=6334
      - QDRANT__SERVICE__HTTP_PORT=6333
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:6333/health"]
      interval: 30s
      timeout: 10s
      retries: 3

volumes:
  qdrant_data:
```

## Sicherheit

Aktivieren [API-Schlüssel-Authentifizierung](https://qdrant.tech/documentation/guides/security/):

```yaml
environment:
  - QDRANT__SERVICE__API_KEY=your-secret-key
```

# Überwachung

## Metrik

Qdrant entlarvt [Prometheus-Metriken](https://qdrant.tech/documentation/guides/monitoring/) bei `/metrics`:

```bash
curl http://localhost:6333/metrics
```

Schlüsselkennzahlen:

- `qdrant_collections_vector_count` - Vektoren insgesamt
- `qdrant_rest_responses_duration_seconds` - Latenz abfragen
- `qdrant_memory_usage_bytes` - Speicherverbrauch

## Schnappschüsse

Erstellen [Sicherungen](https://qdrant.tech/documentation/concepts/snapshots/):

```bash
# Create snapshot
curl -X POST http://localhost:6333/collections/blog_posts/snapshots

# List snapshots
curl http://localhost:6333/collections/blog_posts/snapshots

# Restore (copy snapshot to storage/collections/blog_posts/snapshots/)
```

# Häufig Gotchas

## 1. Port Confusion

- **6333** = REST API
- **6334** = gRPC API (verwenden!)

## 2. Missanpassung der Vektor-Dimension

```
Error: expected dim: 384, got 768
```

Ihr Einbettungsmodell und Ihre Sammlung müssen stimmen:

- `all-MiniLM-L6-v2`: 384 Abmessungen
- `nomic-embed-text`: 768 Abmessungen
- OpenAI `text-embedding-3-small`: 1536 Abmessungen

## 3. Langsame erste Abfrage

HNSW Lazy-loads in den Speicher. Aufwärmen nach dem Start:

```csharp
await client.SearchAsync("blog_posts", new float[384], limit: 1);
```

## 4. Array Filtern

Verwendung `Match.Any` für Array-Felder:

```csharp
new Match { Any = new RepeatedStrings { Strings = { "AI", "ML" } } }
```

# Aufkommen

## Offizielle Qdrant-Dokumentation

- [Überblick](https://qdrant.tech/documentation/overview/) - Fangen wir an.
- [Begriffsbestimmungen](https://qdrant.tech/documentation/concepts/) - Kernkonzepte
- [Sammlungen](https://qdrant.tech/documentation/concepts/collections/) - Erstellen und Verwalten von Sammlungen
- [Punkte](https://qdrant.tech/documentation/concepts/points/) - Arbeiten mit Vektoren
- [Suchen](https://qdrant.tech/documentation/concepts/search/) - Abfragen
- [Filtern](https://qdrant.tech/documentation/concepts/filtering/) - Filterbedingungen
- [Indexierung](https://qdrant.tech/documentation/concepts/indexing/) - HNSW und Nutzlastindizes
- [Quantifizierung](https://qdrant.tech/documentation/guides/quantization/) - Speicheroptimierung
- [Sicherheit](https://qdrant.tech/documentation/guides/security/) - Authentifizierung
- [Überwachung](https://qdrant.tech/documentation/guides/monitoring/) - Metrische und Telemetrie
- [Schnappschüsse](https://qdrant.tech/documentation/concepts/snapshots/) - Sicherung und Wiederherstellung

## Client-Bibliotheken

- [Qdrant .NET Client](https://github.com/qdrant/qdrant-dotnet) - Offizielles C# SDK
- [NuGet-Paket](https://www.nuget.org/packages/Qdrant.Client) - Letzte Veröffentlichung

## Verwandte Artikel

- [Teil 4: Umsetzung von ONNX und Qdrant](/blog/semantic-search-with-onnx-and-qdrant)
- [Teil 5: Hybride Suche & Auto-Indexierung](/blog/rag-hybrid-search-and-indexing)
- [Überblick über die RAG-Serien](/blog/rag-primer)

## Quellencode

Alle verfügbaren Codes unter: [github.com/scottgal/mostlylucidweb](https://github.com/scottgal/mostlylucidweb)

- `Mostlylucid.SemanticSearch/Services/QdrantVectorStoreService.cs` - Qdrant-Integration