Files
meldestelle/infrastructure/event-store/redis-event-store
StefanMo b35c4087a2 Fix: Test-Commit für VCS-Integration (MP-8) (#15)
* MP-8 OTHER Implementiere JWT-Authentifizierungs-Filter im Gateway

* Fix(ci): Update upload-artifact action to v4

* Fix(ci): Add start command for Keycloak and failure logs

* Fix(ci): Remove invalid 'command' property from Keycloak service

* Fix(ci): Use KC_DEV_MODE env var to start Keycloak

* Fix(ci): Keycloak service was removed from GitHub Actions services and replaced with a manual docker run step that starts Keycloak with the start-dev command.

* dev(ci): vereinheitliche Keycloak auf 26.4.2; aktiviere Health im CI (MP-8)

* Fix(ci): Stabilize Keycloak startup in integration tests via matrix

- Add `dev-file` Keycloak variant to matrix for stability fallback.
- Improve wait logic and health checks for Keycloak and Postgres.
- Unify Keycloak version to 26.4.2 across codebase.
- Add log dumps on failure.

* Fix(ci): Die betroffene Datei docs/Visionen-Ideen/Infrastruktur-Strategie_DSGVO-Konformität.md endet aktuell mit genau einer leeren Zeile (Zeile 87). Das entspricht der Regel MD047 („Files should end with a single newline character“). Damit ist deine Korrektur korrekt.

* Fix(ci): Repository-wide auto-fix for Markdown files was implemented with a GitHub Actions workflow and a local helper script. EditorConfig and markdownlint ignore files were added to ensure consistent formatting. Instructions for using the auto-fix both via GitHub Actions and locally were provided.

* fix(gradle): build.gradle.kts jsBrowser testTask disabled

* fix(gradle): build.gradle.kts jsBrowser testTask disabled

* Fix(ci): Stabilize integration tests with Keycloak matrix build (MP-8)

Introduces a matrix strategy (`keycloak_db: [postgres, dev-file]`)
in the integration-tests workflow to mitigate flaky Keycloak starts
when using the Postgres service container.

- Adds a `dev-file` Keycloak variant for stability fallback.
- Improves wait logic and health checks for Keycloak/Postgres.
- Unifies Keycloak version to 26.4.2 across codebase (Dockerfile, Compose,
  ADR, README, tests).
- Adds log dumps on failure in CI.
- Ensures `KC_HEALTH_ENABLED=true` is set.
- Updates related documentation (README, Schlachtplan).
- Includes broader Docker SSoT cleanup (versions.toml as source,
  script updates, env file cleanup, validator hardening).

This resolves recurring CI failures related to Keycloak startup and
ensures required checks for PRs (#15) are reliable, while also
improving overall Docker build consistency.

* feat(docs, ci): Implement YouTrack SSoT strategy with Dokka sync (MP-8)

- Add Dokka multi-module Gradle configuration and KDoc style guide.
- Add GitHub Actions workflow (docs-kdoc-sync.yml) and Python script
  (youtrack-sync-kb.py) to sync Dokka GFM output to YouTrack KB.
- Extend front-matter schema (bc, doc_type) and update relevant pages/stubs.
- Adapt CI scripts (validate-frontmatter, check-docs-drift, ci-docs link ignore).
- Update README.md to reference YouTrack KB.

* feat(docs, ci): Implement YouTrack SSoT strategy with Dokka sync (MP-8)

- Add Dokka multi-module Gradle configuration and KDoc style guide.
- Add GitHub Actions workflow (docs-kdoc-sync.yml) and Python script
  (youtrack-sync-kb.py) to sync Dokka GFM output to YouTrack KB.
- Extend front-matter schema (bc, doc_type) and update relevant pages/stubs.
- Adapt CI scripts (validate-frontmatter, check-docs-drift, ci-docs link ignore).
- Update README.md to reference YouTrack KB.

* Fix(ci): Replace OpenAPI validator with Spectral

Replaces the deprecated 'char0n/swagger-editor-validate' action,
which failed due to sandbox issues in GitHub Actions, with the
modern '@stoplight/spectral-cli'.

This ensures robust OpenAPI specification validation without
requiring a headless browser environment. The 'generate-api-docs'
job now depends on the successful completion of the Spectral validation.

Part of resolving CI failures for PR #15 (MP-8).

* Fix(ci): Specify spectral:oas ruleset for OpenAPI validation (MP-8)

* Fix(ci): Remove explicit ruleset argument for Spectral validation (MP-8)

* Fix(ci): Added a .spectral.yaml file to fix Spectral linting errors. Corrected markdown lint issues in two documentation files. Updated README.md with a new guidelines section to fix link validation errors.

* Fix(ci): Markdownlint errors were fixed by adding required blank lines. The Guidelines Validation error was resolved by updating the README.md link. The API Documentation Generator workflow was stabilized by updating paths, tasks, and validation steps.

* Fix(ci): Alle vier fehlerhaften GitHub-Action-Prüfungen wurden behoben. Fehler in der OpenAPI-Spezifikation, Probleme mit der Markdown-Linting-Analyse und Validierungsfehler bei Querverweisen wurden korrigiert. Die README.md enthält nun alle erforderlichen Links zu den Richtlinien.

* Fix(ci): Markdown linting errors in docs/api/README.md were fixed by specifying languages in fenced code blocks. OpenAPI specification errors in documentation.yaml were resolved by correcting example property types to strings. Cross-reference validation errors in README.md were fixed by adding the missing link to project-standards/coding-standards.md.

* Fix(ci): Duplicate heading errors in docs/api/members-api.md were fixed. Cross-reference validation errors for docker-architecture.md were resolved. All originally reported issues passed validation successfully.

* Fix(ci): The markdown heading levels in docs/api/members-api.md were corrected from h5 to h4 to fix linting errors. The missing cross-reference link from technology-guides/docker/docker-development.md to docker-overview.md was added. These fixes resolved the original validation and linting errors causing the process to fail.

* Fix(ci): Duplicate heading warnings in docs/api/members-api.md were resolved. Cross-reference validation for docker-development.md to docker-architecture.md was fixed. A new unrelated warning about docker-production.md was identified but not addressed.

* refactor(ci,docs): Simplify CI pipeline and migrate docs to YouTrack SSoT

BREAKING CHANGE: Documentation structure radically simplified

- Consolidate 9 GitHub Actions workflows into 1 main pipeline (ci-main.yml)
- Remove redundant workflows: ci-docs, markdownlint-autofix, guidelines-validation, api-docs
- Delete documentation migrated to YouTrack: api/, BCs/, Visionen-Ideen/, reference/, now/, overview/
- Keep only ADRs, C4 diagrams, and essential dev guides in repo
- Update README.md with YouTrack KB links
- Create new docs/README.md as documentation gateway
- Relax markdown-lint config for pragmatic developer experience

Kept workflows:
- ssot-guard.yml (Docker SSoT validation)
- docs-kdoc-sync.yml (KDoc → YouTrack sync)
- integration-tests.yml (Integration tests)
- deploy-proxmox.yml (Deployment)
- youtrack-sync.yml (YouTrack integration)

Related: MP-DOCS-001

* refactor(ci,docs): Simplify CI pipeline and migrate docs to YouTrack SSoT

BREAKING CHANGE: Documentation structure radically simplified

- Consolidate 9 GitHub Actions workflows into 1 main pipeline (ci-main.yml)
- Remove redundant workflows: ci-docs, markdownlint-autofix, guidelines-validation, api-docs
- Delete documentation migrated to YouTrack: api/, BCs/, Visionen-Ideen/, reference/, now/, overview/
- Keep only ADRs, C4 diagrams, and essential dev guides in repo
- Update README.md with YouTrack KB links
- Create new docs/README.md as documentation gateway
- Relax markdown-lint config for pragmatic developer experience

Kept workflows:
- ssot-guard.yml (Docker SSoT validation)
- docs-kdoc-sync.yml (KDoc → YouTrack sync)
- integration-tests.yml (Integration tests)
- deploy-proxmox.yml (Deployment)
- youtrack-sync.yml (YouTrack integration)

Related: MP-DOCS-001

* refactor(ci,docs): README.md und einige andere Dokumentationen überarbeitet.
ports-and-urls.md hinzugefügt.
Related: MP-DOCS-001

* refactor(ci,docs): Die Markdownlint-Fehler in README.md und docs/README.md wurden behoben, indem die Überschriftenebenen angepasst, überflüssige Satzzeichen am Ende entfernt und die notwendigen Leerzeilen um Überschriften, Listen, Tabellen und Codeblöcke eingefügt wurden. Das problematische Leerzeichen am Ende in docs/README.md wurde ebenfalls entfernt. Die Dateien entsprechen nun den vorgegebenen Markdownlint-Regeln und sollten die CI-Validierung bestehen.
Related: MP-DOCS-001

* refactor(ci,docs): Docker guideline cross-references were fixed and normalized to lowercase labels. Validation scripts confirmed zero cross-reference warnings and consistent metadata. Documentation was updated with a changelog and enhanced README navigation.
Related: MP-DOCS-001

* refactor(ci,docs): Docker guideline cross-references were fixed and normalized to lowercase labels. Validation scripts confirmed zero cross-reference warnings and consistent metadata. Documentation was updated with a changelog and enhanced README navigation.
Related: MP-DOCS-001

* refactor(ci,docs): Dead links in docs/architecture/adr were fixed by updating URLs to stable sources and adding an ignore pattern for a placeholder link. Specific ADR files had their broken links replaced with valid ones. The markdown-link-check GitHub Action is expected to pass with zero dead links now.
Related: MP-DOCS-001

* refactor(ci,docs): Links in ADR checked
Related: MP-DOCS-001

* refactor(ci,docs): Links in ADR checked
Related: MP-DOCS-001

* refactor(ci,docs): Markdown Regeln ausgebessert
Related: MP-DOCS-001

* refactor(ci,docs): Markdown Regeln ausgebessert
Related: MP-DOCS-001

* refactor(ci,docs): Markdown Regeln ausgebessert
Related: MP-DOCS-001

* Chore: Rerun CI checks with updated branch protection rules
2025-11-07 12:26:33 +01:00
..
2025-10-14 12:47:16 +02:00
2025-10-14 12:36:00 +02:00

Redis Event Store Module

Überblick

Dieses Modul stellt eine konkrete Implementierung der event-store-api unter Verwendung von Redis Streams als Event-Store-Backend bereit.

Architektur

Das Modul folgt dem Provider-Pattern:

  • event-store-api: Provider-agnostische Interfaces (EventStore, EventSerializer)
  • redis-event-store: Redis Streams-spezifische Implementierung

Verwendung

Dependency Hinzufügen

dependencies {
    implementation(projects.infrastructure.eventStore.redisEventStore)
}

Konfiguration

Das Modul verwendet Spring Boot Auto-Configuration. Konfigurieren Sie Redis über application.yml:

redis:
  event-store:
    host: localhost
    port: 6379
    password: null  # Optional
    database: 1     # Separate database for event store (default: 1)
    connectionTimeout: 2000
    readTimeout: 2000
    usePooling: true
    maxPoolSize: 8
    minPoolSize: 2
    consumerGroup: event-processors
    consumerName: event-consumer
    streamPrefix: "event-stream:"
    allEventsStream: all-events
    claimIdleTimeout: 60s
    pollTimeout: 100ms
    maxBatchSize: 100
    createConsumerGroupIfNotExists: true

Code-Beispiel

@Service
class MyEventService(
    private val eventStore: EventStore,
    private val eventConsumer: RedisEventConsumer
) {

    // Event speichern
    suspend fun saveEvent(aggregateId: Uuid, event: DomainEvent) {
        eventStore.appendEvent(
            aggregateId = aggregateId,
            event = event,
            expectedVersion = EventVersion.ANY
        )
    }

    // Events abrufen
    suspend fun loadEvents(aggregateId: Uuid): List<DomainEvent> {
        return eventStore.loadEvents(aggregateId)
    }

    // Events konsumieren
    fun startConsuming() {
        eventConsumer.consumeEvents { event ->
            println("Received event: $event")
        }
    }
}

Features

  • Event Sourcing mit Redis Streams
  • Optimistic Locking mit Event Versioning
  • Consumer Groups für parallele Event-Verarbeitung
  • Event Replay-Fähigkeit
  • Pub/Sub für Event-Benachrichtigungen
  • Jackson-basierte Serialisierung
  • Connection Pooling mit Lettuce
  • Kotlin Coroutines Support

Redis Streams

Das Modul nutzt Redis Streams für Event Sourcing:

  • Stream pro Aggregate: event-stream:{aggregateId}
  • All Events Stream: event-stream:all-events
  • Consumer Groups: Für parallele Verarbeitung
  • Message IDs: Für Event-Ordering und Replay

Beans

Das Modul registriert folgende Spring Beans:

  • eventStoreRedisConnectionFactory: Separate Redis ConnectionFactory für Event Store
  • eventStoreRedisTemplate: StringRedisTemplate für Event-Operationen
  • eventSerializer: Jackson-basierter Event-Serializer
  • eventStore: EventStore Implementierung
  • eventConsumer: RedisEventConsumer für Event-Verarbeitung

Gleichzeitige Verwendung mit redis-cache

⚠️ WICHTIG: Wenn Sie sowohl redis-cache als auch redis-event-store im selben Service verwenden:

Unterschiedliche Databases

Die Module verwenden separate Redis Databases, um Konflikte zu vermeiden:

  • redis-cache: Database 0 (Standard)
  • redis-event-store: Database 1 (Standard, konfigurierbar)

Konfigurationsbeispiel

# Beide Module in einer application.yml
redis:
  # Cache Konfiguration
  host: localhost
  port: 6379
  database: 0  # Cache verwendet Database 0

  # Event Store Konfiguration (nested)
  event-store:
    host: localhost
    port: 6379
    database: 1  # Event Store verwendet Database 1
    consumerGroup: event-processors

Bean-Namen

Die Module verwenden unterschiedliche Bean-Namen zur Vermeidung von Konflikten:

Komponente redis-cache redis-event-store
ConnectionFactory redisConnectionFactory eventStoreRedisConnectionFactory
Template redisTemplate eventStoreRedisTemplate
Serializer cacheSerializer eventSerializer

Keine Konflikte

Die Module sind so designed, dass sie ohne Konflikte gleichzeitig verwendet werden können:

  • Separate ConnectionFactories mit @Qualifier Annotations
  • Separate Property-Prefixes: redis vs redis.event-store
  • Unterschiedliche Database-Nummern: 0 vs 1
  • Unterschiedliche Bean-Namen: Explizite Qualifier verhindern Kollisionen

Event Versioning

Das Modul unterstützt Optimistic Locking:

// Erwartete Version spezifizieren
eventStore.appendEvent(
    aggregateId = aggregateId,
    event = myEvent,
    expectedVersion = EventVersion.of(5)  // Erwartet Version 5
)

// Beliebige Version akzeptieren
eventStore.appendEvent(
    aggregateId = aggregateId,
    event = myEvent,
    expectedVersion = EventVersion.ANY
)

Bei Version-Konflikten wird eine ConcurrencyException geworfen.

Consumer Groups

Das Modul unterstützt Consumer Groups für parallele Event-Verarbeitung:

// Consumer 1
eventConsumer.consumeEvents(
    consumerName = "consumer-1"
) { event ->
    // Verarbeite Event
    processEvent(event)
}

// Consumer 2 (in der gleichen Consumer Group)
eventConsumer.consumeEvents(
    consumerName = "consumer-2"
) { event ->
    // Verarbeite Event parallel
    processEvent(event)
}

Events werden automatisch auf verfügbare Consumer verteilt.

Event Replay

Sie können Events von einem bestimmten Zeitpunkt oder Message-ID replaying:

// Replay alle Events eines Aggregates
val events = eventStore.loadEvents(aggregateId)

// Replay Events ab einer bestimmten Version
val eventsFromVersion = eventStore.loadEvents(
    aggregateId = aggregateId,
    fromVersion = EventVersion.of(10)
)

Serialisierung

Das Modul verwendet Jackson für Event-Serialisierung:

  • Automatische Kotlin-Modul Integration
  • Polymorphe Serialisierung für verschiedene Event-Typen
  • Custom Serializer können via @Bean überschrieben werden

Performance

  • Connection Pooling: Wiederverwendbare Verbindungen via Lettuce
  • Batch Processing: Konfigurierbare Batch-Größe für Consumer
  • Non-blocking I/O: Reaktive Operations mit Kotlin Coroutines
  • Stream-basiert: Effiziente Event-Speicherung mit Redis Streams

Troubleshooting

Redis Verbindungsfehler

RedisConnectionFailureException: Unable to connect to Redis

Lösung: Überprüfen Sie Redis-Server und Netzwerk-Konfiguration. Stellen Sie sicher, dass Redis Streams unterstützt werden (Redis 5.0+).

Concurrency Exception

ConcurrencyException: Expected version X but found Y

Lösung: Dies ist normales Verhalten bei Optimistic Locking. Implementieren Sie Retry-Logik oder verwenden Sie EventVersion.ANY.

Consumer Group Fehler

Consumer group already exists

Lösung: Setzen Sie createConsumerGroupIfNotExists: true in der Konfiguration oder löschen Sie die Consumer Group manuell.

Bean-Konflikte mit Cache

Wenn Sie Fehler wie "Multiple beans of type RedisConnectionFactory" erhalten:

Lösung: Die Module verwenden bereits unterschiedliche Bean-Namen mit @Qualifier. Stellen Sie sicher, dass Sie beide Module korrekt konfiguriert haben (siehe Abschnitt "Gleichzeitige Verwendung").

Best Practices

  1. Separate Databases: Verwenden Sie immer separate Redis Databases für Cache und Event Store
  2. Event Versioning: Verwenden Sie Optimistic Locking für kritische Aggregates
  3. Consumer Groups: Nutzen Sie Consumer Groups für horizontale Skalierung
  4. Error Handling: Implementieren Sie Retry-Logik für transiente Fehler
  5. Monitoring: Überwachen Sie Stream-Größen und Consumer Lag

Weitere Informationen