StorageService

Kind: Service

Source: atloria-monorepo/apps/screenshot-worker/src/app/storage/storage.service.ts

Azure Blob Storage service for storing screenshots

Features:

• Upload screenshots to Azure Blob Storage
• Generate public URLs
• Support for metadata
• Auto-create containers

StorageService is a NestJS backend service responsible for persisting screenshot files into Azure Blob Storage and returning accessible URLs for later retrieval. It encapsulates container management (including auto-creation), upload operations, and optional blob metadata handling so the screenshot-worker can store outputs reliably without duplicating storage logic.

Methods

Dependencies

• ConfigService

Where it refuses work

• StorageService stops the work with Error when !this.containerClient — “Azure Blob Storage not initialized”, in 2 places.

When something fails

• StorageService handles failure in 4 places: it logs it and continues in 2, and lets it reach the caller in 2.

Diagram

mermaid
sequenceDiagram
  autonumber
  participant Worker as Screenshot Worker
  participant Storage as StorageService
  participant Azure as Azure Blob Storage

  Worker->>Storage: uploadScreenshot(buffer, filename, container, metadata?)
  Storage->>Azure: ensureContainerExists(container)
  Azure-->>Storage: container ready
  Storage->>Azure: uploadBlob(container, filename, buffer, metadata)
  Azure-->>Storage: upload result (etag/version)
  Storage->>Azure: generatePublicUrl(container, filename)
  Azure-->>Storage: public URL
  Storage-->>Worker: { url, blobName, container, etag? }

Usage

ts
import { Injectable } from '@nestjs/common';
import { StorageService } from './storage/storage.service';

@Injectable()
export class ScreenshotJobService {
  constructor(private readonly storage: StorageService) {}

  async persistScreenshot(pngBuffer: Buffer, jobId: string) {
    const container = 'screenshots';
    const blobName = `jobs/${jobId}.png`;

    const result = await this.storage.uploadScreenshot(pngBuffer, blobName, {
      container,
      contentType: 'image/png',
      metadata: {
        jobId,
        source: 'screenshot-worker',
      },
    });

    // Store result.url in DB / emit event, etc.
    return result.url;
  }
}

AI Coding Instructions

• Keep Azure-specific logic (container creation, blob client setup, URL generation) inside StorageService; call it from jobs/workers/controllers rather than re-implementing storage concerns.
• Always pass a deterministic blobName (include job IDs / timestamps) to avoid accidental overwrites; if overwrites are intended, document it explicitly.
• When attaching metadata, ensure values are strings and keep keys stable (Azure metadata constraints); avoid putting sensitive data in metadata since it can be exposed.
• Handle container creation idempotently (create-if-not-exists) and surface meaningful errors (auth, missing connection string, permissions) to the caller for retry/backoff logic.
• Prefer returning the generated public URL from the service (single source of truth for URL format/SAS/public access strategy) and avoid hardcoding URL templates elsewhere.

Relationships

• DEPENDS_ON → configservice

Referenced By

• AppService (DEPENDS_ON)
• ScreenshotService (DEPENDS_ON)
• BatchScreenshotService (DEPENDS_ON)
• FlowReplayService (DEPENDS_ON)
• StorageModule (MODULE_PROVIDES)
• StorageModule (MODULE_EXPORTS)