Media Pipeline on Backblaze B2: A Production Case

Presigned uploads, virus scanning, thumbnails, and CDN delivery at scale. The team prioritized reversible migrations while migrating on a Nestlancer-style stack—gateway at the edge, domain NestJS services, Prisma on Postgres, async work through RabbitMQ, and media on Backblaze B2.

Where we started

The team inherited user-generated uploads landing on local disk in early Nestlancer iterations. The triggering incident was clear: virus scan and thumbnail jobs blocked API pods during traffic spikes. Leadership funded the refactor when customer-visible latency and support load rose together—not when the diagram looked messy.

Architecture before

• Single deploy artifact coupling unrelated domains
• Synchronous cross-module HTTP with partial timeouts
• Mixed read/write traffic on one database primary
• Ad-hoc file storage complicating virus scan and CDN caching

Migration timeline

Discovery

Mapped user-generated uploads landing on local disk in early Nestlancer iterations and inventoried which routes could move behind the gateway without user-visible changes.

Strangler cutover

Routed new traffic through NestJS handlers while legacy paths drained over 8 weeks.

Reliability hardening

Added Backblaze B2, RabbitMQ, gateway with explicit SLOs owned by a 8-engineer platform squad.

Stabilize and learn

Ran game days for DLQ replay, replica lag failover, and presigned upload expiry edge cases.

Patterns we applied

• Backblaze B2 presigned uploads moved bytes off API pods; workers handled thumbnails.
• RabbitMQ workers absorbed notification and indexing fanout with DLQ replay runbooks.
• API gateway centralized auth, OpenAPI, and aggregation for mobile clients.

Code sketch from the cutover

@Injectable()
export class PostsRepository {
  constructor(
    @Inject('PRISMA_WRITE') private write: PrismaClient,
    @Inject('PRISMA_READ') private read: PrismaClient,
  ) {}

  listPublished() {
    return this.read.post.findMany({ where: { status: 'PUBLISHED' } });
  }
}

Results

• moved to presigned B2 uploads with async workers—API CPU dropped 25%
• On-call pages for queue backlog fell after outbox lag dashboards went live
• Product teams could ship blog and portfolio changes without redeploying payments
• Support tickets citing 'stale listings' dropped once read paths moved to replicas

Retrospective checklist

• Game day DLQ replay documented with ordering notes
• Replica lag runbook tested in staging monthly
• Presigned upload TTL aligned with mobile retry policy
• Gateway error envelope consistent across all domain services
• Post-incident templates link to dashboards—not screenshots

Lesson

Microservices did not arrive on day one. The team earned splits by proving operational ownership per domain—not by copying a reference diagram. The Nestlancer patterns above were adopted only after metrics justified the coordination cost.