AWS Details How One Customer Scaled to One Million Lambda Functions

2026-07-09 · Source: InfoQ · Field: Technology & Digital — Cloud Computing & IT Infrastructure, Software Development & Engineering · Depth: Intermediate, quick

Summary

AWS has outlined how ProGlove, an industrial-wearables manufacturer, was able to scale its SaaS platform to run more than one million AWS Lambda functions spread across thousands of dedicated customer accounts. The company achieved this by implementing a one-account-per-tenant model, extensive automation with AWS Organizations, Step Functions, and CloudFormation StackSets, and aggressive scale-to-zero policies, which kept idle costs below one US dollar per month per account. ProGlove addressed initial operational friction beyond 50 accounts and "self-inflicted DDoS" from rigid cron-style schedules by adopting jittered execution windows and event-driven triggers. Observability costs were managed by consolidating high-priority failures into a central dead-letter queue and removing unused Amazon SQS queues. While automation introduced compromises like slower StackSet rollouts, these were deemed acceptable for improved isolation, cost visibility, and operational consistency.

Key takeaway

For DevOps Engineers scaling multi-tenant serverless platforms, prioritize tenant isolation with dedicated AWS accounts from the outset. You should aggressively automate infrastructure provisioning using tools like CloudFormation StackSets and implement scale-to-zero policies to control costs. Proactively manage observability spend and replace rigid cron schedules with jittered, event-driven triggers to prevent self-inflicted DDoS and maintain operational consistency as you grow.

Key insights

ProGlove scaled to one million Lambda functions by isolating tenants, automating infrastructure, and optimizing observability costs.

Principles

• Isolate tenants for stronger security and cost attribution.
• Automate infrastructure actions early and extensively.
• Treat observability spend as a first-class scaling constraint.

Method

ProGlove used AWS Organizations, Step Functions, and CloudFormation StackSets to create and update thousands of customer accounts from a single pipeline. They replaced rigid timers with jittered execution and event-driven triggers.

In practice

• Implement a one-account-per-tenant model.
• Use CloudFormation StackSets for multi-account deployment.

Topics

• AWS Lambda
• Serverless Architecture
• Multi-tenant Systems
• CloudFormation StackSets
• AWS Organizations
• Observability Costs
• Infrastructure Automation

Best for: AI Architect, MLOps Engineer, Entrepreneur, DevOps Engineer, Software Engineer, IT Professional

Related on AIssential

• Building a fault-tolerant metrics storage system at Airbnb — Scaling observability requires robust multi-tenant isolation, automated management, and strategic multi-cluster deployment to ensure reliability.
• Scaling on Demand: How Serverless Architectures Empower Agile and High-Performing Platforms — Serverless architecture offers macOS developers scalable, cost-efficient, and maintenance-free application development.
• Building Blocks for Foundation Model Training and Inference on AWS — Foundation model scaling now requires integrated infrastructure for pre-training, post-training, and inference, emphasizing compute, network, and storage.
• Kubernetes Autoscaling Demands New Observability Focus Beyond Vendor Tooling — Modern Kubernetes autoscaling requires observability focused on provisioning intelligence, not just static infrastructure health.
• EP196: Cloud Load Balancer Cheat Sheet — Cloud-native architectures enhance scalability and efficiency through specialized components like load balancers, CQRS, Docker, and serverless functions.
• I Built a 6-Service AI Document Pipeline on AWS for Under $5/Month…Here’s the Architecture — A multi-modal AI pipeline can automate complex customer support triage cost-effectively using serverless AWS services.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by InfoQ.