AWS dropped powerful updates to S3 Storage Lens in late 2025 – full prefix visibility, new performance metrics, and direct export to S3 Tables. Here's what changed and how to start using it today.
Canonical_url
🚀 Big News for S3 Power Users: Amazon S3 Storage Lens Just Got Supercharged!
If you're managing petabytes (or even exabytes) of data in Amazon S3, you've probably used S3 Storage Lens to get organization-wide visibility into usage, costs, and activity. But until recently, it had some limits that could feel frustrating when your buckets grew massive or your apps started showing weird performance hiccups.
AWS dropped a game-changing update in December 2025 — and it's now rolling strong across most regions. Here are the three headline upgrades that make S3 Storage Lens way more powerful:
1. Performance Metrics That Actually Tell You Why Things Are Slow
Tired of guessing why your application latency spikes or why certain workloads feel throttled? S3 Storage Lens now delivers eight new performance metric categories (available at org, account, bucket, and prefix levels).
Key ones include:
- Read/Write request sizes (spot those tiny-object nightmares)
- Concurrent PUT 503 errors (hello, throttling detection!)
- Cross-Region data transfer volumes
- Unique objects accessed
- First-byte and total request latency
These metrics help you pinpoint real bottlenecks — like inefficient access patterns or unnecessary cross-region hops — so you can tune your apps and save real money on performance-related surprises.
2. Billions of Prefixes? No Problem Anymore
Before: Prefix metrics were capped — only the "top" prefixes (largest 1%, up to 10 levels deep) got detailed reporting.
Now: The new Expanded Prefixes Metrics Report gives you visibility across every single prefix in your bucket — even if we're talking billions of prefixes per bucket. No more size thresholds or depth limits (up to 50 levels deep now supported in some contexts).
You get full storage usage, activity, data protection, and status-code metrics for your entire prefix tree. Perfect for data lakes, ML training datasets, log archives, or any workload with deep, sprawling folder structures.
(Pro tip: When you enable this, give it a day or two for the full prefix crawl to complete – it's massive!)
3. Export Straight to S3 Tables — Query Like a Pro, No ETL Hassle
The killer feature for analysts and data teams: Metrics can now auto-export daily directly into Amazon S3 Tables (AWS-managed, Apache Iceberg-backed tables).
No more building pipelines or transforming CSVs/Parquets yourself. Just point your Storage Lens dashboard to an S3 Table bucket, and the data lands ready-to-query in tables like expanded_prefixes_activity_metrics.
Fire up Amazon Athena, QuickSight, EMR, Redshift — or even bring your own tools — and start asking questions like:
- "Which buckets grew the most last month?"
- "Show me storage costs by class across all accounts"
- "Where are we hitting the most 503s on PUTs?"
Bonus: This opens the door to agentic AI workflows with natural language queries via S3 Tables integrations.
Why This Matters for DevOps & Cloud Engineers
Whether you're hunting cost leaks, debugging slow data pipelines, ensuring compliance across millions of prefixes, or just trying to keep your apps snappy — these updates turn S3 Storage Lens from a "nice-to-have" dashboard into a must-use observability powerhouse.
Quick Start Tip
Head to the S3 console → Storage Lens → Create/Edit dashboard:
- Opt into Advanced tier
- Enable performance metrics
- Turn on Expanded prefixes report
- Select export to S3 Tables (or your own bucket for CSV/Parquet if preferred)
Give it 24 hours for daily metrics to flow in, then start exploring.
Have you already played with the new performance metrics or S3 Tables export? Drop your favorite insight — or any weird finding — in the comments below. Curious to hear real-world wins (or "aha!" moments) from the field.
Your S3 just got smarter. Time to put these superpowers to work! 💪
Support if you found this helpful😉
No Money 🙅🏻♀️ just Subscribe to me YouTube channel.
Linktree Profile: https://linktr.ee/DevOps_Descent
GitHub: https://github.com/devopsdescent

Top comments (0)