I Found 487 GB Of Data Nobody Had Looked At In Three Years. Deleting It Felt Like Archaeology.

2026-04-02 · Source: Data Science on Medium · Field: Technology & Digital — Data Science & Analytics, Cloud Computing & IT Infrastructure · Depth: Intermediate, quick

Summary

A storage audit on an analytics database revealed 487 GB of unused data, leading to a significant cost reduction. The audit was initiated after the CFO questioned why the company was paying for 2.1 TB of storage when only 800 GB was reported as "active." The largest unreferenced table, "user_activity_log," accounted for nearly 25% of the total storage. This table, which tracked user clicks for a discontinued product analytics experiment, had not been updated since February 14, 2023, and was not being queried by any active processes. Deleting this dormant data, after verifying its inactivity and creating backups, resulted in a monthly hosting bill reduction of $340.

Key takeaway

For Data Engineers managing cloud database costs, regularly auditing storage for dormant data is crucial. Identifying and safely removing unused tables, like the 487 GB "user_activity_log" in this case, can lead to direct savings, such as the $340 monthly reduction achieved here. Implement a routine to query table sizes and last access dates to prevent unnecessary expenditure.

Key insights

Regular storage audits can identify dormant data, reducing costs and improving database efficiency.

Principles

• Unused data incurs ongoing storage costs.
• Data lifecycle management prevents data sprawl.

Method

Audit storage by querying table sizes, identify unreferenced or inactive tables, verify last update and query activity, then delete with backups.

In practice

• Query `information_schema.tables` for disk usage.
• Check `created_at` for last row insertion.
• Verify no active queries before deletion.

Topics

• Storage Audit
• Database Management
• Data Lifecycle Management
• Cost Optimization
• Unused Data

Best for: Data Engineer, DevOps Engineer, IT Professional

Related on AIssential

• 14,000 Files, 40GB: Fixing a Slow Delta Table in Azure Synapse — Excessive small files in Delta tables significantly degrade Spark query performance by increasing transaction log processing time.
• We Spent $2 Million Leaving AWS. Then We Spent $3 Million Coming Back. — Repatriating cloud infrastructure can lead to significant, unforeseen costs, often exceeding initial cloud expenses.
• I Deleted 4TB of Production Data by Trusting dbt’s --full-refresh Flag — Misunderstanding a command's full implications can lead to catastrophic data loss, especially without validated backups.
• Protect Performance and Reduce Surprise Costs with Default Warehouse — Default Warehouse automates SQL warehouse selection for ad hoc queries, optimizing cost and performance.
• AI, MCP, and the Hidden Costs of Data Hoarding — Unchecked use of Model Context Protocol (MCP) leads to "data hoarding," incurring hidden costs and hindering developer skill growth.
• Detecting Join Duplication — Silent join duplication corrupts data pipelines; systematic auditing of join keys, row counts, and match coverage is essential.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Data Science on Medium.