Beyond Copy and Paste: Data Migration Challenges at Scale
During September 2025, the SNIA Developer Conference brought storage and data infrastructure experts together in Santa Clara. Datadobi CTO Carl D’Halluin and VP of Product Marketing Steve Leeper took the stage to present War Stories from the Storage Trenches: Moving Data Across NFS, SMB, and S3, a session focused on what really happens when enterprise data moves from one system to another.
Their goal was simple: challenge the assumption that migration is just copy and paste.
Terms like “migrate” and “copy” are often used casually, but the work behind them can vary dramatically depending on system architecture, protocol behavior, and implementation details. In some cases, a move may be little more than a metadata update. In others, it can involve transferring content, ownership, permissions, and other critical attributes between very different platforms.
At enterprise scale, these data migration challenges multiply quickly.
5 Data Migration Challenges That Hide Beneath the Surface
The SNIA session highlighted several issues that can turn an apparently simple migration into a complex engineering exercise:
1. Files change while the migration is underway.
Enterprise data is rarely static. Users edit files, background tasks update content, and applications continue creating new data while the migration is in progress. Multiple incremental passes may be needed to capture those changes and keep the destination consistent with the source.
2. Protocols and platforms behave differently.
Standards such as SMB and NFS are not always implemented identically across vendors. Even subtle differences can create incompatibilities. In multi-protocol environments where S3, NFS, SMB, and proprietary APIs coexist, additional translation can add complexity and latency.
3. Character encoding and naming can cause failures.
Non-Latin character sets such as Japanese, Cyrillic, and Arabic can expose encoding mismatches between systems. The result may be unreadable filenames, failed transfers, or data that appears to have moved successfully but is no longer represented correctly.
4. Path and filename limits vary.
Different platforms may support different maximum filename lengths, path depths, or total path lengths. Deeply nested directories and long naming conventions can create truncation issues, break links, or result in lost metadata when the target cannot reproduce the source structure exactly.
5. Permissions do not always translate cleanly.
Permission and access models can vary significantly between platforms and protocols. Without careful mapping, users may lose access to migrated data or gain access they should not have.
Why Data Migration Validation Matters
Moving the bytes is only part of the job. Reliable migration also requires orchestration, error handling, and data migration validation to confirm that the destination contains complete, usable, and appropriately accessible data.
Validation helps teams identify where source and target behavior differs, distinguish expected transformations from actual failures, and address exceptions before cutover. A migration is not successful simply because a copy job finishes; the result has to work as intended.
Bridging the Gap Between Expectation and Reality
The session also shared a real-world example in which a stakeholder assumed enterprise migration could be handled with a simple drag-and-drop operation. The story highlighted a common disconnect between how straightforward migration can appear and the engineering work required to make it reliable.
For technical teams, the lesson is clear: successful migration depends on anticipating edge cases, understanding platform-specific behavior, and using tools designed for the realities of large-scale data movement.
For stakeholders, education matters too. Aligning expectations around complexity, testing, validation, and cutover helps everyone understand what success actually requires.
Beyond Copy and Paste
Enterprise migration is ultimately about preserving data integrity, accessibility, permissions, and context as information moves between systems that may behave very differently. That is why the most important data migration challenges often sit beneath the surface. Addressing them early, and validating the result, can make the difference between a copy that completes and a migration that succeeds.
War Stories from the Storage Trenches: Moving Data Across NFS, SMB, and S3
Watch the presentation to explore the real-world complexities that can emerge when moving data across NFS, SMB, S3, and heterogeneous storage environments.