Carl D'Halluin

"How hard can it be to copy files and objects from one storage system to another? I could write a script myself or use one of the free tools out there. Easy, right?" 

That was my thinking too until I stepped into the world of enterprise data mobility and encountered its harsh realities. In this talk, we’ll explore 15 years of lessons learned from the trenches of real-world data migration, replication and archival. We’ll dive into the technical challenges that consistently derail projects: filename encoding issues, unreadable or garbled data, metadata problems, incompatible or un(der)documented protocols, inconsistent timestamp precision, size and scale limitations, illegal path names, incompatible retention mechanisms, and plenty of performance problems and plain old bugs. 

For application developers, we’ll cover how to design software that plays nicely with a broad range of storage systems ensuring compatibility, avoiding user pain, and building resilience into your code. For storage system engineers, we’ll explore how to design backends that operate seamlessly in multi-vendor environments, support broad use cases, and anticipate the quirks of real-world client behavior. For storage administrators, we’ll show how to future-proof your storage environment and prevent vendor lock-in. Join us as we navigate the minefield of NFS, the quagmire of SMB, and the abyss of S3, and discover why "just copying data" is one of the most misunderstood challenges in modern infrastructure.

Carl D'Halluin

CTO

Datadobi