Lecture Notes on Storage Concepts

Introduction

Speaker: Mike Johnson (MJ), training content creator and DevOps engineer at MinIO

Topics Covered:

• Block Storage
• File Storage
• Object Storage

About the Speaker

• 22 years in IT
• 10 years in cloud administration
• 20 years in sysadmin
• Experience with containers and Kubernetes
• Certifications in AWS, Azure, Docker, and Kubernetes

Key Concepts to Cover

• Block Storage
• File Storage
• Object Storage
• Performance and cost considerations
• Common use cases
• Visual side-by-side comparison

Block Storage Basics

• Treats data as a sequence of fixed block sizes
• Data can be spread across multiple blocks
• Block sizes can be adjusted to optimize performance
• Stored in a way to provide best performance (handled by software)
• Limited metadata capacity (e.g., file name)
• Strong consistency and high structure
  • Data unlikely to corrupt on single device
  • Loss of device leads to data loss
• Retrieved as blocks via iSCSI, fiber, serial, SATA, etc., then reassembled as files

File Storage Basics

• Files stored as whole, accessed in original format
  • Path-based access (e.g., corporate directory structure)
• Limited metadata (create date, modify date, file size)
• Files can be locked to prevent corruption
• Often sits atop block and object storage
• Accessed via SMB or NFS protocols, adding overhead

Object Storage Basics

• Files stored in distributed shards with metadata, object ID, attributes
• Allows for unlimited metadata, enhances search capabilities
• Versioning to retain data integrity and regulatory compliance
• Unlimited scalability by adding back ends
• Accessed over REST API using HTTP
  • Versatile usage across various platforms and applications

Conclusion

• Block, file, and object storage have unique characteristics and use cases
• Consider performance and cost for each type
• Each type uses different access protocols and methods
• Object storage offers flexible scalability and advanced metadata capabilities