Data Storage Day

Opening Remarks

Storage Virtualization and the Evolution of Storage Networking

This session looks at the evolution of storage networking from its origins in primitive disk drives and file servers up through elaborate virtualized storage solutions. We explore how new storage functionality comes to market by inserting itself into the appropriate layers in a stack-like architecture called the Storage I/O Path, which originates with an application and ends on a piece of storage media.

The storage I/O path has been evolving quite rapidly in the last few years. New layers are appearing in the form of inline network devices, hypervisors, and object-based storage models. Meanwhile, traditional layers like logical volume managers and RAID arrays are being collapsed into all-in-one file systems. This lecture sets the stage for the remaining 3 talks at LISA Data Storage Day. It provides the foundation for understanding the various ways to integrate solid state storage, data tiering and life-cycle management, and data protection into existing storage environments.

Topics include:

• The storage I/O path
• Storage virtualization
• The evolution of SAN and NAS solutions
• Virtual I/O
• LAN-based storage virtualization and virtual file systems
• Logical volume management
• Principles of object-based storage

Application and File System Acceleration with Solid State Storage

The data storage industry is experiencing a mini renaissance as a flurry of new solid state storage devices come to market. SSDs are now available in a broad spectrum of price, form factor, and performance. It is hard enough for a storage architect to sort out the myriad of choices, but that is nothing compared to the challenge of figuring out how to deploy SSD technology in an affordable and effective way.

This lecture focuses on practical approaches for integrating SSD technology into your existing storage infrastructure. We explore different insertion points in the storage I/O path, and reveal a variety of methods for accelerating databases, email systems, virtualization platforms, content repositories, local file systems and network file systems (NAS).

Topics include:

• Types of flash media: SLC, MCL, eMLC, DRAM
• Packaging options for solid state storage
• Differentiating random from sequential disk I/O patterns
• SQL database acceleration
• MS Exchange acceleration
• File system tiering and acceleration
• Dynamic tiered storage in SAN arrays
• 3rd-party caching appliances
• Pairing solid state storage with deduplication

Lunch Break in the Exhibit Hall

Visit the Vendor Exhibition
Find out more about the exhibition here.

A Crash Course on Object-based Storage: Massive Scalability, Cloud Stores, Deduplication, and Even Tape

The hype surrounding cloud computing has spilled over into the storage industry, inspiring tremendous innovation in the fields of object storage and large scale file management. Conventional storage technologies (enterprise SAN and NAS systems) cannot scale limitlessly, not to mention that they are too rigid to expand and contract with the ebb and flow of the cloud. The cloud demands a whole new class of file systems and storage management approaches. As luck would have it, the problems being tackled in the name of cloud storage are applicable to large scale storage management in private facilities. In other words, so-called "cloud technologies" are relevant regardless of whether one ever plans to store data in a third party cloud.

Object-based storage does not describe any one technology or type of solution. Rather, is a generic term that describes an approach to addressing and manipulating discrete units of storage. Object-based storage is essential for deduplication, self-healing, massive scalability, geographic distribution, caching, tiering, etc. This session explains the key principles of object-based storage and provides examples and technologies based on those principles.

Topics include:

• The limitations of RAID and conventional file systems
• Object addressing versus file and block addressing
• Content-based addressing
• Hashing and hash collisions
• Fundamentals of deduplication: hashing, chunking, indexing
• New implementations of dedupe: primary storage, virtual desktops, rich media
• Tape-enabled file systems and "active archiving"
• Object-level data redundancy
• Erasure-coded data protection

Break

Designing and Optimizing Enterprise Backup Systems

Backup and restore continue to be one of the greatest pain points in the data center, and to add insult to injury most IT organizations now report that they are paying more per terabyte for data protection than they are for primary storage. Before one can define the solution, one needs to get a handle on the problem, so this seminar begins by identifying the technology snags and hidden costs of enterprise data protection.

Even the best of the enterprise software vendors are not able to deliver comprehensive data protection under a single pain of glass. For most organizations, the winning data protection strategy is likely to involve different product types, often from different vendors. The toolbox includes products such as: enterprise backup software, disk backup appliances, deduplication hardware and software, tape systems, archive managers, hypervisor backup software, cloud solutions, application-specific backup software, snapshots inside the storage devices, etc. This seminar reveals ways of identifying the right tool for the job and devising a sustainable and scalable backup strategy.

Topics include:

• Identifying and eliminating common backup bottlenecks
• Scaling out traditional backups
• The role of VTLs versus conventional disk-enabled backup
• Leveraging deduplication (for what it is good for)
• CDP and granular incremental backups
• Hypervisor-enabled backup and replication
• Life cycle management solutions to offset backup costs/complexity
• Active archives and backup strategies for fixed content
• A simple model for comprehensive data protection