File Systems and File Management

Jump straight to an abstract:

The Abstracts

The File Systems Evolution
Christian Bandulet

File Systems impose structure on the address space of one or more physical or virtual devices.  Starting with local file systems over time additional file systems appeared focusing on specialized requirements such as data sharing, remote file access, distributed file access, parallel files access, HPC, archiving, security etc.. Due to the dramatic growth of unstructured data files as the basic units for data containers are morphing into file objects providing more semantics and feature-rich capabilities for content processing.    This presentation will categorize and explain the basic principles of currently available file systems (e.g. local FS, shared FS, SAN FS, clustered FS, network FS, WAFS, distributed FS, parallel FS, object FS, ...). It will also explain technologies like NAS aggregation, NAS clustering, scalable NFS, global namespace, parallel NFS, storage grids and cloud storage.     All of these files system categories are complementary. They will be enhanced in parallel with additional value added functionality. New file system architectures will be developed and some of them will be blended in the future

Learning Objectives

• Understand the basic principles of different files system architectures
• Understand how file systems evolved over time
• Being prepared to discuss, position and recommend the most appropriate file system for a customer solution

Scale-Out Storage Systems
Nicholas Kirsch

This is a overview of scale-out storage systems and their underlying file system technologies - primarily focused on network-attached storage systems. In this presentation, scale-out will be defined in contrast with scale-up storage systems, the market, user, and technology needs driving new class of storage systems will be explained, as well as a survey of open-source and commercial implementations available today.

Learning Objectives

• What is a Scale-Out Storage system and how does it differ from a "traditional" storage system?
• How do different implementations of Scale-Out technologies differ - what are the pros/cons of the various design choices?

File Systems for Object Storage Devices
Paul Massiglia, Tushar Tambay

Object-based storage devices (OSDs) may well be the “next big thing” in file-oriented data storage. Already popular in the high-performance computing arena, they are poised to enter more more general enterprise computing environments. By distributing storage management and enabling secure data transfer between storage devices and clients, OSDs promise significant improvements in scaling and administrative simplicity. But making effective use of OSDs requires a new breed of file system—one that  makes use of the new devices effectively to deliver the promised benefits.     This tutorial will describe the salient properties of OSDs, explain how file system technology is evolving to exploit the scalability and administrative simplicity they offer, identify the mature and emerging segments of the OSD-based file storage market, and show how technology that has been successful in HPC can be beneficially employed in the general data center environment. Standardization activities, notably the parallel NFS (pNFS) protocol for addressing OSDs will be discussed.

Learning Objectives

• Appreciate the advantages of OSDs over conventional disk storage and the environments in which those advantages are meaningful.
• Understand how OSD file systems differ from disk-based file systems.
• Understand the current state of OSD file system maturity and the segments in which they are deployed, and learn how OSD technology and standardization are developing to broaden the appeal of OSDs throughout enterprise computing.

Storage Tiering and the Impact of Flash on File Systems
Jonathan Goldick

With all of the talk about how storage systems will be impacted by large amounts of relatively inexpensive Flash little has been said about how file systems will need to change to take advantage of it.   This tutorial will cover how file systems are evolving tiered architectures to leverage Flash.

Learning Objectives

• Understanding how Flash will impact file systems.
• Characteristics of Flash that are relevant to file systems.
• How tiering works in a file system and whether it’s better handled at the array.

File Systems
Dominic Kay, Mark Maybee

Deduplication can be accomplished in different ways in a file system.  This tutorial will focus on block-level deduplication. While conceptually simple, an implementation can be quite complex as it must address multiple issues:    scalability - when the lookup table no longer fits in memory.  performance - impact of table lookups and writes dependent on reads.  space accounting - space now be shared between files and file systems.  administrative model - keeping model simple.    We will talk about these issues in detail.  This tutorial will also cover expanding the notion of deduplication beyond the storage device to include in-memory and over-the-wire deduplication.

Learning Objectives

• Handling scalability in deduplication.
• Understanding performance issues associated with deduplication.
• Dealing with space accounting with deduplication.