Shared Data Stored on Network Attached Storage (NAS)

Introduction

I have worked in shops where there are a significant number of high end technical workstations deployed throughout the enterprise all needing to share the same application related data. Most open systems shops seem to only service one, possibly two platforms and all the related data is only accessed by a single host. Fifteen years ago NAS appliances became of age and provided for better ways for managing data apart from just consuming it and dealing with both the logical and physical limits that localized block storage traditionally provided.

Advantages of NAS:

• Centralized data management consumed by all hosts on the network.
• Backups can be offloaded onto other hosts than where the data is being consumed.
• Strategic component for supporting a disaster recovery strategy.

Disadvantages of NAS:

• Potential for less performance. This has been highly debated. There are trade-offs here in that you at the end of the day, you have a pipe and multiple layers between physical storage and the presentation of a file system in some form. An example from the past that is applicable here is that of a sound system. If any component (microphone, cables, sound board, speaker) from the source to the speaker is inferior, so will the sound quality.

• Consumption over a network is configured by a set of rules for what host and which user can access the data. It is harder, though not impossible, to manage where that data is being consumed.

The advantages outweigh the disadvantages in my book. Any potential performance degradation (I’ve actually seen better performance when tuned correctly) is minuscule and is outweighed by the efficient centralized management of the data itself, particularly for backups and disaster recovery. Managing data is a universal issue that takes good architectural design to provide a system for defining responsibility and accountability beyond how the data is consumed.