Local NAS Storage Best Practices and Policy

Webpage currently being edited......


Network Attached Storage

A storage device that provides file storage and filesharing to other computers over a network. 

PSD Computing has noticed the proliferation of inexpensive SMB and SOHO-class NAS Devices around the Division.  Though these devices can be useful resources when they are properly configured and maintained, we've encountered devices where data was permenantly lost due to a failure to follow best practices.

We strongly encourage you to contact PSD Computing before you purchase a NAS device so we review your needs and possibly provide more appropriate alternatives.   If a local NAS device makes sense for you, we would like provide our expertise in the selection of the device and then be involved in the inital setup of the NAS.

NAS Device Configuration

The appropriate configuration of your device will determined by the features your device supports and by your specific needs.  Some options, like the various available configurations for the drives each have their benefits and liabilities.  It's particularly important to understand the liabilities of a particular configuration and to make sure that you deal with their implications.

RAID Setup

The RAID level describes how data is written across multiple drives.  Each RAID level provides different levels of performance & data protection.  They each have different impacts on how must of the total drive space is available for data storage.

RAID 0 -  Also called "Striping".  Requires at least two drives. 

Pro:  Provides maximum available space and maximum performance.  

Con:  Provides NO data redundancy and increases the chance of data loss due to drive failure.   Data on RAID 0 arrays is alternately written across all the drives in the array so the loss of any one drive results in irrecoverable data corruption across the entire array.

Best Use:  RAID 0 should be used when you need the highest level of performance.  Ideally, it should only be used when you can afford to loose all the data on the array (e.g. data on the array can be easily re-generated or restored from backup).

RAID 1 - Also called "Mirroring".  Data is written to multiple drives.  

Pro:  Each drive in a mirrored set contains the same data so a RAID 1 configuration provides a high level of data protection.  On some devices, reading data from mirrors is also faster because data is read from muliple drives in parallel.  

Con:  Keeping identical data on more than one drive decreases the amount of storage available to you.  A mirror of Two 2TB drives  (4TB total capacity) only provides 2TB of unformatted storage capacity.

Best Use:  RAID 1 should be used when you require high availability and high data protection, and where you can give up at least half of the capacity of your drives.

RAID 5 - Requires at least three drives.  One drive's capacity is lost for data parity.   

Pro:  Provides data partity to increase data protection.  RAID 5 allows for the failure of any one drive without any loss of data.  Overall performance should be better than RAID 1.

Con:  Not available on small devices with 2 drives.  Some loss of capacity (one drive in the array).

Best Use:  RAID 5 should be used when you need good performance with protection against data loss from drive failure.

RAID 6 - Requires at least four drives.  Two drives' capacity is lost for data parity.   

Pro:  Provides more data partity than RAID 5 to increase data protection.  RAID 6 allows for the failure of any Two drive without any loss of data.  Overall performance should be better than RAID 1.

Con:  Not available on small devices with 4 drives.  More loss of capacity than RAID 5 (two drives in the array).

Best Use:  RAID 6 should be used when the most important considerations are availability and protection of your data, and when you can lose the capacity of two drives.

Sharing Protocols

User Authentication

Backups and Cloud Sharing