Processor ® Free Subscription
Used HP, Used IBM, Used Compaq, Used Cisco, Used Sun
Home |  Register |  Contact Us   
This Week's Issue
Browse All Issues
Search All Articles
Product News & Information
Company
News & Information
General Feature Articles
News
Opinions



Tech & Trends Email This
Print This
View My Personal Library

General Information Add To My Personal Library
December 19, 2008 • Vol.30 Issue 51
Page(s) 22 in print issue

Power-Efficient Storage Systems
Reduce The Power & Energy Used By Your Storage Equipment
Storage systems consume as much as 24 to 40% of total IT power usage and have become a primary target in green IT initiatives. The challenge is how to create a more power-efficient storage infrastructure while dealing with explosive growth in that storage.

Unlike other green IT initiatives, storage is unique; the classic problem is that 80% of the data has not been accessed in the past 90 days and will likely not be accessed again. However, retaining that data is still important for legal and corporate governance reasons.

To respond, most major hard drive manufacturers have released “green drives.” These drives typically save power by spinning down or going into a standby mode. Simply implementing power-efficient drives is not enough—if the data is not placed on the drives in an efficient manner to make sure that old data is not mixed with active data, the drives will never have the opportunity to spin down or will spin up and spin down so frequently that they actually consume more power than the drives they replaced.

Identify & Migrate Inactive Data

Migration of inactive data off of primary storage and onto power-managed secondary storage is the logical first step in creating a power-efficient storage infrastructure. An alternative is to migrate this data out of the data center altogether and leverage one of the cloud storage service providers.

The key ingredient to migrating the inactive data set is identifying the data that has not been accessed that is eligible for migration. Another decision point is how to accomplish access to the old data. Some tools identify and move the data; other tools add the ability to transparently access that data.

The final step in the migration phase is to implement a process by which you will monitor this data set. “A real-time, heads-up display is critical for taking action on data the moment data becomes eligible for migration,” says Stephen Harding, director of marketing at Tek-Tools Software (www.tek-tools.com).

When migrating data for power efficiency, the goal is different compared to the traditional ILM models of the late ’90s. The ILM model typically migrated data as the capacity of the volume filled; with power efficiency, the goal is to free up storage to the point that primary storage shelves can be physically turned off. According to Harding, “to accomplish this requires a more real-time interaction with the data.”

Select Secondary Storage

After determining how data will be identified and migrated, the selection of secondary storage is made. The first step in this phase is to determine where to store the data. You must choose to either store the data in-house or utilize a cloud storage provider. The next step is to determine how to store the old data.

“Leveraging cloud-based storage for an archive delivers two immediate benefits that intersect well with the need to reduce power consumption and the need to deal with flat or shrinking IT budgets,” says Adrian Herrera, senior director of marketing at Nirvanix (www.nirvanix.com), a cloud storage provider. The advantage of a cloud-based storage solution is that there is no on-premise equipment to consume power. Also, it is a pay-as-you-grow purchasing model and, as a result, defrays the large initial acquisition costs that an on-premise solution may have.

Moving data to and from the cloud can be done by utilizing a NAS-style gateway, an on-premise software application that makes any available server into a cloud gateway, presenting the cloud as a standard file system. The second option for migration is to have the application access the cloud directly via the cloud provider’s API set.

Further considerations related to power, storage, and even replication are the responsibility of the cloud storage provider.

For those IT professionals who would rather keep data under their roofs or are concerned with the long-term cost of cloud storage, a local power-optimized secondary storage platform may be more appropriate.

When looking for a on-premise power-managed second tier, Chris Santilli, CTO at Copan Systems (www.copansystems.com), suggests that, “A power-efficient archive needs more than just green drives; it needs to be a complete solution that can not only power down or power off drives as required, but also it needs to have the intelligence to place data on the system to maximize the power efficiency.” In other words, you want to make sure drives that are powered off stay in that state for as long as possible.

“Proper power management and data placement allows for the very dense packing of the physical drives, optimizing costs and improving floor space,” continues Santilli. This dense packing is accomplished by alternatively mounting the drives to reduce vibration and to make sure that only 25% of the drives are powered on at any given point in time.

Lastly, data integrity is often a concern in power-managed systems. The fear of the drive actually turning back on is often raised as an objection. This can be addressed by periodically powering the drives on and exercising them by validating the data on those drives. “As a result of powering the drives off and with proper periodic data validation, the drive service life can actually increase beyond the expected norm,” concludes Santilli.

Focus On Primary Storage

With data moved to secondary storage, either at an offsite archive or a power-managed internal storage system, the focus needs to return to the primary storage and how you will demonstrate power reduction. It can be justified by simply showing the delay in future primary storage, which would obviously require budget dollars and power.

In many cases, so much space has been saved off of primary storage that terabytes of free space are available. The challenge is that this old data is not sequential but spread out in varying segments across drive shelves. Restriping arrays across fewer shelves and powering off those shelves can have further reclamation of power. This can be a time-consuming task, but it may be well worth it if three or more shelves can be powered off.

Achieving power-efficient storage is challenging, but by identifying and migrating old data to power-efficient secondary storage and then reorganizing the remaining active data, significant power savings can be had.

by George Crump


Key Steps To Power-Efficient Storage

• Identify what data you have

• Migrate to power-efficient secondary storage

• Restructure primary storage to power down open shelves



After Migration, Green Steps

• Leverage inline compression to reduce active storage

• Leverage deduplication to reduce inactive but not migrated data set

• Look for thin provisioning and intelligent data movement when purchasing new storage


Share This Article:    del.icio.us: Power-Efficient Storage Systems     digg: Power-Efficient Storage Systems     reddit: Power-Efficient Storage Systems

 

Home     Copyright & Legal Notice     Privacy Policy     Site Map     Contact Us

Search results delivered by the Troika® system.

Copyright © by Sandhills Publishing Company 2010. All rights reserved.