Even with a sour economy, power efficiency is still a hot-button issue among storage buyers. In fact, recent research shows that a sizable chunk of customers are willing to spend money on IT solutions that are power efficient; for example, the spring 2009 Purchasing Intentions survey conducted by SearchStorage.com showed that while only 3.5% of 566 storage buyer respondents said energy efficiency was the most important factor in the choice of a disk array, about 32% said it was a major factor in the decision. But even customers who are already sold on the importance of power efficiency will need help to actually realize those power savings, and that's where you come in. So how do you get them from Point A (a data center full of power-hungry hardware) to Point B (a much less ravenous room of equipment)? And can it be done on a range of storage devices: backup, archive and primary storage hardware?
First, keep in mind this simple principle: If power-managed disks in a storage system are not being accessed for a period of time, they either spin down or power down, saving power and cooling costs. The challenge, however, is that the data on these systems has to be inactive long enough that the drives will actually spin down. "Long enough" is a user-settable time period, often minutes.
Backup, archive power efficiency
In a backup-to-disk scenario, it's almost a no-brainer: Power-managed drives will be inactive long enough that they could go into a power saving mode. Assuming a backup window of eight to 10 hours and minimal restore activity, a backup-to-disk system should be able to power down for four to six hours per day, adding up to a measurable power savings.
Beyond backup, there's power savings to be had on hardware used for archiving. While some in the industry argue that archive data doesn't belong on disk, partly owing to the energy consumption of those disks (which compares unfavorably with tape media), it's relatively simple to decrease the power consumption of disk archives in power-managed drives by powering them down while idle.
And since the purpose of an archive is to store a customer's less frequently accessed data, there should be plenty of idle time. To maximize the amount of idle time on a disk archive, solution providers should migrate data to the archive just once a week. More importantly, you should put tools in place (from companies such as Tek-Tools and Aptare) to monitor data being accessed from the archive to determine whether there's data that actually belongs on primary storage rather than on the archive disk.
Occasional access of a file from archive is fine, but if a file shows up as very active, a de-migration back to primary storage might make more sense. By weeding out the files that are being frequently accessed from archive, you can increase the amount of idle time on the archive and thereby increase the power savings.
Power efficiency on primary storage
In addition to backup and archive scenarios, you can also help customers decrease their power usage on primary storage. For instance, while it will take some skill on your part, the layout of physical drives in array groups can be designed for optimal use of the power saving capabilities of these drives.
Primary storage will need to be monitored to assess access frequencies of data, with an eye toward grouping data with similar access frequencies together. That monitoring can be done using the tools mentioned above.
When monitoring primary storage to find good candidates for power savings, look for folders and volumes that are not accessed for a few hours, especially overnight. An example might be user home directories. After the users go home for the night and the backups of those systems are complete, they may sit idle for hours at a time. Moving the volume that contains the home directories from an array group with active drives to an array group that is on top of power-managed drives could save quite a bit of power. Identifying these types of poor associations between volumes and drives requires tools and storage experience but could save customers a significant amount of money because of reduced power consumption. And it allows the VAR to shine.
This leads to another way to fine-tune the power efficiency of your customer's environment: arranging the backup schedule so that power-managed areas of storage are backed up first at the end of the day, right after users go home, so those drives can go into idle mode and not be woken by the backup job.
An alternative way to achieve power efficiency is using a method that we describe in our article "Archiving in Place." If you provide only power-managed arrays to your clients when they buy new equipment, as they add new arrays to their environment, only move the most active data to the new environment. Instead of archiving the oldest 80%, promote the newest 20% to the new array. Then, over time, as the older data on the older array becomes accessed less frequently, the system begins to power down on its own eventually becoming extremely power-efficient.
While spin-down drives can have a big impact on a customer's power usage, to have the highest level of power efficiency possible, your customers need to re-architect what data lives where within the data center, involving changes around backup, archive and primary storage.
About the author
George Crump is president and founder of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. With 25 years of experience designing storage solutions for data centers across the United States, he has seen the birth of such technologies as RAID, NAS and SAN. Prior to founding Storage Switzerland, George was chief technology officer at one of the nation's largest storage integrators, where he was in charge of technology testing, integration and product selection. Find Storage Switzerland's disclosure statement here.
