The technology is relatively simple. After a period of inactivity the drives in an array slow down and are then put into a mode where they stop spinning altogether. Each of these modes have an increase in power savings but with them comes additional delay in response time. The delay can range from sub-second to a few seconds. Drive and array vendors will offer different levels of power savings. Nexsan for example has levels ranging from unloading the heads, to slowing the RPMs of the drive down, to putting the drives into a sleep mode.
For the benefits of power efficient drives to be realized they must be allowed to not be accessed long enough for the spin down technologies to engage. That means your data needs to be organized in such a way that your very active data set is not sharing a LUN with the potentially inactive data set. This can be particularly difficult to design if you are creating a LUN with drives from separate shelves.
If you are using a multi-shelf system and want to mix drive types, make sure your vendor supports it and be aware that the only power efficiency you will derive is on those drives. The shelf itself will need to continue to run at full power. Single shelf systems or systems that only provide power efficient storage are easier to manage and with today's available drive capacities offer considerable scalability.
Designing a separate LUN or LUNs to receive mostly idle data is the first step. The next is to analyze that data and find data that is infrequently accessed and then to group that data together. This can be done with tools like those offered by Tek-Tools or APTARE. Folders can be monitored for access frequency, those that aren't accessed daily are ideal candidates to move to the power efficient LUN. Another technique that we discuss in our article "Archiving in Place" is to, when purchasing a new array, route newly created data to that array and let the older array idle off as access to it becomes less frequent.
Power efficiency is not just for really old data that has not been accessed in the last six months. It is for data that is being accessed infrequently, every other day for example. It should even be considered for data that is not being accessed overnight, user home directories as another example. While we often talk about archive data being about 80% of the data in the environment, infrequent data may be as much as 99% of the data in the enterprise. Imagine the power savings if 99% of your data was on power efficient storage.
Track us on Twitter: http://twitter.com/storageswiss
Subscribe to our RSS feed.
George Crump is founder of Storage Switzerland, an analyst firm focused on the virtualization and storage marketplaces. It provides strategic consulting and analysis to storage users, suppliers, and integrators. An industry veteran of more than 25 years, Crump has held engineering and sales positions at various IT industry manufacturers and integrators. Prior to Storage Switzerland, he was CTO at one of the nation's largest integrators.