Summer is starting and with higher energy prices there might be some incentives for raising the thermostats in your Data Centers. However, one thing we have noticed over the years of supporting hard drives in the field is that higher temperatures cause disks to fail more often. When it comes to cost control, keeping a balance between disk failures and higher temperatures seems to be a question we are asked about often.
There seems to be a correlation between disks that are higher in racks and failures, although it could be a coincidence. Most people have a hazy understanding of their Data Center airflows , other than they notice hot spots and cooler spots within them. But hard drives need to be kept cool, so keeping your shelves of hard drives in the cool spots might help your drives last a bit longer. Unfortunately, moving your cabinet of drives to a formerly cool spot might make it a hot spot.
One thing is certain and that is heat rises. Keeping your data center on a lower floor and moving your servers and drives closer to the floor will keep them slightly cooler. Consolidating your storage islands onto fewer platforms and trying to rationalize your storage onto appropriate platforms for their usage patterns may save you money, buying new systems may look good, but often just consolidating you current storage will save you money by saving energy and extending the lifespan of your systems.
There are no magic solutions to the cost problem of cooling your data center. Consolidating and rationalizing your storage configurations will save your company money. If you have any proven ideas, send them our way and we will post them.
