My company makes long term disk-based archiving products, so I'll give you my elevator pitch about how we approach this concept: First, yes indeed you need to spin up the drives periodically but not because of lubrication. Rather, it's because the components in the HDA "outgas" and particles condense on the media. Running the drive and scanning all cylinders breaks these particles loose before they can accumulate to dangerous levels and the rotation of the media causes the flow of air through the filtration system. In our system, every two months we read every file on the disk and verify its cryptographic hash. If there are any errors we repair using either the RAID parity drives (and then re-check the hash) or we use a second copy we keep on a different RAID shelf. Asynchrounous to this process, we have a raw media scan every two weeks on the RAID shelves which verifies that each drive thinks its data is readable (and if not we re-construct it and map out the newly bad blocks). Error stats are sent to the end user and to our tech support center so that if numbers start to climb we can start looking at a forklift upgrade of the hardware and automatic data migration of the user data. Protecting the data is also what you might call a "disk firewall", which means we have a modified disk driver which won't write to the repository unless the data is coming from a cryptographically signed process, and likewise, the RAID controller won't respond to drivers other than the modified driver. These last steps help guard against hostile administrators or rogue processes running at admin privilege levels (not claiming this is foolproof but it means that only a specifically targeted attack on our system has any hope of corrupting the disk data).