A friend recently circulated an email reporting that her laptop had died, apparently due to hard drive failure. (Yes, it could be fixed, but it’s an old laptop, so she has decided to upgrade.) More importantly, she reported that several IT types had diagnosed the problem as relating to the fact that she typically relied on stand-by rather than turn-off, generally only using the latter about once a month. This practice, they advised, had caused the hard drive to burn out. They further advised that she should be doing a full turn-off at least once a week.
IANACG, but this diagnosis sounds wrong to me. Stand-by shouldn’t be using the hard drive. And, if it does, turning off the computer once a week rather than once a month shouldn’t solve the problem. So, what’s the Straight Dope? And, even assuming stand-by doesn’t imperil the hard drive, does it cause other problems? How often should Windows be fully shut down? And is hibernate any better for any of these issues?
That doesn’t sound right to me either. In standby mode the hard drive usually spins down, so that the platter isn’t spinning at all, but the drive is still powered up waiting for a read or write command. If anything, having the hard drive in a spun down state most of the time is healthier than having it always spinning, and many people leave their computer on with the platters spinning 24/7 and the hard drives should last years like that. The process of spinning down and back up of the hard drive can be stressful, but any modern drive should be well engineered for this.
I think it was probably just a bum drive and they are grasping at straws to figure out what when wrong. The reality is that sometimes a hard drive will just go bad after a month or a year, while the one right behind it on the manufacturing line will last for ten years being spun up and down thousands of times. This is especially true with the inexpensive consumer drives. Backups are your friend.
If she is putting the system into and out of standby more than five or ten times a day, though, it may be best to either leave the computer on or alter her usage habits a bit. While this normally shouldn’t hurt it, it will put more stress on the drive than just leaving it running. The same would be true if she shuts the system down instead of putting it in standby, though, so I think they are wrong in their diagnosis.
If your computer is well maintained, without a lot of buggy programs running, you should really only need to do a full restart of they system when an update or installation calls for it. On a few of my systems this tends to be about once a month, and every couple of weeks on my more cluttered systems. Windows has gotten much better about being internally bug free enough that OS crashes are almost always bad hardware or a buggy third-party driver.
When providing technical support to the less computer savvy, I often tell them once a week, whether they need it or not. This is because they’ve probably loaded their system up with programs that will eventually hurt the system performance if left to run indefinitely. It’s easier to give them a general schedule to follow than to expect them to diagnose when programs are starting to have problems. This isn’t to save their hard drive or anything in the system, but just to reset all the running programs to a more stable state.
Hibernation is really only better if you need to conserve all your battery power. Instead of storing all the program states in RAM (which uses some battery power) and continuing to power the core devices in the system, it saves the program states to the hard drive and then completely turns off the computer. The trade off is that it takes longer to start back up as loading the state from the hard drive is slower than from RAM.
Ah, the old IBM ‘Deathstar’. IBM once produced a range of hard disks which had precisely this problem. Google[sup]TM[/sup] ‘ibm deathstar’ for endless tales of woe.
Many laptops run much closer to the edge of overheating than is common on desktop machines. If hers was one of them, then going to standby rather than a power-off condition would mean the machine stays warmer. That could eventually cause earlier failure in the hard drive.
But I agree with others that this sounds unlikely. I’d be more likely to think it was a bad hard drive, or one of a badly-designed model of hard drives.
Our corporate IT department has sent out a notice to all of us who use IBM T42 laptops, regarding problems with drive crashes when the laptops are put in standby or hibernate mode and subjected to bumps or drops. Apparently, some of the active protection features of these models don’t work as well or at all in these modes. We’ve had a high rate of failure with drives in these models; I know of six or seven colleagues who’ve been affected by this, as have I. So there may be some validity to the story she’s heard, particularly if she commonly moves the machine around while it’s in standby.
Thanks for the replies. I think rackensack probably has hit the nail on the head. My friend works for a law firm, so it wouldn’t be surprising if the knocks-and-bumps problem has come up with laptops on the road. As it happens, this doesn’t explain my friend’s problem - her laptop never moves - so it probably was, as others have suggested, simply a problem with the drive. But now I know how her IT folks came up with the diagnosis. And learned something useful, as I have on occasion moved my laptop in standby mode. No more.