h a l f b a k e r yBusiness Failure Incubator
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Many hard-disk drives have a feature where the operating system can tell the drive to turn off its spindle motor to save power, prevent wear, or reduce noise when the drive is not being used (when the machine is idle.) However, there are many situations where it is desirable to shut down the drive in
this manner, but one is running a program that still needs some non-volatile storage space. For example, if you are running a word-processor for many hours on end, there is no need for the ability to access the entire drive - only that portion of the platter where your file is stored. THUS, I propose a hard-disk where "linear storage zones" exist. Normally, a file (if not fragmented) will be written in a shape resembling an arc - the platter will rotate as the write head will stay in a fixed position for a brief moment. However, the proposed drive would have the ability to move its read/write head even when the spindle motor is not running. When the drive enters the "low power" mode and shuts off the rotor, it would first turn the spindle to a known "linear storage" position. Here, as the name implies, files would be stored as patterns oriented along the radius of the platters. Thus, the drive would have a special zone in which reading and writing could be done only when the drive is not spinning. The zone could possibly include several hundred MB of data. Thus, there would be no need to keep the drive spinning when running a word-processor, as there would be a small amount of nonvolatile storage space that you could save your work on even when the drive spindle is not turning. The technology could be used to save significant amounts of power on portable computers, as most users rarely run software that requires rapid access to more than several hundred MB of storage at one time.
[link]
|
| |
My croissant. I don't know enough about the internal hardware of a HD to know how much space you'd get like this, but it's an interesting idea. Problem is, the drives don't park unless you're not using them. You'd also have a problem trying to read the file with the disk running, later, unless you mean to use it as just a swap file. Instant super fragmentation. |
|
| |
//if you are running a word-processor for many hours on end, there is no need for the ability to access the entire drive - only that portion of the platter where your file is stored. THUS, I propose a hard-disk where "linear storage zones" exist.//
How many drives do you have on your hard disk? The way I'm setting my new one up will have 15 drives overall. That way, whatever I'm woiking on puts me on the outer (faster) edge of it's assigned drive. |
|
| |
With MTBFs in the hundreds of thousands of hours range, I doubt you'll save much wear and tear on a modern hard drive. Once you look at all the OS and hardware changes that would be involved it might be cheaper to create a RAM disk or use a FlashRAM or PCMCIA hard drive. |
|
| |
[thumbwax] Do you mean 15 partitions? |
|
| |
My (admittedly limited, possibly archaic) understanding is that the head maintains the correct distance from the surface by the aerodynamics of the spinning drive, and that stopping the drive with the head over the data surface destroys these aerodynamics and causes a head crash. If that's correct, you wouldn't be able to read or write from a stationary disk surface. |
|
| |
I thought about this and came to the conclusion that if you spun the head and inched the drive forward during normal operation then you would be able to read the disk while it was stationary. |
|
| |
Then I realised that you'd have to spin the drive head up and down to read the stationary drive so it just shifts the problem. |
|
| |
How about a non-volatile write cache built into the drive which writes itself to the disk whenever the disk spins up? |
|
| |
The amount of data that could be stored on an 'arc' would be extremely small, even if the necessary hardware were included to allow for it. A more interesting notion would be to design the electronics for hard drives and CD-ROMs so that they can read data, albeit somewhat slowly, while the drive is 'spinning up'. Writes could probably only be done safely while the drive was stable at speed, but reading data during the spin-up time shouldn't pose a problem. |
|
| |
Ever run FileMon (www.sysinternals.com)? You will see that on the average computer there are many processes that constantly access the hard drive - even when idle. This idea may have worked well with DOS, Windows 3.1, or even 95; but not today. |
|
| |
Woah, and I thought Other:gravity was the home of fishbones. This category is unbelievable. |
|
| |
Doesn't the bulk of mechanical wear (on any device)
occur during start up and shut down? Wouldn't this only
serve to aggravate this instead of reducing it? |
|
| |