1

Fact or fiction without any friction, I have heard in passing on the web only about 2 times, that hard drives will re-write the data onto the platter when Only reading the data off of that platter.

Question applies to modern normal consumer hard drives with spinning platters, and read/write heads.

Do hard drives really re-write data when only reading data from the drive?

If i am supposed to have a specific problem, then: For very long term storage of data, if this is actually true, then reading the data once a year could insure a better magnetic polarity differance. For long term storage not only would it be good to spin it up and get the motor moving, but to do a read of the data one time.

1
  • What is your probelm you'd like solved. I'm considering flagging this question.
    – wizlog
    Oct 18, 2011 at 16:49

2 Answers 2

4

Reading occurs more often than just when you execute a file. So it may appear to some that this is the case. But the reading that occurs is always at the request of running software.

Wizlog is correct, when the drive is ONLY reading, it will not re-write the data. However, metadata is usually recorded when the files are read. For instance, in Windows you can view the Date Last Accessed, which is not the last time the file was modified, but the date the file was last opened. While the file data may not have been modified when the file was accessed, the meta-data associated to that file was written to record the fact the file was accessed.

Also, the system may be reading and writing at any given time when instructed to do so by software.

In an investigation when people wish to view the contents of a drive without in any way modifying the data, you'll use special read-only hard drive interfaces which will allow the system to read the data while preventing any meta-data, or any other data, from being written to the drive.

2
  • Thanks, the persons on the web saying it portrayed that they had some knowledge of the internal workings of the hard drive. Thier statements were never countered or questioned, in the threads. Mabey they were confused by the updating of the minor data.
    – Psycogeek
    Oct 18, 2011 at 17:07
  • Core memory worked that way: The only way to tell whether a core's magnetization was "left" or "right" was to send current through the x/y wires to set it, say, left... and then a sense wire would detect the magnetic field change, if it had changed. If it didn't change then it must have been "left" already; if it did change, then it had been "right" and would have to be put back to that state. We don't need to do this on hard drives because the platter is always moving under the heads - hence always inducing a waveform that reflects the polarities of the magnetic domains. Feb 2, 2015 at 14:52
3

No. It does not. When your HDD is reading data, it uses the read part of the read-and-write head. When writing data, it uses the write part of the head. When doing both, it uses both.

1
  • There is no "read part" or "write part" of the head. There are two sets of electronics, one for writing and one for reading, but they are both connected to the same head coil. Some high-end analog audio tape recorders did have separate record and playback heads (and a third head for erasing) but hard drives don't need that. Apr 2, 2017 at 12:31

Not the answer you're looking for? Browse other questions tagged .