Surface Test in order of effectiveness & Recommendation

[Octavehl]

Hi
In below link are details on Surface Test.
http://www.hdsentinel.com/help/en/61_surfacetest.html

I have the Pro version and my understanding is that following are the order of effectiveness in detecting issues and possibly resolving for potentially higher reliability in future.
1. READ Test
2. WRITE test
3. WRITE + read test
4. Read +. WRITE + read test (refresh data area)

Is the last "Reinitialize disk surface" better than 4? In another words, if I'm willing to invest a lot of time, then ideally I should do this.

I ask, because if most of 1-4 methods are (24/11/15 edit: possibly) destructive, then as long as I have time, I would want to do the most thorough surface test (unless of course difference of time requirements between 4 and THE most effective method is astronomically different like 12hrs vs. 40hrs).

Thank you

[hdsentinel]

Thanks for your message.

This is not really the order of effectiveness, just a partial list of the possible surface test.

You did not mention the most important one, the Reinitailise Disk Surface test - but it's the most effective in repairing problems related to the disk surface, so we surely need to consider this.

Generally, to reveal the problems, detect possible unreadable / slower areas and sectors which can be utilised after retries, the READ test is the perfect way.

And if this shows problem(s), the best way to repair is to use the Reinitialise Disk Surface test.
This forces the hard disk to repair such sectors by
- reallocation of the sectors to replace them with the spare area
- or if the sectors were just weak but not physically damaged and can be re-used, then restore their status to normal

If you prefer a quick, complete overwrite / clearing of the drive, then the WRITE test may be enough also to repair the sectors, but not in all cases.

If you want to also verify the effectiveness of the write test, then the Write + Read test is a good way which verifies that the written data can be read back and accurate.

This is a simpler, quicker version of the Reinitialise Disk Surface test, which peforms additional overwrite passes to increase the effectiveness.
Yes, this takes longer time - but usually more effective.

The Read + Write + Read test is completely different. It is absolutely NOT designed to repair problems, as if the read scan detects error, then for safety reasons the wite pass is not initiated at all.
As its description suggests, this test is designed to refresh the data area of the device, to keep the data readable and "fresh".
This may be useful mostly for flash-based storage, like memory cards / pendrives where the devices may not be able to hold the stored data for longer time. Usually after such refresh, the device works better, faster, like a new one.

For diagnostics and repair, please check http://www.hdsentinel.com/faq.php#tests for the list of recommended tests:

1) Disk -> Short self test
2) Disk -> Extended self test
3) Disk -> Surface test -> Read test
4) Disk -> Surface test -> Reinitialize disk surface

(the Read + write + read test is not mentioned there, as it is not really designed for this purpose).

If speed is more imprtant than effectiveness, then you may try

Disk -> Surface test -> Read test first to detect if there is any problem which may require correction / attention. And if it shows problem, then you may use

Disk -> Surface test -> Write + Read test

or even

Disk -> Surface test -> Write test

as these are faster than the Reinitialize disk surface test.

[Octavehl]

Thanks for the reply on my post and link for the relevant FAQ.

Doing what’s recommended for “refresh” of SDD, am I able to run Disk->Surface Test->Refresh data area for my only SDD on my laptop which has my OS+data? Any risks?

Also as separate side question - to leverage the useful temperature offset function documentation suggests measuring HD temperature with infrared thermometer – where on the HD physically do you actually sample the temperature?

Thanks

[hdsentinel]

Excuse me, but it is not really possible to refresh the SSD if the OS is active on that
To perform the refresh, the device need to unmounted, to prevent any other software (or Windows itself) from accessing the device during this procedure. And because the system drive contains many open files (for example the swapfile, registry, temporary files/folders used by other software etc.) it is not possible to lock the drive for exclusive use and especially to unmount it.

This is a bit similar as chkdsk work: chkdsk /f can't be used on system drive while Windows and other applications are active..

Generally the refresh function designed to refresh memory cards, pendrives, or additional (secondary) drives. With them, it is possible to unmount the device for the duration of the test.

For system drive, the only way is to boot from an other media, for example a live Windows CD/pendrive and start Hard Disk Sentinel Pro Portable. Then the original SSD can be locked (as the installed Windows is not active at that point) and then it is possible to perform the refresh.
Currently investigating the possibilities of performing refresh on the active system drive, so it may be possible later.

The refresh (as it performs a write back to the data sectors) is always risky. If you experience a power failure / power loss, the currently written data area may be inconsistent, the data may be corrupted / damaged. So use with care.


About the temperature measuring: I always recommend to measure at the bottom side, near to the central point of the rotating disks, where the axis of the disk platters located. As this point receives the highest mechanical stress (due to the constant spinning) it's temperature may be important - and this should be relatively close to the reported temperature value.

Please see the Help -> Appendix -> Temperature calibration for more details about the temperature calibration, to correct possible improper temperature values. This may not required on most modern hard disks, but several older hard disk drives (Maxtor, Samsung, older WD) affected as they may report incorrect tempertures.

[Octavehl]

Hello. Sorry I realized I had not properly thanked you for the help here. Thank you!