If I do the read test and everything shows up perfect (all green blocks, zero errors), does it make sense to also run some other surface tests to be 100% sure everything is fine (such as write test and others)?
In other words, can other surface tests show/reveal/detect problems that the basic read test can't? I just want to be sure. Thank you very much.
Is the basic surface read test enough to detect any potential surface issues?
-
hdsentinel
- Site Admin
- Posts: 3019
- Joined: 2008.07.27. 17:00
- Location: Hungary
- Contact:
Re: Is the basic surface read test enough to detect any potential surface issues?
The Disk menu -> Surface test -> Read test is a good start, but in many cases, it may be not "enough".
For example with SSD or SMR hard disk drive (assuming that the drive is new/empty so generally unused) the drive may not actually reads the data sectors, but responds quickly from a look up table that "the sector is empty".
Please check this topic:
https://www.hdsentinel.com/forum/viewtopic.php?t=14522
which explains exactly similar situation when a Read test on a new/empty hard disk shows very high speed (because the drive did not read the appropriate sectors) and then, when the drive filled, we can see the "real" performance.
So to be 100% sure, I'd use a Write type test too (and maybe the hardware self tests eg. Disk menu -> Short self test, Extended self test) as suggested at
https://www.hdsentinel.com/faq.php#tests
This suggests (in addition to the Read test) the Reinitialise Disk Surface test - which may take too long time for a high capacity drive, so you may use the "light" version of this test, the Disk menu -> Surface test -> Write+Read test instead, which performs a single pass overwrite and then read back and verify if we read back what the first pass written.
This would surely verify if all sectors could be written and read too.
For example with SSD or SMR hard disk drive (assuming that the drive is new/empty so generally unused) the drive may not actually reads the data sectors, but responds quickly from a look up table that "the sector is empty".
Please check this topic:
https://www.hdsentinel.com/forum/viewtopic.php?t=14522
which explains exactly similar situation when a Read test on a new/empty hard disk shows very high speed (because the drive did not read the appropriate sectors) and then, when the drive filled, we can see the "real" performance.
So to be 100% sure, I'd use a Write type test too (and maybe the hardware self tests eg. Disk menu -> Short self test, Extended self test) as suggested at
https://www.hdsentinel.com/faq.php#tests
This suggests (in addition to the Read test) the Reinitialise Disk Surface test - which may take too long time for a high capacity drive, so you may use the "light" version of this test, the Disk menu -> Surface test -> Write+Read test instead, which performs a single pass overwrite and then read back and verify if we read back what the first pass written.
This would surely verify if all sectors could be written and read too.
Re: Is the basic surface read test enough to detect any potential surface issues?
Wow, thanks for a quick and comprehensive answer.
The drive is external WD Elements 10 tb (CMR HDD). It's actually filled up, so I don't really want to do write test at this point (and lose data lol). So I guess, the read test is the only option for me here? By the way, the short self-test and SMART values are all OK.
Also, would you consider filling up the drive (simply copying stuff to it), a good write test in it of itself? Especially when monitored by your program? The whole process of filling up the drive with my data over the last few days was monitored by HDSentinel and it never reported any issues during the copying. If there would be any surface/write/read issues present, is it safe to say your program would already detect it and show it?
Oh, and just one more question (sorry if it's stupid). Can write tests reformat the drive? Or does the file system always stay the same? NTFS stays NTFS?
Thank you very much, I have been quite paranoid last couple of days.
The drive is external WD Elements 10 tb (CMR HDD). It's actually filled up, so I don't really want to do write test at this point (and lose data lol). So I guess, the read test is the only option for me here? By the way, the short self-test and SMART values are all OK.
Also, would you consider filling up the drive (simply copying stuff to it), a good write test in it of itself? Especially when monitored by your program? The whole process of filling up the drive with my data over the last few days was monitored by HDSentinel and it never reported any issues during the copying. If there would be any surface/write/read issues present, is it safe to say your program would already detect it and show it?
Oh, and just one more question (sorry if it's stupid). Can write tests reformat the drive? Or does the file system always stay the same? NTFS stays NTFS?
Thank you very much, I have been quite paranoid last couple of days.
-
hdsentinel
- Site Admin
- Posts: 3019
- Joined: 2008.07.27. 17:00
- Location: Hungary
- Contact:
Re: Is the basic surface read test enough to detect any potential surface issues?
Yes, if the drive already contains data (filled at least once completely), then the Disk menu -> Surface test -> Read test would be perfect.
I thought you may mean a new/empty/unused drive.
On a drive with data, yes, I'd only recommend the Disk menu -> Surface test -> Read test.
> Also, would you consider filling up the drive (simply copying stuff to it), a good write test in it of itself?
> Especially when monitored by your program? The whole process of filling up the drive with my data over
> the last few days was monitored by HDSentinel and it never reported any issues during the copying.
Yes, that's a good way too, as it tests MOSTS sectors (not all, but assuming the partition(s) cover the entire disk, this method may help too).
> If there would be any surface/write/read issues present, is it safe to say your program would already detect it and show it?
Yes, exactly as you wrote: during any write/read operations (performed by any software, so eg. by copying data) the drive would "recognise" and record possible problems - and these would be detected/reported by Hard Disk Sentinel too.
> Oh, and just one more question (sorry if it's stupid). Can write tests reformat the drive?
> Or does the file system always stay the same? NTFS stays NTFS?
The write tests (Disk menu -> Surface test -> Write test, Write+Read test, Reinitialise Disk Surface test) clears ALL data, including the partition table.
So no, it does not keep (or format) the previous NTFS file system. After these write tests, you'd need to open Windows Disk Management (by the 9th button in the main button bar of Hard Disk Sentinel) and need to initialise the drive, make one or more partition(s) and format them as you prefer (just like for a completely new, empty drive).
I thought you may mean a new/empty/unused drive.
On a drive with data, yes, I'd only recommend the Disk menu -> Surface test -> Read test.
> Also, would you consider filling up the drive (simply copying stuff to it), a good write test in it of itself?
> Especially when monitored by your program? The whole process of filling up the drive with my data over
> the last few days was monitored by HDSentinel and it never reported any issues during the copying.
Yes, that's a good way too, as it tests MOSTS sectors (not all, but assuming the partition(s) cover the entire disk, this method may help too).
> If there would be any surface/write/read issues present, is it safe to say your program would already detect it and show it?
Yes, exactly as you wrote: during any write/read operations (performed by any software, so eg. by copying data) the drive would "recognise" and record possible problems - and these would be detected/reported by Hard Disk Sentinel too.
> Oh, and just one more question (sorry if it's stupid). Can write tests reformat the drive?
> Or does the file system always stay the same? NTFS stays NTFS?
The write tests (Disk menu -> Surface test -> Write test, Write+Read test, Reinitialise Disk Surface test) clears ALL data, including the partition table.
So no, it does not keep (or format) the previous NTFS file system. After these write tests, you'd need to open Windows Disk Management (by the 9th button in the main button bar of Hard Disk Sentinel) and need to initialise the drive, make one or more partition(s) and format them as you prefer (just like for a completely new, empty drive).
Re: Is the basic surface read test enough to detect any potential surface issues?
Wow, thank you so much for these very clear answers. I feel so much calmer now. Have a great day man. Your program is the best 