Page 1 of 1

Incorrect "Performance" information 100% ???

Posted: 2019.02.15. 22:06
by Nero
Incorrect "Performance" information 100% ???
I do not have this disk anymore, but the read / write speed of the disk has dropped practically to zero. And the disk with such a low speed was not able to write or read anything. Unless "Performance" concerns a different parameter?

Re: Incorrect "Performance" information 100% ???

Posted: 2019.02.18. 11:01
by hdsentinel
No, it is not "incorrect" of course.

The Performance value determines if there is any issue (for example cable/communication problem or similar) which may reduce disk performance/transfer rate.
Also it shows if there is any issue related to attributes related to disk performance (for example "seek time performance" value if supported/provded by the drive).
Plus it shows (for older IDE drives) when the drive has cable issues or using PIO mode which may also dramatically reduce performance.

I'd more than happy to see such situation and check, examine how things can be even further improved.
So in such / similar situations, please try to use Report menu -> Send test report to developer option. Then it is possible to examine the complete status of the disk drive and verify the attributes, their values and check which can be outlined better (if any) to indicate lower performance.

> but the read / write speed of the disk has dropped practically to zero. And the disk with such a low speed was not able to write or read anything.

Did you see such behaviour during the complete surface test too? Or just experienced during file operations (read/write files and folders)?
Slower file operations usually related to fragmentation which is completely different situation.

Re: Incorrect "Performance" information 100% ???

Posted: 2019.02.18. 16:37
by Nero
Currently, this is not possible, the disk "theoretically works" but is not detected (or detected only really super-shortly). This may be due to unstable sectors on the disk that I managed to capture earlier. Then all partitions were damaged. All partition tables have been damaged. At the moment I do not have the possibility to recover the data because the recovery of 1.4 GB is costly in a data recovery company and it is not known whether it is possible.
Theoretically, the problem can only arise from unstable sectors, as other SMART parameters were normal.
Usually and most often bad sectors appear at the beginning of the partition or at the end of the partition, I do not know what it depends on.

Re: Incorrect "Performance" information 100% ???

Posted: 2019.02.20. 13:58
by hdsentinel
> Then all partitions were damaged. All partition tables have been damaged.

Yes. This is exactly what I wrote in the other topic (where you mentioned testing/clearing one partition only) that in this case, the complete disk drive is affected, so possible other partitions can be damaged/corrupted.

> Theoretically, the problem can only arise from unstable sectors, as other SMART parameters were normal.

Yes. These should be fixed - and generally avoided, to prevent such issues.
Constant disk monitoring would help to see when the count of such weak sectors begin increasing as then we can do the required steps immediately: checking connections, replacing cables, perform disk testing to reveal and fix the problems.
These are all covered in
https://www.hdsentinel.com/hard_disk_ca ... ectors.php

Otherwise, with no monitoring, we can't see the degradation - until there are LOTS of files are unreadable, LOTS of folders vanish, the complete system is not bootable and so.
The problem is that most users install Hard Disk Sentinel (or any other tool) this time and wants to repair all problems quickly and without any kind of data corruption / data loss. This is simply not possible this time: we'd need to prepare and protect long before the problems occur.
Similarly to seat belt or airbag in your car - designed to prevent problems and reduce any possible injury.


> Usually and most often bad sectors appear at the beginning of the partition or at the end of the partition.

This is not a strict law, but yes, in many cases such sectors appear at the end of the disk drive (not the partition).
This (as a generally common situation) described at Support -> Knowledge base -> Hard disk cases -> Bad sectors
( https://www.hdsentinel.com/hard_disk_ca ... ectors.php )

Re: Incorrect "Performance" information 100% ???

Posted: 2019.02.20. 17:23
by Nero
checking connections, replacing cables, perform disk testing to reveal and fix the problems.
There is nothing to do here. Modern SATA cables simply do not damage themselves. Once - only the problem was the old vein cables (ATA) in older types of disks.

How do bad sectors happen?

- Some program incorrectly saved data bits (bug in the program)
- Some program mistakenly tried to defragment files (bug in the program)
- Some Program while aligning the partitions caused an error and bad blocks appeared (bug in the program)

--------------------------------
Let's look at the situation:
SMART shows that there is one bad sector on the disk (probably 4KB).
Nevertheless, the disk works fully efficiently.
What to do in this case?
Make a backup, is it better to perform an analysis that will allow you to repair a bad sector?

A disk with 1 bad sector can work correctly.
But if you are tempted to analyze, you will cause that the number of bad sectors will increase, even it can show 257 !!!

But how? Previously, there was one sector and the disk was working properly, but after analysis 257 bad sectors were found and it is no longer possible to continue to work on the disk.
a) the disc has stopped being detected (BIOS, OS)
b) damaged file table
c) damaged MFT and important NTFS system files

Each analysis will detect more bad sectors, on the contrary, it will cause new ones to appear until the very end of the analysis, although the analysis will not end, because the disk will be completely damaged and you will lose valuable data.

The best solution is just a copy of the data. A copy of the data will not cause bad sectors to build up. There will still be one bad sector instead of 257 !!!

Unfortunately, I was tempted by the information that the disk can be repaired.
This caused the opposite effect - to the intended one.

In the past I used the program:
Unstoppable Copier 5.2
Recover 99.9% of all data
I erase one of the partitions, it still worked 2 years without problems.
And now I tried in a different way and lost data.
I did it because I did not have a replacement hard disk.

Re: Incorrect "Performance" information 100% ???

Posted: 2019.02.28. 10:40
by hdsentinel
> There is nothing to do here. Modern SATA cables simply do not damage themselves.

Excuse me, but this is absolutely incorrect.
There are lots of problems with (usually older) SATA cables, they simply do not allow proper connection. If you attempt, they may be moved easily (loose) and do not attach properly.
Yes, sometimes they can work - even for months - until problems begin appearing.

Also it is extremely common that
- hard disk (older, slower) replaced with SSD
- the cable is not changed
but the cable may not able to properly support newer SATA 6G speeds - resulting communication errors soon.

Sorry, but this is wrong.


> Once - only the problem was the old vein cables (ATA) in older types of disks.

Not really.
With older IDE/PATA drives, yes, there were lots of issues, but usually caused by that the cable was simply too long or the drives incorrectly connected: the "longer" end should be connected to the motherboard (controller), the shorter end to the MASTER drive, the middle to the SLAVE drive.
Just as this is not practical and may not always possible due to length, personally I see 100's of PCs (with problems) where the IDE cable connected improperly (devices/ends swapped). But this is other story ;)

So even if sounds surprising, SATA cables CAN cause communication problems. In "best" case, these reduce performance - but data corruption / data loss may happen because of them.
See https://www.hdsentinel.com/hard_disk_ca ... _error.php
which describes (I receive dozens of reports about similar situation every day....)


> How do bad sectors happen?
> - Some program incorrectly saved data bits (bug in the program)
> - Some program mistakenly tried to defragment files (bug in the program)
> - Some Program while aligning the partitions caused an error and bad blocks appeared (bug in the program)

Excuse me, but this is the biggest incorrection I *ever* read in the last 20 years.

Of course any bug in any program does NOT cause any bad sectors.
Just to prevent false auumptions, I write again.
Of course any bug in any program does NOT cause any bad sectors.

Bad sectors happen completely differently.

1) Hard disk (or SSD) attempts to read back a sector and the read attempt failed completely, for example because of mechanical damage (small, microscopic scratch, dust, head print, damaged head / surface resulted particle etc.)
In this case, the drive internally marks the sector as bad and all further reads/writes targeting this sector re-directed automatically to a sector in the spare area.
The count of reallocated (bad) sectors increase - and ideally the drive works perfectly (even if it has bad sectors) because the spare sector can be read/written without problems.
This is why some bad sectors can be even acceptable ( https://www.hdsentinel.com/faq.php#health )
This is caused by hardware failure, completely independent from any software, OS and so.

2) the sector can be read - but the stored checksum does not match with the checksum calculated previously.
Then the sector is usually marked as WEAK (pending). Chkdsk/format reports as "bad sectors" incorrectly and disk surface tests show them and can be corrected.
Usually related to problems outside the hard disk / SSD, for example
- cable/connection issues
- insufficient power (very common with USB 2.5" hard disks used with single USB connection which may not provide enough power for proper operation)
- accidental power loss, reset during write
But in some cases, the drive may be simply not "good" enough to record data for longer time. If the internal heads damaged (or in case of power issues), then it is possible that the data recorded can't be properly read back after some time. Or when the SSD reached low health due to wearout, then it can be still written - but after some time, it is possible that the data can't be read back.
This is when you may expect new and new weak / pending sectors caused by disk scan (we discussed in your other posts) and this is when the Disk menu -> Surface test -> Reinitailise disk surface test can be better: as it forces overwriting all sectors multiple times before clearing and verification.
This helps the drive to repair problems and restore the sector to good, working state - or perform reallocation if required.


> SMART shows that there is one bad sector on the disk (probably 4KB).
> Nevertheless, the disk works fully efficiently.
> What to do in this case?
> Make a backup, is it better to perform an analysis that will allow you to repair a bad sector?

The bad sector is already repaired: it replaced by the spare sector.
There is nothing to repair now - as it is already fixed.

Yes, in this case (especially if the bad sector is new, just appears) I usually always recommend a backup of important files, to be sure and perform testing. NOT to repair this sector (which already repaired) but to reveal any possible OTHER sectors and stabilize them NOW (when we can keep an eye on the drive and already saved the most important data).
This is exactly described at
https://www.hdsentinel.com/faq.php#tests

And if the tests show more problems then we'd need to consider possible factors: loss of data, downtime, maintenance costs etc.
For a PC used for web browsing, this may be not critical.
For a hospital, it may be life critical. In a company, possible unwanted downtime can be also expensive, so better to prepare and make replacement.

But if the tests show no problems, confirming that the bad sector is already fixed and there are no other problems, we can acknowledge this and reset the error counter, to focus on possible new problems only.
This is described at
https://www.hdsentinel.com/faq_repair_h ... _drive.php


> A disk with 1 bad sector can work correctly.

Yes of course, this is completely true. Then the health is 99%, it is still Excellent.
( https://www.hdsentinel.com/faq.php#health )


> But if you are tempted to analyze, you will cause that the number of bad sectors will increase, even it can show 257 !!!

Yes: if the drive is not stable, then it is completely normal and expected to reveal problems.
As I already wrote (if I remember, many times) those problems would SURELY appear later, even with normal use, when you simply want to copy/move files, save important data on the drive.
Which is better:
1) having a complete backup and then performing a test to reveal such issues in controlled environment, where you can be prepared, monitor status (and change) analyse possible corrupted files / data area and attempt to repair, fix issues or (in worst case) replace
or
2) do not perform any backup or any test and feel the drive is stable (which is usually NOT the case when a bad sector detected) and face problems, corrupted/damaged/lost files, unbootable systems - usually at the worst time?

When one bad sector detected, there is good chance that MORE follows. Not in all cases, personally I have (and use) drives with 1-2-5 bad sectors, they work stable for years.
But I receive messages every day that "Hard Disk Sentinel reported one bad sector, I started to copy files and during the backup (!) the health went down to 50% and then during the test it went down to 5% or less". But no data lost because all files could be saved - and then the test CONFIRMED that the drive is far from perfect.
If the test would show no issues, then with 1-2 bad sectors, the disk is perfectly usable.


> But how? Previously, there was one sector and the disk was working properly, but after
> analysis 257 bad sectors were found and it is no longer possible to continue to work on the disk.

Yes. Because the drive had MORE problems - but they are not detected until the appropriate sector accessed for ANY purpose
(including backup, saving, reading, testing, etc.)

This is exactly described in a VERY COMMON case:
https://www.hdsentinel.com/hard_disk_ca ... ectors.php
when problems remain unnoticed for years, until the drive is (almost) completely filled.
If you read that case: "Initially, the hard disk health reported was 90 %." and then, on the bottom "The health determined after the test reduced to 7%, indicating that now the drive is properly identified the problems."

This does NOT mean that the disk surface test CAUSED the problems - the test only revealed the problems in a controlled, managed, monitored environment (after saving the most important files to ensure we're safe now).

Even NEW, unused disks can have such problems (!) which may remain undetected without testing.
This is why I always recommend intensive testing even new drives:
https://www.hdsentinel.com/faq.php#tests
exactly to reveal issues THEN, long before the drive is filled with TBs of important data.

> a) the disc has stopped being detected (BIOS, OS)
> b) damaged file table
> c) damaged MFT and important NTFS system files

Yes of course, I'm afraid with a drive with lots of problems these are completely "normal" and "expected" (even if we do not really like).
This is the purpose of testing, to reveal - or confirm if the data area (which includes the spare sector, used instead of the original bad sector) is error-free.


> Each analysis will detect more bad sectors,

If the drive is not stable (see what I wrote about head damage or insufficient power used when writing) then it is completely normal and expected.
This is why in this case we'd need to perform more effective repair functions: Disk menu -> Surface test -> Reinitialise disk surface, exactly as suggested at https://www.hdsentinel.com/hard_disk_ca ... ectors.php

This would stabilize, fix problems, repaired 10000's of hard disks, SSDs, pendrives, memory cards....
And if the drive reports issues even AFTER this, then I'm afraid it should be no longer used for storage (it may have low health in this case).

> on the contrary, it will cause new ones to appear until the very end of the analysis,
> although the analysis will not end, because the disk will be completely damaged and you will lose valuable data.

Sorry, but at this point (if you did not have backup) you ALREADY lost your valuable data.
Any attempts (even backup / copy of the affected sectors) would do 100% same: increase the problem count.

> The best solution is just a copy of the data.

Yes, this is true: but ideally you'd have backup configured previously, when the disk was stable. Then all data could be surely read.

> A copy of the data will not cause bad sectors to build up. There will still be one bad sector instead of 257 !!!

If your backup files which are (if you're lucky) stored on other, good sectors, then it may be true: then yes, as the problematic sectors not touched, no more problems detected/reported.

This is why (upon problems) it is better to attempt to COPY files, important data first, exactly to minimise possible delays, retries, restarts (which can cause even more damages and more unreadable files).
For example, if there is a small dust/metal particle floating inside the HDD, it may damage other sectors too when the head reaches it.

This is why you'd need to separate:
- perform backup to save data
- THEN perform test in order to verify if the disk is (at least) usable, reveal and stabilize the problems in all possible ways.

But these are completely different things. We can't say "file copy" is better than "disk surface test" as these are designed for completely different situations/functions.


> Unfortunately, I was tempted by the information that the disk can be repaired.
> This caused the opposite effect - to the intended one.

Yes: I'm afraid if the drive reports new and new issues, then in most cases it can't be repaired. But you can attempt to stabilize - and THEN the test would confirm if the drive can be used - or not.

The tool you mentioned designed to copy files. Again if you're lucky, then the files stored on the good areas and most of them could be copied.
But for this, you do not need any special tools: you can copy with Explorer, Total Commander or any other file manager tools.
But if we speak about important, critical data, I'm afraid we can't always expect luck....