VOGONS


First post, by CYRIX

User metadata
Rank Newbie
Rank
Newbie

Hello, recently a friend was gifted to me 5 different old hard drives, and I decided to scan them with the HD REGENERATOR program, but one of them SEA GATE 40 GB turned out to have 3 DELAYED sectors. I tried to fix them by use setting the REGENATE ALL SECTORS option, but still these DELAYED sectors are still there. I tried with other programs like VICTORIA but this program brought out even more errors and defects. Then I also tried with HDAT2, but there the results were that the hard drive is clean without any errors.

After a lot of reading on the internet, for weeks, I came across your forum. I've been trying different options with MHDD for almost a month now, carefully reading the entire manual, and then trying everything step by step with the ERASE, then ERASE DELAYES, then REMAPPING options. I ran another scan with HD REGENERATOR, but again those DELAYED sectors are still there! How can I remove these DELAYED sectors? Please someone with more experience help. Thanks in advance!

P.S.
I specify that, I tested these programs under pure DOS, no Windows! I even start MHDD with BIOS off so it can work directly with the hard disk controller!

When all computers were pizza boxes , I wanted a tower case. Now that all computers are towers, the classic pizza-boxes look awesome!

Reply 1 of 15, by kixs

User metadata
Rank l33t
Rank
l33t

That means some original sectors were bad and were replaced with spares that are located elsewhere on the disk. So it takes a little longer to access them when doing sequential reading/writing.

I wouldn't worry about it. But keep in mind that the HDD is failing and don't use it for anything special.

Requests here!

Reply 2 of 15, by Ryccardo

User metadata
Rank Member
Rank
Member

Speaking only for Victoria and MHDD as it's the ones I know:

When doing a surface test (whether read only, fully destructive, or "read + erase delays") you see a grid with access times, normally (for such a relatively modern disk) it will be the fastest rank and jump up by one every now and then while switching cylinders;
If a new bad sector is discovered while reading, the disk (unless configured otherwise, as typical for DVR and RAID designated models) will try to reread the sector on its own, which will show up in the grid;
In most firmwares only WRITING to a bad sector will cause it to be remapped (if there are still some spare sectors available, and if the physical sector is really bad), which as Kixs said will still cause a performance penalty but hopefully not as much as the multiple read attempts, so "erase delays" just means "write sectors that are arbitrarily too slow" and is implied in a full disk write test 😀

Depending on the exact drive model there may be an arguably superior option to remapped sectors: to run a self-scan (think of it as a "medium level format" as sector markers, but not the servo reference, are recreated), completely leaving out any unwritable ones - tedious at best, requiring multi-thousand-dollars professional gear at worst 😀 🙁- look on HDDOracle for more about this if you're desperate enough or just want to learn more!

Have you given a look at the SMART values yet?

Reply 3 of 15, by verysaving

User metadata
Rank Newbie
Rank
Newbie

What's the model of the drive?
Post a SMART log of it, just to see its health.
If there are some reallocated sectors none of the
software you use can't deal with them.
But since it's a SEAGATE, you could have some luck
by connecting it to a serial terminal and checking
its internal defect lists.

take a look here :

http://www.hddoracle.com/viewtopic.php?f=6&t=193
http://www.hddoracle.com/viewtopic.php?f=46&t=192

Reply 4 of 15, by momaka

User metadata
Rank Member
Rank
Member

40 GB Seagate HDD you say? Is it a Barracuda 7200.7 or Barracuda ATA IV or V model by any chance? If yes, I wouldn't worry about those bad sectors at all.
I have numerous Seagate IDE HDDs from these old models, and a good few of them have at least one bad sector here and there. The worst one, a ST340016A (Barracuda ATA IV, if I'm not mistaken) has racked up ~38 bad sectors up to date. Started with 36 bad sectors when I got it 2nd hand back in 2010. I used it for a good few years and it never had any issues or slowdowns. Only after parking the system / PC it was in for 7 years and then pulling it out of storage did the bad sector count increase. No data loss as far as I can tell, though. Given the HDD is 14 years older now than when I got it and that it still works perfectly fine, I'd say that's still a very good track record, even with the few bad sectors.

When I worked in IT back around 2010-2012, I ran across a lot of 7200.7 models (both IDE and SATA). I can say that in at least 25% of these, I saw drives with bad sectors. But here's the kicker - all of these HDDs continued to work fine and eventually all of the PCs they were in were retired due to old age rather than failed drive. The other HDDs I frequently ran across were the WD800xx models (IDE and SATA again). In contrast, these rarely had bad sectors. And what's even worse, they regularly failed "out of the blue" with a click-of-death. So IME, they were worse overall than the equivalent Seagate HDDs of the era. At least the Seagate HDDs would start racking up a lot of bad sectors and get slow if they were to fail (did see a few fail, but much less often compared to the Western Digital's.)

Anyways, that's my experience with these older IDE/SATA HDDs.
Now, if this was one of the newer SATA Seagate HDDs from the 7200.10, 7200.11, 7200.12, or newer series, I'd keep an eye on those bad/slow sectors a lot more often. With these newer series, once a few appear, it's just a matter of time until more appear. Sometimes the reallocated / pending sector count will settle, and sometimes it work. With one 7200.12 drive I have, the bad sector count has settled (well, more or less) to around ~750 reallocated sectors and ~1300 pending. 😮: That drive also still works OK, but I sometimes get data corruption when I do large file compression on one of the 3 partitions. I imagine the magnetic medium on the actual disk where that partition is located is going / gone weak. Nothing I can do about it, though. I just use that HDD with a lot of caution and don't put anything important on it. Still has some functional life left, so why not. 😉
...

So TL:DR: don't worry about the few slow/bad sectors on that old Seagate HDD. Just keep on using it and keep any eye on the SMART logs from time to time to see if the Reallocated Sector Count or Pending Sector Count starts increasing very quickly with use. If it doesn't, the HDD should be fine to use (even long term.)

Reply 5 of 15, by CYRIX

User metadata
Rank Newbie
Rank
Newbie
kixs wrote on 2024-08-18, 18:30:

That means some original sectors were bad and were replaced with spares that are located elsewhere on the disk. So it takes a little longer to access them when doing sequential reading/writing.

I wouldn't worry about it. But keep in mind that the HDD is failing and don't use it for anything special.

As far as I understand, DELAYED sectors actually mean that these were actually ex BAD sectors that had already been replaced with spare sectors from the spare area, and because of that they are slow to access... and maybe that's why they're called DELAYED sectors.
Did I understand correctly?

When all computers were pizza boxes , I wanted a tower case. Now that all computers are towers, the classic pizza-boxes look awesome!

Reply 6 of 15, by Horun

User metadata
Rank l33t++
Rank
l33t++

Yeah HD REGENERATOR does not totally describe what "Delayed sectors" are from what I can find.
This quote makes me think of scam stuff: "HDD Regenerator is a disk repair program (hence the name) that can repair damaged drives without affecting or altering existing data. As a result, the previously unreadable data in the bad sectors becomes accessible." All it does is move the indexed sector to end of drive (making drive size smaller), but does it really read the data from the bad sector proper?
There is no easy way to do that unless you do a low level nibble read of an HD's bad sectors to get the data back and even that is less than 50% possible..forensic techs sometimes spend weeks trying to recover any data from damaged sectors and this app says it can ?
I would use a Seagate tool to scan and fix/reallocate if needed since it a Seagate drive.
Sorry just my old age and pennies worth 😁

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 7 of 15, by Rwolf

User metadata
Rank Member
Rank
Member

These sector errors can also be caused by unexpected power-outages during a write operation. In that case there is nothing wrong with the disk surface, its just a half-written sector with a bad checksum, and the disc then refuses to read it and just hangs on reading until the sector is remapped or overwritten with good data.

(I had a bunch of these coming from some systems testing where the testers pulled the power randomly to check recovery operations.)

Reply 9 of 15, by analog_programmer

User metadata
Rank Oldbie
Rank
Oldbie
Horun wrote on 2024-08-19, 03:29:

There is no easy way to do that unless you do a low level nibble read of an HD's bad sectors to get the data back and even that is less than 50% possible..forensic techs sometimes spend weeks trying to recover any data from damaged sectors and this app says it can ?

Actually this app successfully revives some kind of marked as "bad" sectors, but I'm sure it can not bring back the remapped ones. I'm still using it on old mechanical drives.

In this particular case it seems like low level format will be more suitable instead of HDD Regenerator.

from СМ630 to Ryzen gen. 3
engineer's five pennies: this world goes south since everything's run by financiers and economists
this isn't voice chat, yet some people, overusing online communications, "talk" and "hear voices"

Reply 10 of 15, by verysaving

User metadata
Rank Newbie
Rank
Newbie

HDD Regenerator (aka HDD Degenerator) is just a joke, it just relocates
some corrupted sector to defect list (GLIST) like other tools like
MHDD, Victoria o HDAT2 can do without claiming of "regenerating" anything.
Just take a disk with some bad sector, take a look at the SMART attributes

#5 "relocated sector count" ,
#197 "current pending sectors",

before and after running the program, you will see clearly that one
will increase and the other will decrease.

The Seagate LLF tool it's just useful if you need to RMA the drive,
it's nothing more than the other "zerofillers" you may find on the web,
that actually write just zeroes on the whole LBA space.

You CAN'T Low Level Format those drives, but on such old (SEAGATE) disk
the firmware of the drive containt FULL code to selfscan itself (removed
by Seagate from 7200.11 and newer drive family) and, if it has no physical
issue you will get something like a "brand new" drive.

Rwolf pointed out what the real problem on the drive may be,
just few scattered bad spots on the disk caused by improper power
cut during write operation. I've seen this hundreds of time ...

Another attribute to look for is #199, "Ultra DMA CRC errors",
bad connections at cable level may cause this attribute to increase
and "fool" the firmware to reallocate lot of sectors that are
absolutely fine.

As Momaka said, if it's just 3 relocated sectors the disk is fine.

Without a look at SMART attributes I think other discussion is pointless.

Reply 11 of 15, by analog_programmer

User metadata
Rank Oldbie
Rank
Oldbie
verysaving wrote on 2024-08-19, 23:39:
HDD Regenerator (aka HDD Degenerator) is just a joke, it just relocates some corrupted sector to defect list (GLIST) like other […]
Show full quote

HDD Regenerator (aka HDD Degenerator) is just a joke, it just relocates
some corrupted sector to defect list (GLIST) like other tools like
MHDD, Victoria o HDAT2 can do without claiming of "regenerating" anything.
Just take a disk with some bad sector, take a look at the SMART attributes

#5 "relocated sector count" ,
#197 "current pending sectors",

before and after running the program, you will see clearly that one
will increase and the other will decrease.

I've never noticed such a behavior. And if this is true, why it takes hours to scan HDD with volume of couple of gigabytes? In this case it just can take already pending sectors and relocate them while imitating scan process and will be advertised as "the fastest scanning tool".

What about those old BIOSes with LLF tools? Are they just zero-filling tolls?

from СМ630 to Ryzen gen. 3
engineer's five pennies: this world goes south since everything's run by financiers and economists
this isn't voice chat, yet some people, overusing online communications, "talk" and "hear voices"

Reply 12 of 15, by Ryccardo

User metadata
Rank Member
Rank
Member
analog_programmer wrote on 2024-08-20, 05:10:

I've never noticed such a behavior. And if this is true, why it takes hours to scan HDD with volume of couple of gigabytes? In this case it just can take already pending sectors and relocate them while imitating scan process and will be advertised as "the fastest scanning tool".

Mainly because there's no standard way to get a list of bad sectors at the block device (or lower) level 😀 , and SMART (where it exists and works) just counts - relatively independently - how many were identified by the drive...

analog_programmer wrote on 2024-08-20, 05:10:

What about those old BIOSes with LLF tools? Are they just zero-filling tolls?

Generally they're the real deal, but you have to know the history of HDDs to understand why they're useless on most every 40 GB disk:

  • The first HDD option (for a PC with a 16-bit ISA bus - let's ignore what came first as it's not terribly relevant and completely incompatible) is the original MFM controller for the IBM 5170, which connects to 1 or 2 HDDs that, crucially, are like larger and nonremovable floppy drives: linear seeking, it can only know when it's on track (cylinder) 0 and all others can only be accessed by relative movement, not LLF'd out of the factory because different controllers only supported different geometries, can't remap sectors (that's why some filesystems, including FAT12/16/32 when you do a "full format" or a Scandisk surface test or a Chkdsk /R, can blacklist parts of the disk on their own "high level": on a modern drive that can remap you'd need to run out of remappable sectors first for such a write+verify, pass/fail test to detect something)
  • Then the RLL controllers, with higher "bit clock", can get more out of (better quality examples of) the same kind of "passive" disks, but more critically they're accessed in the same way (same customary port addresses and meaning)
  • Then IDE drives, meaning literally Integrated Drive Electronics - the combination of a "passive" disk with a controller in a compact package, but also almost 100% equivalent

And a (consumer grade) low level formatter in the strict sense (something that takes a disk, even if completely magnetically neutral, and creates the sector headers and therefore the C/H/S geometry) only exists for these (equivalent) drives!

After that, still (correctly) under the IDE category, "smart" drives started coming out with zoned recording and the like, meaning even the "physical" CHS values delivered on the ISA bus are virtual (because the real geometry isn't constant); with factory-prerecorded servo markers, enabling higher head positioning precision and therefore density (at the cost of the user, or even full time pro without actual factory equipment, not being able to LLF them anymore), and later of "smarter" integrated controllers with things like bad sector remapping (which introduces a "medium level format" mapping not necessarily linearly CHS/LBA sectors to physical ones, which is (re)created with the "self-scan" mentioned above) and SMART and ATA security (which I bet has caused more bad than good) and AAM and APM, etc etc 😀

So, most every IDE/SATA drive designed after give or take 1996 will reject the inapplicable "format track" commands a LLF program (whether that's a file, option ROM, or built into the BIOS) issues, and for extra style points a few around that time will accept them and completely corrupt themselves 😁

And if the above was too simple - in modern computing there's sometimes a 4th layer between the "medium" and the LBA, with disks that can be configured to have different sizes of LBA sectors! 😁

Reply 13 of 15, by hyoenmadan

User metadata
Rank Member
Rank
Member

These days the HDD repair and servicing market is almost like the Smartphone repair market: Lots of incompatible, propietary and highly locked and encrypted tools paired with propietary FPGA specific hardware. The specific knowledge to make these tools, the schematics of the devices and module-data/roms/special bootloaders which your specific tool brand provides... Most of the time comes from hacked/stolen data from the factories and R&D labs in Asia and other 3rd world places where the devices are assembled, and then sold to you as an expensive product made by them, just as with the Smartphone "fix tools". And ofc once you sign with them to get their "tools" they will always keep a track on you so you can't leak their data online. Their hardware has to be returned to them if broken, or if you will not use it anymore.

I always wondered why Open Source communities don't pursue an alternative tool to the proprietaries like ACELabs or the ones on China (which looks like hacked/stolen themselves a version of the ACELabs kit and made one which only recently has become usable enough). Most probably is the special data available for the drives isn't clean, or doesn't have a clean origin itself, so it can't be used legally. And unfortunately that's the only data available.

Reply 14 of 15, by verysaving

User metadata
Rank Newbie
Rank
Newbie

@analog_programmer
What drives did you scan? some old disks are actual very slow
or can be degraded so their scanning could take lot of time.
Try making a scan of the disks with MHDD and take note of their
real speed.

LLF software works properly only on older MFM/RLL disk, the
"SEAGATE 40 GB" that started this thread must be of course an
IDE drive.

Reply 15 of 15, by analog_programmer

User metadata
Rank Oldbie
Rank
Oldbie
Ryccardo wrote on 2024-08-20, 17:32:

Mainly because there's no standard way to get a list of bad sectors at the block device (or lower) level 😀 , and SMART (where it exists and works) just counts - relatively independently - how many were identified by the drive...

Ok, this makes sense, but still it's not necessary every pending or OS marked "bad" sector to be relocated as a "cure". Probably after some unsuccessful reading/writing attempts of bad sector it's relocated as a last option. A couple of times I was able to fully restore some damaged files due to bad sectors with HDD Regenerator and this doesn't work through relocating of bad sectors.

I think HDD Regenerator uses similar techniques as VGACopy (working DOS software for floppy disks with magnetically bad sectors).

Maybe some people mess HDD Regenerator with Victoria. I think the latter is the c*ap that always relocates "bad"/pending sectors.

Ryccardo wrote on 2024-08-20, 17:32:
Generally they're the real deal, but you have to know the history of HDDs to understand why they're useless on most every 40 GB […]
Show full quote

Generally they're the real deal, but you have to know the history of HDDs to understand why they're useless on most every 40 GB disk:

  • The first HDD option (for a PC with a 16-bit ISA bus - let's ignore what came first as it's not terribly relevant and completely incompatible) is the original MFM controller for the IBM 5170, which connects to 1 or 2 HDDs that, crucially, are like larger and nonremovable floppy drives: linear seeking, it can only know when it's on track (cylinder) 0 and all others can only be accessed by relative movement, not LLF'd out of the factory because different controllers only supported different geometries, can't remap sectors (that's why some filesystems, including FAT12/16/32 when you do a "full format" or a Scandisk surface test or a Chkdsk /R, can blacklist parts of the disk on their own "high level": on a modern drive that can remap you'd need to run out of remappable sectors first for such a write+verify, pass/fail test to detect something)
  • Then the RLL controllers, with higher "bit clock", can get more out of (better quality examples of) the same kind of "passive" disks, but more critically they're accessed in the same way (same customary port addresses and meaning)
  • Then IDE drives, meaning literally Integrated Drive Electronics - the combination of a "passive" disk with a controller in a compact package, but also almost 100% equivalent

And a (consumer grade) low level formatter in the strict sense (something that takes a disk, even if completely magnetically neutral, and creates the sector headers and therefore the C/H/S geometry) only exists for these (equivalent) drives!

After that, still (correctly) under the IDE category, "smart" drives started coming out with zoned recording and the like, meaning even the "physical" CHS values delivered on the ISA bus are virtual (because the real geometry isn't constant); with factory-prerecorded servo markers, enabling higher head positioning precision and therefore density (at the cost of the user, or even full time pro without actual factory equipment, not being able to LLF them anymore), and later of "smarter" integrated controllers with things like bad sector remapping (which introduces a "medium level format" mapping not necessarily linearly CHS/LBA sectors to physical ones, which is (re)created with the "self-scan" mentioned above) and SMART and ATA security (which I bet has caused more bad than good) and AAM and APM, etc etc 😀

So, most every IDE/SATA drive designed after give or take 1996 will reject the inapplicable "format track" commands a LLF program (whether that's a file, option ROM, or built into the BIOS) issues, and for extra style points a few around that time will accept them and completely corrupt themselves 😁

And if the above was too simple - in modern computing there's sometimes a 4th layer between the "medium" and the LBA, with disks that can be configured to have different sizes of LBA sectors! 😁

Thanks for this detailed answer. This explains why BIOS LLF tools are missing from mid '90s and newer motherboards.

verysaving wrote on 2024-08-20, 23:25:
@analog_programmer What drives did you scan? some old disks are actual very slow or can be degraded so their scanning could take […]
Show full quote

@analog_programmer
What drives did you scan? some old disks are actual very slow
or can be degraded so their scanning could take lot of time.
Try making a scan of the disks with MHDD and take note of their
real speed.

For the last year I've scanned with HDD Regenerator all of my old "retired" mechanical IDE ATA HDDs with different volumes from 250 MB up to 80 GB. And the process is slow, much slower compared to "surface scan" in DOS/Windows Scandisk. Fortunately these drives are in good shape and I use them for test drives with old motherboards.

But I know and use this HDD Regenerator software for at least two decades and in the past it saved me several times from corrupted system and other files on mechanical HDDs after unexpected power failures, thus saving me OS reinstallations and some valuable personal data. This is something that doesn't always happen with Scandisk's full+surface scan.

verysaving wrote on 2024-08-20, 23:25:

LLF software works properly only on older MFM/RLL disk, the
"SEAGATE 40 GB" that started this thread must be of course an
IDE drive.

Nope. LLF tool from older BIOSes (some s.7 boards have it) also works with some of the smaller and older IDE ATA HDDs. See Ryccardo's post for details.

from СМ630 to Ryzen gen. 3
engineer's five pennies: this world goes south since everything's run by financiers and economists
this isn't voice chat, yet some people, overusing online communications, "talk" and "hear voices"