Update: After running the tests I described, I ran "write zeros to drive" from the WDigital Diagnostic Tool, rebooted and proceeded to format c: again, this time with no "trying to recover allocation unit" issues at all. I'm even more confused now!

Originally, FORMAT used to detect and sort out bad sectors on MFM/RLL drives and old IDE drives.
Once found, bad sectors were market as bad in FAT and ignored. Just like with floppy disks.

However, modern HDDs do hide their bad sectors automatically once they do encounter them.
They do have spare sectors for compensation, which are only used up in such a situation.

That's why a bad sector report is so fatal nowadays.
The report means that the drive's on-board computer failed to compensate.
The cause could be loose particles from the surfaces.
If they aren't caught by the filters in the HDD anymore, they may hit the surface again and again, causing more particles to fly around.

Edit: That's just an idea, of course. Maybe it's also the drive's mechanic, the lubrication etc.
If you like, let a diagnostic program run multiple read/write tests, so the actuator gets moved a lot.
This may solve issues with sticky lubrication.
CheckIt for DOS can do that in the HDD test menu.
However, old CheckIt is limited to FAT16/2GB partitions maybe.
Not sure if it's really helpful thus. 🤷‍♂️

Jo22 wrote on 2022-08-18, 23:52:

Originally, FORMAT used to detect and sort out bad sectors on MFM/RLL drives and old IDE drives. Once found, bad sectors were m […]
Show full quote

Originally, FORMAT used to detect and sort out bad sectors on MFM/RLL drives and old IDE drives.
Once found, bad sectors were market as bad in FAT and ignored. Just like with floppy disks.

However, modern HDDs do hide their bad sectors automatically once they do encounter them.
They do have spare sectors for compensation, which are only used up in such a situation.

That's why a bad sector report is so fatal nowadays.
The report means that the drive's on-board computer failed to compensate.
The cause could be loose particles from the surfaces.
If they aren't caught by the filters in the HDD anymore, they may hit the surface again and again, causing more particles to fly around.

Edit: That's just an idea, of course. Maybe it's also the drive's mechanic, the lubrication etc.
If you like, let a diagnostic program run multiple read/write tests, so the actuator gets moved a lot.
This may solve issues with sticky lubrication.
CheckIt for DOS can do that in the HDD test menu.
However, old CheckIt is limited to FAT16/2GB partitions maybe.
Not sure if it's really helpful thus. 🤷‍♂️

So basically FORMAT detected and proceeded to ignore from that point on the bad sectors? Even when I ran the tests after and "wrote zeros to drive" on the diagnostic tool? Wouldn't that "reset" the disk to factory per se, making it recognize the bad sectors again? That's why I was confused that there were no errors when I ran FORMAT again. Won't the ignored bad sectors just keep degrading the drive?

Lubrication / mechanic stiffness can make sense, I guess I'll have to run more tests. If you have more suggestions as to diagnostic programs for DOS/Win98SE I'm all ears. This wouldn't be such a pain if I didn't need to install lots of stuff, and I feel as I'm building this on sand with bad sectors possibly "festering". I'm not even sure the risk is worth it. Thanks for your help

With a "Smart" HD once a sector is classed as bad it is never put back into use, unless a true low-level format is done which you should never do on any IDE drive as it can actually ruin it.
Using a tool to write all 0 or 1 does not low level, they just write to the current good sectors...
The platters may not have even coatings and certain sectors are read more than others like the FAT for Files, Directories and Partition tables so they can fail quicker but the remaining sectors can be just fine, one of the reasons for re-allocating sectors. IIRC
Yes lubrication/mechanic stiffness does come into play when drives sit too long unpowered, also the electronics (electrolytic caps) usually need some healing which can happen within minutes or short hours so that the calibrates properly.
Probably am miss stating a few things but finding a few bad sectors is not always a sign of quick eminent failure but is something to keep an eye on.
Hopefully it is not media flake which Jo22 mentioned, that can cause lots of problems 😀

Horun wrote on 2022-08-19, 01:20:

With a "Smart" HD once a sector is classed as bad it is never put back into use, unless a true low-level format is done which yo […]
Show full quote

With a "Smart" HD once a sector is classed as bad it is never put back into use, unless a true low-level format is done which you should never do on any IDE drive as it can actually ruin it.
Using a tool to write all 0 or 1 does not low level, they just write to the current good sectors...
The platters may not have even coatings and certain sectors are read more than others like the FAT for Files, Directories and Partition tables so they can fail quicker but the remaining sectors can be just fine, one of the reasons for re-allocating sectors. IIRC
Yes lubrication/mechanic stiffness does come into play when drives sit too long unpowered, also the electronics (electrolytic caps) usually need some healing which can happen within minutes or short hours so that the calibrates properly.
Probably am miss stating a few things but finding a few bad sectors is not always a sign of quick eminent failure but is something to keep an eye on.
Hopefully it is not media flake which Jo22 mentioned, that can cause lots of problems 😀

The thing is, in the first time, FORMAT only manifested (albeit multiple times) the "trying to recover allocation unit" so it's a vague message. How can I know for sure in the future that the message was referring to bad sectors? I just don't want to reinstall and config everything again next year for example!

Horun wrote on 2022-08-19, 01:20:

With a "Smart" HD once a sector is classed as bad it is never put back into use, unless a true low-level format is done which you should never do on any IDE drive as it can actually ruin it.

good luck doing low level format of IDE drive 😀 only the very first pre 1990 drives allowed it at all - ones with stepper motor head mechanism or dedicated servotrack platter. Modern drives are pseudo hard sectored, one of the manufacturing steps is recording special track/sector markers on platters https://en.wikipedia.org/wiki/Servowriter

> Modern HDDs generate position feedback signals from special magnetic patterns called servo patterns which are written in designated areas on the disk surface known as servo sectors. The generated feedback signals are called position error signals (PES)… the servo sectors are created at the time of manufacturing and are never overwritten or erased. Then, the closed-loop servomechanism decodes the position information written in these sectors to accomplish adequate control tasks.

IBM patent US7079347B2 “Method and apparatus for providing a marker for adaptive formatting via a self-servowrite process” from 2002:
>Traditional servo writing has been performed in a clean room environment with external sensors invading the head disk assembly to provide the precise angular and radial position information to write the servo patterns. For example, an external clock head was typically disposed on the disk outer diameter.
>Currently, determining the number of available tracks from the self servowrite process requires a trial and error process or requires that the track count be sent ahead from the servowriter to the function test station.

andre_6 its possible your disk:
-indeed had bad sectors
-format marked them as bad in the FAT structure
-ScanDisk scan skips FAT sectors market bad and only tests healthy ones, thus showed no errors
-zero filling the drive finally forced the Drive firmware to remap bad sectors, or drive already remapped them during the format. Modern drives have dedicated P-List/G-List/T-List tables for that purpose https://datarecovery.com/rd/what-are-p-lists-and-g-lists/
-now the logical disk surface is all healthy again

plug it into windows computer and run https://hddscan.com or even https://crystalmark.info/en/software/crystaldiskinfo/
the relevant parameters are
-Reallocation Sector Count
-Reallocation Event Count
-Current Pending Errors Count

hddscan also offers Short and Extended SMART tests. Correction. I was thinking of GSmartControl
https://gsmartcontrol.shaduri.dev/screenshots
It lets you view selt-test logs, Sometimes scammers selling smart cleaned drives will forget to erase logs and you can see funny business like past Self-Test run at 60000 hours on a "0 hours" drive 😀

PS theer are hacks to Erase modern drives Smart status. This is how there are tons of "never used" 2-4TB drives on AMAZON with 2012 manufacturing date. You buy one, plug into the computer and it indeed shows 0 read/written and 0 hours powered on .. except its a server pull with >60000 hours spend in hot server rack. Extended Smart test is sometimes able to show leftovers from previously run deep smart tests reveling the scam.

rasz_pl wrote on 2022-08-19, 02:24:

good luck doing low level format of IDE drive :) only the very first pre 1990 drives allowed it at all - ones with stepper motor […]
Show full quote

Horun wrote on 2022-08-19, 01:20:

With a "Smart" HD once a sector is classed as bad it is never put back into use, unless a true low-level format is done which you should never do on any IDE drive as it can actually ruin it.

good luck doing low level format of IDE drive 😀 only the very first pre 1990 drives allowed it at all - ones with stepper motor head mechanism or dedicated servotrack platter. Modern drives are pseudo hard sectored, one of the manufacturing steps is recording special track/sector markers on platters https://en.wikipedia.org/wiki/Servowriter

> Modern HDDs generate position feedback signals from special magnetic patterns called servo patterns which are written in designated areas on the disk surface known as servo sectors. The generated feedback signals are called position error signals (PES)… the servo sectors are created at the time of manufacturing and are never overwritten or erased. Then, the closed-loop servomechanism decodes the position information written in these sectors to accomplish adequate control tasks.

IBM patent US7079347B2 “Method and apparatus for providing a marker for adaptive formatting via a self-servowrite process” from 2002:
>Traditional servo writing has been performed in a clean room environment with external sensors invading the head disk assembly to provide the precise angular and radial position information to write the servo patterns. For example, an external clock head was typically disposed on the disk outer diameter.
>Currently, determining the number of available tracks from the self servowrite process requires a trial and error process or requires that the track count be sent ahead from the servowriter to the function test station.

andre_6 its possible your disk:
-indeed had bad sectors
-format marked them as bad in the FAT structure
-ScanDisk scan skips FAT sectors market bad and only tests healthy ones, thus showed no errors
-zero filling the drive finally forced the Drive firmware to remap bad sectors, or drive already remapped them during the format. Modern drives have dedicated P-List/G-List/T-List tables for that purpose https://datarecovery.com/rd/what-are-p-lists-and-g-lists/
-now the logical disk surface is all healthy again

plug it into windows computer and run https://hddscan.com or even https://crystalmark.info/en/software/crystaldiskinfo/
the relevant parameters are
-Reallocation Sector Count
-Reallocation Event Count
-Current Pending Errors Count

hddscan also offers Short and Extended SMART tests.

PS theer are hacks to Erase modern drives Smart status. This is how there are tons of "never used" 2-4TB drives on AMAZON with 2012 manufacturing date. You buy one, plug into the computer and it indeed shows 0 read/written and 0 hours powered on .. except its a server pull with >60000 hours spend in hot server rack. Extended Smart test is sometimes able to show leftovers from previously run deep smart tests reveling the scam.

Thank you very much for your thorough reply, so much to learn. In any case, will the disk remain usable as normal or will it deteriorate faster than normal or perform at a lower speed? When it comes to it, I prefer to just get another one in order to have peace of mind, and keep this one as a backup one for some basic stuff (i.e. not installing a full catalog of Win98 era stuff for days only to be forced to start all over again....yet again!)

check the smart, if its a single reallocation and wont grow after couple of months you are fine
its also possible format error was about some FAT logic having nothing to do with disk surface health, and there are no reallocations at all

rasz_pl wrote on 2022-08-19, 03:19:

check the smart, if its a single reallocation and wont grow after couple of months you are fine
its also possible format error was about some FAT logic having nothing to do with disk surface health, and there are no reallocations at all

I'll have to find an XP PC for HDD Scan, in any case I'll keep it as an basic emergency option. The first FORMAT command didn't take *that* long to complete, but the many "trying to recover allocation unit" messages were very discouraging. It's nice to learn that an issue like this isn't necessarily doom and gloom, thanks for your valuable input and all users before you

what's the model of the HD, I mean the string like "WD1200BB-22GUC0" on the label.
Also a SMART log would be very useful to check the actual health of the drive.

To get it and test the disk properly, you can download MHDD from here and make
a bootable disk to start it.

http://hddguru.com/software/2005.10.02-MHDD/

if you need DOS then also https://www.hdsentinel.com/hard_disk_sentinel_dos.php "optionally the complete S.M.A.R.T. details if /SMART switch is used"

Maybe some one can help explain the data from a SMART drive "read" in a simple manner because sometimes the data delivered by some SMART tools can be hard to understand.
I understand most of it but might be wrong in a few areas.. Example from CrystalDiskInfo.
The "raw" data in Hex is the important numbers as to current status, not the "Current, Worst, Threshold" as those are from the HD's SMART rom table as maximums/minimums and are not the current recorded exact condition of a certain Attribute, which is written to a eprom part in Hex. One of the DOS based I tried in the past also gives the data without a good translation.
So if the raw data for say Reallocated Sectors Count has a full readout of Current: 199, Worst 199, Threshold 140, RAW 00000004. The actual relocated sectors is 4 Hex which is also 4 decimal for 4 re-allocated sectors.
This is on one of my older drives and has 60800+ hours on it and the ReAllocation Counter RAW is 3 (the 3 times it corrected errors fixing 4 sectors).
Another drive nearly new has something similar with C 100, W 100, T 10 and RAW 00000000 so no relocated sectors yet....
If that does not sound correct please explain, thanks.

verysaving wrote on 2022-08-20, 00:53:

what's the model of the HD, I mean the string like "WD1200BB-22GUC0" on the label.
Also a SMART log would be very useful to check the actual health of the drive.

rasz_pl wrote on 2022-08-20, 01:44:

if you need DOS then also https://www.hdsentinel.com/hard_disk_sentinel_dos.php "optionally the complete S.M.A.R.T. details if /SMART switch is used"

Horun wrote on 2022-08-20, 03:24:

Maybe some one can help explain the data from a SMART drive "read" in a simple manner because sometimes the data delivered by some SMART tools can be hard to understand.
I understand most of it but might be wrong in a few areas.. Example from CrystalDiskInfo.

So here is the report from HSDOS with SMART log, file attached

S.M.A.R.T.
1 : Raw Read Error Rate Threshold=51 Value=200 Worst=179 Data=000000000000 Status=OK Flags=Error-rate, Statistical, Critical
3 : Spin Up Time Threshold=21 Value=167 Worst=162 Data=000000001057 Status=OK Flags=Performance, Statistical, Critical
4 : Start/Stop Count Threshold=40 Value=94 Worst=94 Data=0000000019DF Status=OK Flags=Self Preserving, Event Count, Statistical
5 : Reallocated Sectors Count Threshold=140 Value=200 Worst=200 Data=000000000000 Status=OK Flags=Self Preserving, Event Count, Statistical, Critical
7 : Seek Error Rate Threshold=51 Value=200 Worst=200 Data=000000000000 Status=OK Flags=Error-rate, Statistical, Critical
9 : Power On Time Count Threshold=0 Value=77 Worst=77 Data=0000000042DD Status=OK (Always passing) Flags=Self Preserving, Event Count, Statistical
10 : Spin Retry Count Threshold=51 Value=100 Worst=100 Data=000000000000 Status=OK Flags=Event Count, Statistical, Critical
11 : Drive Calibration Retry Count Threshold=51 Value=100 Worst=100 Data=000000000000 Status=OK Flags=Event Count, Statistical, Critical
12 : Drive Power Cycle Count Threshold=0 Value=94 Worst=94 Data=000000001975 Status=OK (Always passing) Flags=Self Preserving, Event Count, Statistical
194: Disk Temperature Threshold=0 Value=117 Worst=99 Data=000000000021 Status=OK (Always passing) Flags=Self Preserving, Statistical
196: Reallocation Event Count Threshold=0 Value=200 Worst=200 Data=000000000000 Status=OK (Always passing) Flags=Self Preserving, Event Count, Statistical
197: Current Pending Sector Count Threshold=0 Value=200 Worst=200 Data=000000000000 Status=OK (Always passing) Flags=Event Count, Statistical
198: Off-Line Uncorrectable Sect... Threshold=0 Value=200 Worst=200 Data=000000000000 Status=OK (Always passing) Flags=Event Count, Statistical
199: Ultra ATA CRC Error Count Threshold=0 Value=200 Worst=200 Data=000000000013 Status=OK (Always passing) Flags=Error-rate, Statistical
200: Write Error Rate Threshold=51 Value=200 Worst=200 Data=000000000000 Status=OK Flags=Error-rate, Critical

The SMART says there is nothing wrong with the drive based on the important ones:
1 : Raw Read Error Rate = 0
5 : Reallocated Sectors Count = 0
7 : Seek Error Rate = 0
196: Reallocation Event Count = 0 (should be 0 if 5 is 0)
198: Off-Line Uncorrectable Sect = 0
Just for you info:
9 : Power On Time = 17117 Hrs
12 : Drive Power Cycle Count = 6517 times

Horun wrote on 2022-08-20, 21:03:

The SMART says there is nothing wrong with the drive based on the important ones: 1 : Raw Read Error Rate = 0 5 : Reallocated Se […]
Show full quote

The SMART says there is nothing wrong with the drive based on the important ones:
1 : Raw Read Error Rate = 0
5 : Reallocated Sectors Count = 0
7 : Seek Error Rate = 0
196: Reallocation Event Count = 0 (should be 0 if 5 is 0)
198: Off-Line Uncorrectable Sect = 0
Just for you info:
9 : Power On Time = 17117 Hrs
12 : Drive Power Cycle Count = 6517 times

Thanks for the feedback, that's encouraging, if anyone spots anything else of note please feel free to point it out, I'm trying to hold out for all the information I can before embarking on the long program install journey

Attribute 199: Ultra ATA CRC Error Count is not 0,
so check your IDE cables and the connection to
the drive and to the motherboard.
Loose cable and bad connections were pretty common
with IDE drives, altough even with serial drive
they are not rare!

Other attributes indicates an healty drive, but
I would make a full scan of the drive just to be
sure.
And at the end would check that attribute 199 has not
increased.

If there's no data on the disk, a full erase and
then a full scan with MHDD wuold be ideal.

Good thought verysaving ! I figured those CRC errors could board/controller/bios related since no one knows what it was connected to before or what is connected to now ;p

verysaving wrote on 2022-08-21, 00:32:

If there's no data on the disk, a full erase and
then a full scan with MHDD wuold be ideal.

Horun wrote on 2022-08-21, 01:13:

Good thought verysaving ! I figured those CRC errors could board/controller/bios related since no one knows what it was connected to before or what is connected to now ;p

Was caught off guard by how long it took to erase the disk, so here it is:

21.08.2022 15:24:53 MHDD>FULLSCAN
21.08.2022 15:24:56
21.08.2022 15:24:56 MHDD>SCAN
21.08.2022 15:27:08 Scan started
21.08.2022 15:27:08 MODE: IDE
21.08.2022 15:27:09 Device: WDC WD1200BB-00GUA0
21.08.2022 15:27:09 F/W: 08.02D08
21.08.2022 15:27:09 SN: WD-WMALA1456203
21.08.2022 15:27:09 -------------------------------
21.08.2022 15:27:09 Lap : 1
21.08.2022 15:27:09 LBA scan: 0 to 234441647
21.08.2022 16:04:34 Time spent: 00:37:23
21.08.2022 16:04:34 Blocks < 3ms = 798971
21.08.2022 16:04:34 Blocks < 10ms = 120285
21.08.2022 16:04:34 Blocks < 50ms = 124
21.08.2022 16:04:34 Blocks < 150ms = 0
21.08.2022 16:04:34 Blocks < 500ms = 0
21.08.2022 16:04:34 Blocks > 500ms = 0
21.08.2022 16:04:34 No warnings, no errors
21.08.2022 16:04:34 Done
21.08.2022 16:33:49
21.08.2022 16:33:49 MHDD>SMART ATT
21.08.2022 16:33:49 Getting SMART attributes...
21.08.2022 16:33:49 SMART READ ATTRIBUTES
21.08.2022 16:33:49 SMART SWITCH: ON
21.08.2022 16:33:50 SMART SWITCH: OFF
21.08.2022 16:33:50 HDD: WDC WD1200BB-00GUA0; FW: 08.02D08; SN: WD-WMALA1456203
21.08.2022 16:33:50 --------------------------------------------------------
21.08.2022 16:33:50 SMART attributes:
21.08.2022 16:33:50 Name Val Worst Raw
21.08.2022 16:33:50 Att # 1 : Read error rate : 200 179 0
21.08.2022 16:33:50 Att # 3 : Spin up time : 167 162 4166
21.08.2022 16:33:50 Att # 4 : Number of spin-up times : 94 94 6625
21.08.2022 16:33:50 Att # 5 : Reallocated sectors count : 200 200 0
21.08.2022 16:33:50 Att # 7 : Seek error rate : 200 200 0
21.08.2022 16:33:50 Att # 9 : Power-on time : 77 77 17132
21.08.2022 16:33:50 Att # 10 : Spin-up retries : 100 100 0
21.08.2022 16:33:50 Att # 11 : Calibration retries : 100 100 0
21.08.2022 16:33:50 Att # 12 : Start/stop count : 94 94 6519
21.08.2022 16:33:50 Att # 194 : HDA Temperature : 107 99 43
21.08.2022 16:33:50 Att # 196 : Reallocate event count : 200 200 0
21.08.2022 16:33:50 Att # 197 : Current pending sectors : 200 200 0
21.08.2022 16:33:51 Att # 198 : Offline scan UNC sectors : 200 200 0
21.08.2022 16:33:51 Att # 199 : Ultra ATA CRC Error Rate : 200 200 19
21.08.2022 16:33:51 Att # 200 : Write error rate : 200 200 0

For comparison, here's the SMART before erase and scan:

21.08.2022 2:24:39 SMART attributes:
21.08.2022 2:24:39 Name Val Worst Raw
21.08.2022 2:24:39 Att # 1 : Read error rate : 200 179 0
21.08.2022 2:24:39 Att # 3 : Spin up time : 167 162 4166
21.08.2022 2:24:39 Att # 4 : Number of spin-up times : 94 94 6625
21.08.2022 2:24:39 Att # 5 : Reallocated sectors count : 200 200 0
21.08.2022 2:24:39 Att # 7 : Seek error rate : 200 200 0
21.08.2022 2:24:39 Att # 9 : Power-on time : 77 77 17117
21.08.2022 2:24:39 Att # 10 : Spin-up retries : 100 100 0
21.08.2022 2:24:39 Att # 11 : Calibration retries : 100 100 0
21.08.2022 2:24:39 Att # 12 : Start/stop count : 94 94 6519
21.08.2022 2:24:39 Att # 194 : HDA Temperature : 115 99 35
21.08.2022 2:24:39 Att # 196 : Reallocate event count : 200 200 0
21.08.2022 2:24:40 Att # 197 : Current pending sectors : 200 200 0
21.08.2022 2:24:40 Att # 198 : Offline scan UNC sectors : 200 200 0
21.08.2022 2:24:40 Att # 199 : Ultra ATA CRC Error Rate : 200 200 19
21.08.2022 2:24:40 Att # 200 : Write error rate : 200 200 0

So basically, am I in the clear to start installing stuff or would I have to be on top of this for some time to see if there are any changes?