Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > alt.comp.os.windows-10 > #186413

Re: chkdsk finds errors, fails to correct them

Show all headers | View raw

On Thu, 7/31/2025 1:04 PM, Jason wrote:
> Now and again, I run chkdsk on partitions to look for  
> trouble. Lately, on C:, chkdsk reports bitmap errors and 
> asks if I want to restart to fix them offline. I do that 
> but chkdsk again reports errors upon restarting. This 
> happened three times in a row. What's up and should I 
> worry? BTW, C: is a 1G SSD should that make any 
> difference.
> 

As you know, Microsoft on the one hand, claims the NTFS API
has not changed in years, thus "everything is compatible"
as they spin their new OSes. They refuse to change the NTFS
release number.

But, changes have been made, and they "aren't exactly compatible".
A glaring example, is the ability to make 2MB clusters in W10/W11,
which Windows 7 cannot read.

Some of the changes are "wear optimization" for SSDs. That means
fewer nuisance writes done by the OS. For example, the "Accessed"
time stamp would be modified on a HDD drive, on some file or other,
almost constantly. And that's because, in the HDD era, we didn't worry
about writes. The HDD has a TBW rating, which does imply a workload
sensitivity versus SKU, but that did not bother anyone at the time.

One of the optimizations was to stop maintaining $BITMAP on the SSD,
while the OS was running. A copy was stored in memory. This upset
Macrium at the time, which does disk verification by examining
structures on disk, before a backup is done. If Macrium had used
the memory copy of $BITMAP, that would not have happened and
"Macrium would not have noticed". But for Microsoft, this is why
you cannot go around arbitrarily changing things, because it
causes your "partners" to have to issue "emergency patches"
(6.3.1865 era) after you fiddle something on a Patch Tuesday.
The VirtualBox guys had to scramble at one point and release
two emergency patches, after Microsoft broke their stuff
(apparently without warning by the looks of it).

The visual symptoms of this, the "blowback" as it were, generally
only showed in the W10 release where it first showed up. Sufficient
bandaids were applied, that people stopped seeing sizing errors
caused by the handling of $BITMAP. For a short period of time,
you could not even trust the pie chart capacity in the partition
Properties dialog, because of this stuff.

*******

So now let us consider our maintenance possibilities. We know that,
quite frequently, the $BITMAP on the disk at rest is wrong. In other
environments, storage drivers must have been modified, to do a $BITMAP
calc while the file system was being mounted.

OK, so the CHKDSK on W10 likely does not care to fix this. Because
that version of CHKDSK knows about the "lazy evaluation" of $BITMAP.
There is no reason to fix it... because it will break again, surely ?

But Win7 still cares about the $BITMAP, and the help for the Win7 CHKDSK
is not going to offer any comments about the poor design choices in
Windows 10/11. Windows 7 didn't get the memo.

If you have a Win7 installer DVD (I bought one of those hologram DVDs
long ago), you can boot one of those, instead of installing, select
"Troubleshooting", and there should be a Command Prompt option in there.

   chkdsk /?

Look at the options carefully, as each generation of CHKDSK, not only
does it do the normal things, but occasionally there are a couple
"weird" options listed. Just the "regular" options will likely fix
this situation. But on maybe one occasion, I had to use one
of those "weird" options to fix an NTFS (errors would not go away
otherwise).

Now the downside of using Windows 7. Windows 7 sees the Extended Attributes
that W10/W11 (reparse points) have applied to the file system. Every time
Windows 7 CHKDSK sees one of those, it takes about a second to process
it and "mess with it". Obviously (up to a point), running an older
OS is not supposed to cause any long term issues for NTFS. The exception
to this, might be booting and sharing an NTFS partition between
WinXP and W10/W11. That might cause a more important issue, depending
on your W10/W11 setup.

But the Windows 7 CHKDSK will run, and as near as I can determine from
my limited usage of it, it's the software item for the job in this
case, even if it does not understand reparse points perfectly. It might
be a bit slow, because of the number of Extended Attributes. If this
happens, go make yourself a three course meal and "chill" :-) This is
how I got this fat, by "waiting for computers" :-)

Summary: I don't know if this is guaranteed to fix it, but the
         Windows 7 CHKDSK is the tool of last resort to try to get
         errors to go away. You can run the CHKDSK in Troubleshooting mode,
         from a Windows 7 Installer DVD. It is not necessary to have a
         hard drive with Windows 7 on it, to gain access to a Windows 7 CHKDSK.
         There is a potential limit, to how modern the hardware can be,
         and the Windows 7 DVD booting (I have not encountered a problem
         but my stuff is aging and my Zen3 isn't the latest any more).

   Paul

Back to alt.comp.os.windows-10 | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

chkdsk finds errors, fails to correct them Jason <jason_warren@INVALID.ieee.org> - 2025-07-31 13:04 -0400
  Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-07-31 10:42 -0700
    Re: chkdsk finds errors, fails to correct them jason_warren <jason_warren@ieee.org> - 2025-07-31 13:44 -0400
      Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-07-31 11:32 -0700
        Re: chkdsk finds errors, fails to correct them Philip Herlihy <nothing@invalid.com> - 2025-08-01 10:39 +0100
          Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-08-01 04:27 -0700
          Re: chkdsk finds errors, fails to correct them Paul <nospam@needed.invalid> - 2025-08-01 08:03 -0400
        Re: chkdsk finds errors, fails to correct them Jason <jason_warren@INVALID.ieee.org> - 2025-08-01 18:16 -0400
    Re: chkdsk finds errors, fails to correct them "J. P. Gilliver" <G6JPG@255soft.uk> - 2025-08-01 18:25 +0100
  Re: chkdsk finds errors, fails to correct them Paul <nospam@needed.invalid> - 2025-07-31 17:09 -0400
    Re: chkdsk finds errors, fails to correct them "Carlos E.R." <robin_listas@es.invalid> - 2025-08-01 02:41 +0200
      Re: chkdsk finds errors, fails to correct them Paul <nospam@needed.invalid> - 2025-07-31 23:22 -0400
        Re: chkdsk finds errors, fails to correct them "Carlos E.R." <robin_listas@es.invalid> - 2025-08-01 14:18 +0200
          Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-08-01 17:30 -0700
            Re: chkdsk finds errors, fails to correct them "Carlos E.R." <robin_listas@es.invalid> - 2025-08-02 14:20 +0200
              Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-08-03 23:15 -0700
                Re: chkdsk finds errors, fails to correct them Hank Rogers <invalid@nospam.com> - 2025-08-04 07:53 +0000
                Re: chkdsk finds errors, fails to correct them T <T@invalid.invalid> - 2025-08-04 01:21 -0700

csiph-web