Sysop: | Amessyroom |
---|---|
Location: | Fayetteville, NC |
Users: | 25 |
Nodes: | 6 (0 / 6) |
Uptime: | 01:39:00 |
Calls: | 494 |
Calls today: | 3 |
Files: | 1,078 |
Messages: | 68,397 |
"Carlos E.R." <robin_listas@es.invalid> wrote:
There are two identical MFTs: main one is on the outer tracks of an HDD, >>> and second is halfway in the platter. The outer would have faster
access. The middle one is for redundancy, and for corruption recovery.
Are both written at the same time?
I would assume so for both to be exactly the same. The duplicate is in
case one gets corrupted, like the sectors on the platter went bad. As
to how the OS does it, I don't know if the file I/O is duplicated and paralleled, or there is ensuing mirroring.
Filename for the MFT is $Mft, and for its mirror is $MftMirr.
The locations for the MFTs are recorded in the boot sector which is also duplicated at the middle of the drive.
https://www.file-recovery.com/recovery-NTFS-master-file-table-MFT.htm
However, the actual operation is never fully explained. Quite often I
see mention that $MftMirr mirrors, at least, the first 4 records of
$Mft, like "MFTMirr: A backup copy of the first 4 records of the MFT."
I'm not sure how just 4 file records are doable for file system
recovery. Perhaps $MftMirr is just for recovery of $Mft, like $Mft got deleted, so $MftMirr could be used to relocate its sectors, not for
recovery of files.
https://flatcap.github.io/linux-ntfs/ntfs/files/mftmirr.html https://bromiley.medium.com/ntfs-part-7-an-ntfs-story-caf42565855b
This is for NTFS. I doubt FAT (and bitwidth) has any file table
recovery scheme. While interesting, I've not much delved into NTFS
beyond what I had to know, or got curious about at the time.
So, my statement there are 2 identical MFTs appears incorrect. There is--
$Mft for the file system table, and there's $MftMirr pointing to where
$Mft is stored.