$MFT fragments issue

[erketre]

I defragmented my NTFS drives with MyDefrag (for the first time) and I noticed \$MFT became fragmented. MyDefrag makes it so without any reason (no wrapping needed). What's strange it claims \$MFT is not fragmented, but I can see fragments in Defraggler (image) and DiskView! Why? Is there any reason for it? My specs: xp pro sp3, latest MyDefrag.

3 different drives (parts), view from Defraggler:
red - \$MFT
white - EMPTY

[poutnik]

The MFT could be temporary unmovable at time of Mydefrag running,
that may not be valid at time of postanalysis by another tool.

Edit - Jeroen is right - I forgot about unmovable MFT header.

[jeroen]

The Windows defragmentation API can move the entire MFT, except for the first 16 inodes (usually 4 clusters). Technically speaking Defraggler is correct, and MyDefrag has fragmented your MFT, but this is how Windows does it.

It's understandable that you are confused. Disks that were processed by MyDefrag look "strange" in other defraggers, not because MyDefrag has done something wrong but because MyDefrag has some sophisticated tricks up it's sleeve that other defraggers know nothing about.

[erketre]

I see. This should be pointed somewhere in manual.
Is there a way to integrate this file back?

[poutnik]

It is not necessery as this piece will be cached well.

It is even not recommended to join them back, if the MFT header is placed far from ideal MFT placement.
It would in fact decrease the MFT performance.

[jeroen]

Is there a way to integrate this file back?

No. As I said Defraggler is technically correct, but in my opinion they should not show the MFT as fragmented. It is confusing to you and other users, making you think that something is wrong and needs to be corrected.