Unsupported File System(s)

[Froggie]

According to previous posts, unsupported FileSystems, when being imaged, can only do sector-by-sector imaging since HBS knows nothing of "used sectors"/etc. when confronted with unsupported FileSystems.

So, after imaging (FULL at 7.6gB) only my Ubuntu partition, I logged into my dual-BOOT Ubuntu/Windows System and under Ubuntu I first TRIMmed the ROOT partition then ran an update of my Linux System.  Upon completion of that update, I re-logged into my Windows System and, using Hasleo, ran an INCREMENTAL image of that FULL done at the beginning.  I expected an image size equal to or larger (due to updates) than the original FULL... mainly 'cause HBS don't know ext<n> FileSystems.  Instead I wound up with a 4gB INCREMENTAL image... much smaller than the original FULL.  I then ran an additional FULL (following my updates) that resulted in a 9.3gB image.

Looking at my previous paragraph... how is that even possible with no knowledge of "used sectors" in an ext<n> FileSystem?

Inquiring minds really wanna know about this one... 

[n8chavez]

I am unsure of this issue. However, your post looks like a chaotic ransom note that uses cut out letters from magazines.

[Froggie]

I am unsure of this issue.  However, your post looks like a chaotic ransom note that uses cut out letters from magazines.

Maybe I should have taken that approach... 

[CDC9762]

My thought.
Why bring the filesystem into the picture? The program knows it can't use it and does a sector-by-sector backup. When you do the incremental, it does a sector-by-sector comparison, only copying the changed sectors.

[Froggie]

That would normally make sense, BUT... there is little to no activity on the disk holding the images.  If your assumption is correct, HBS would have to read the entire previous image and compare it to the current SOURCE disk... that is not happening.  All the disk activity during the incremental is occurring only on the SOURCE disk, no other place.  That means the decision on the changes is coming from the SOURCE disk... and that can only happen if HBS knows how to interrogate the FileSystem of the EXT4 source partition.

Another INCREMENTAL done against the recent 9+gB inactive ext4 partition shows only changes in the 270+kb range and no activity anywhere except on the SOURCE disk (reading).

I'd love to hear from the Devs...

[admin]

CDC9762's assumption is correct: even when performing incremental backups on a partition sector by sector, HBS will compare blocks and only back up changed blocks.

[Froggie]

The process makes sense, the System's action does not. A sector-by-sector comparison requires the entire reading of both the SOURCE disk and the image itself... sorry, I don't see that at all (see Post #5 above).

Pls tell me what I'm missing... 

[admin]

Each data block has a checksum, so the program only needs to check the checksum to quickly determine whether the data in a block has changed. It seems that I shouldn't discuss the technical details too much here.