Recovery Force® Data Recovery Forum

This is where Apple takes two devices, usually an SSD and a HDD and merges them into a single drive, creating their own flavour of SSHD. I found a good article on how to manually create and delete a Fusion drive at the following site - http://www.techrepublic.com/article/pro ... ion-drive/

Here is a list of terminal commands used (you can go to the article to get details about each command:

>diskutil list
>diskutil coreStorage create LOGICAL_VOL_GROUP_NAME DRIVE_1 DRIVE_2
>diskutil coreStorage createVolume lvgUUID type name size
>diskutil coreStorage delete lvgUUID

Basically, a Fusion Drive is a spanned volume where the SSD drive is the first portion of the volume.

I'm not sure I 100% agree with that. There seems to be more to the system because there's no way to simply build it into a span when I've tried playing with it in R-Studio. While assembling that way may act pretty similar to fusion I think the OS plays some other games with the filesystem.

Correct me if I'm wrong, but it doesn't seem possible to re-assemble a fusion drive outside of Mac OS.

It has been a few years since Fusion Drives hit the market. Every one we received for recovery, thus far, have only come with the Hard Drive portion, without the very critical and necessary SSD. In all cases, but one, the Apple technician (not just Apple Stores) did not remove the SSD with the hard drive and just replaced the hard drive in the system.

In order for anyone to fully recover data from a fusion drive, it is essential that both the SSD and HDD drives are removed from the system and provided to the data recovery technician for recovery.

To follow up on this thread, it should be just a matter of determining the offset of the LVM containers on each drive and then joining them in a JBOD. But, I've never had an opportunity to test this theory, yet.

I have tested it, and things just don't line up and make any sense. I've even taken and written a huge file consisting of a pattern of counting LBA values to the volume, and the values after I analyze in HEX are all over the place. Data that should be at the end is near the beginning (probably the write cache part). Data that should be a consistent counting pattern will break, then resume from a different part of the file, etc. It's a fragmented, out of order, mess that I think only MacOS knows how to really work with and put it back together right now.

Interesting. Do you have any more info on corestorage anywhere on the forum?
A search for corestorage only brought up this article.

Yeah, nothing posted about Core Storage here, yet. What are you trying to find out? Both R-Studio and UFS Explorer support Apple Core Storage now, if that helps.

After several updates in UFS Explorer from January to March 2019 we may be sure that software supports it very well.