migrated raid disks to a new platform. trying to rebuild...

[theraginggeek]

Hello forum, I have a challenging situation and all my google-fu has lead me not very far. let me set the stage.
I have been running Gentoo happily on on old Phenom II Black platform for the better part of the decade with little issues, here and there a drive failure etc. most recently my SSD Root drive died which left me with the need to throw another drive into the machine, except this time I wanted to do something a bit more bold. I figured I had recently done a number of upgrades on my newer main PC, and had enough parts laying around to build a new server using a Ryzen 7 2700X and 32gigs of ram as the core instead of the Phenom II with 12Gigs of Ram. quite a performance upgrade.

So I undertook the effort of migrating all of the hardware last week, prior to shutting down the old machine for good I did take a snapshot of the /etc directory into my cold-storage drive, which is a USB HP Storage Drive about 6TB.

Last night I installed Gentoo onto the new 500Gb SSD, got it to a place where it could boot on it's own, and left it. today I updated world, installed some of my favorite tools as well as mdadm, lvm2 etc in order to support my restore of the raid. From what I recall of the raid this is the configuration as I had it:

4 1.8 TB WD Red Drives, partitioned with 1 partition per drive.
I believe I had them in pairs of raid 1, then raided together as a raid 0 to allow redundancy and striping(RAID10).
on top of this I had LVM running and split the space across 2 usable disk areas mounted to /home and a /shares mountpoint for use as a file server space.

on the new machine when I run lsblk I see:

[NeddySeagoon]

theraginggeek,

Check your kernel for CONFIG_EFI_PARTITION.

Without that, the partition tools will work but the kernel will not be able to read the GPT to find your raid partitions.

[theraginggeek]

checked my kernel config and I see:

odin /usr/src/linux # cat .config | grep EFI | grep PARTITION
CONFIG_EFI_PARTITION=y

so it looks like it should support it. the partitions were made probably over a year ago on the older machine so they should've been saved somewhere

if anyone has any additional ideas I'd appreciate it, else I might have to start building a new raid cluster tonight and see how stale my previous backups were.

[NeddySeagoon]

theraginggeek,

A partition table is a collection of pointers to the parttions.
Wiping the partition table looks scary but your data is unharmed.
Writing a GPT in place og an MSDOS table, destroys the data at the end of the volume, as the backup GPT will be written there.

Use testdisk to scan for partitions. Do not let is write any new partition table yet. Make a note of what it finds as you want to exclude false positives.

It would be best to rebuild the partition table by hand, using the testdisk output for hints.

I missed that sda was a part of your old raid set.
Are you sure you used GPT orignally?

1.8TB fits inside the MSDOS hard 2.0TiB limit.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

[theraginggeek]

I think I made some headway with testdisk, handy tool that. scanning the drives gives me this:

[NeddySeagoon]

theraginggeek,

The next step is to make images of those drives, so that you have an undo function. Unless you have validated backups already
See my signature :)

It looks like