Pro6 hangs at 'Booting; New disk added' without disk added

Recently I bought a (used) Pro6 and 4 x 6 TB WD RED (new) to reorganize my storage.

I have an Ultra6 running 4.2.26 with 18 TB space (3 TB empty) in Raid5 (X-raid) and an Ultra6 running 4.2.26 with 16 TB space (2 TB empty), also in Raid 5 / x-raid.

The Pro6 is runnung 4.2.27 / Raid5 / x-raid (about 18 TB available).

The plan:

- copy data from first Ultra6 to Pro6 using rsync over ssh (succeeded)

- setup first Ultra6 with 6.3.35RC1 and dual dedundancy (Raid 6), switch back to x-raid (succeeded)

- copy data back from Pro6 to Ultra6 using rsync over ssh (partially succeeded)

- same procedure for second Ultra6

- use the 4 x 6TB from the Pro6 to expand the first Ultra6

- use 2 HDD's from first Ultra6 to expand the second Ultra6

- Store the (empty) Pro6 unit for future use

The backup from the first Ultra6 to the Pro6 went well; Ultra6 was mounted to the Pro6, rsync was executed on the Pro6. Throughput was about 600 Mbit, using Jumbo frames and a direct cable on eth1.

The 'reverse backup' (mounting the Pro6 on the Ultra6, running rsync on the Ultra6) went considerably slower (about 400 Mbit).

Assuming this was related more powerful hardware on the Pro6, I thought it would be best to run rsync on the Pro6.

At this point I must have had a blackout, because instead of mounting the volumes the other way around, I decided to swap the HDD's.

Upfront I have put both the kernel and the initrd.gz (slightly renamed) of both OS4 and OS6 on the bootflash and adapted the SYSLINUX.CFG accordingly with a new bootlabel, so I can boot into both OS'es on the same machine using the boot menu. I verified this before start.

What I did not think about is the potential difference between 4.2.26 and 4.2.27.

So I decided to swap the HDD's. First I numbered the disks in both units, using labels (stickers). Then I took out the disks on the Pro6, put in the disks of the Ultra6, and booted the Pro6 into OS6; all went well.

Then I put the disks originating from the Pro6 (running 4.2.27) into the Ultra6 and booted into OS4 (4.2.26). After a while, the display showed "Booting..." on the first line and "New disk added" on the second line. And it kept showing that... :-(

I am a 100% sure the disks were on their original position, and no disk was added.

After 5 minutes, I switched off the unit and put the disks back in the units they came from. The Ultra6 booted fine, the Pro6 (now equipped with it's original 4 x 6TB disks at their original position) stuck at "Booting.... New disk added". It does so for about 12 hours now.

I have no clue as to what's going on inside. I can't telnet or ssh into the OS. Using a serial connection, I see garbage when the unit starts up (using 9600 8n1). Raidar shows the unit 2 times for each eth port (so 4 times in total), saying "Management service is offline" on 2 instances (showing 4 disks), and disks not present on the other 2 instances. It doesn't recognise the OS.

Raidar shows the ip address that's provided by DHCP on eth0, but not the static ip address that is set for eth1 (the cable is connected at both ends, the other end does show its own static address). So it seems to me the unit has partially booted up.

Does anyone have an idea on what's going on inside?

I hope it's secretly rebuilding a raid array, so I probably can restore the 7 TB data that still was not copied back (though it is not essential data). If that's the case, I's be happy to just wait until the unit comes back alive.

However, if it just hangs in an eternal bootloop, I'd rather do a factory default...

I hope someone can shed a light on this, either on what's going on inside or on how I can access the unit over serial. Unfortunately I don't have a VGA breakout cable available at the moment.

Thanks in advance!