Well, I was actually just coming by to drop off a hint on this very problem. In fact I was just deciding on where I should post my message, but I guess here is as good a place as any. :)
Anyways, I had the same problem and it turns out that in my case it's spinning someplace in /etc/rcS.d/S30checkfs.sh . In fact on my device I see it spawning two instances of 'sleep 1' at a time over and over again. Must be sourcing some other file, cause I don't see that behavior in the actual checkfs.sh script. I'll keep looking but in the meantime, in order to access your device you need to add yourself a backdoor of some type. I plugged one of the two drives from my MBWE II (blue lights) into my laptop, started up the RAID in degraded mode (mdadm —auto-detect), mounted the lenny part of the drive (mount /dev/md1 /mnt/md1), and created a new file:
- vi /mnt/md1/etc/rcS.d/S11hack
#---CUT HERE-——
#!/bin/sh
/sbin/ifconfig eth0 192.168.1.2
/usr/sbin/sshd &
exit 0
#-END HERE--
Made that executable (chmod +x /mnt/md1/etc/rcS.d/S11hack), plugged the drive back into the device, fired it up and I'm back in and diagnosing it at least. I'd be curious to know if someone found a real cure for this, rather than an ugly hacl. ;)
Also, thanks to everyone for all on the great work on this so far! Always nice to see Debian running on more devices. It's really starting to earn that "universal operating system" name. :D