Hmmm. /var/db and /var/file-cache are picky file systems... they are actually mounted by a upstart script during the boot process. /proc is a special Linux filesystem which contains information on the running system (/proc/cpuinfo, for instance, gives you info on your processors).
When you get /var mounted, try running this command:
That should force the system into run level 3 (multi-user mode + networking). Alternatively, you can try run level 5.
If you have a Doctor image, you can probably configure MetaDoctor with info about it, and I think "make backup" will do a mem dump backup of the device, including tokens. Essentially a tar is created that lets you re-build the device later with MetaDoctor.
If the init commands don't work, I think you *might* be able to try this to mount /var/db and /var/file-cache:
Though... The best thing to do is to reboot the device, let it stall, get a console on it through novaterm, and then swipe off /var/log/messages and /var/log/dmesg. dmesg will give you a kernel level description of the system and probably note any low level failures, and messages will let you know which module is failing.
EDIT: You can post the messages and dmesg file here, or message them to me if you think they might contain passwords.