well...
by Iconoplast - 5/8/12 2:38 AM
In Reply to: The problem by Jimmy Greystone
I understand what you mean here, and I get what journaling does, but the issue you raise is a bit beside the point. Your correct in that my model can't be booted via USB, unless you want to mess with Open Firmware.
I may not have elucidated my point, so I'll explain what I've discovered I can do...
1) you need a .dmg, or an actual installed OS on another drive, connected via USB,
2) Start your computer (containing the drive you wish to restore to) in Open Firmware [command-option-o-f]
3) when you see the open firmware screen, find your USB port, using the device list command [dev ls]. In my case its:
/pci@f2000000/usb@1b
4) if the external drive you have connected to that USB port is drive that is formatted so that Open Firmware can read the file tree, you should see [Disk@1]. So I get, if Open Firmware sees the HD device:
/pci@f2000000/usb@1/Disk@1
From there, I just make an alias for the above address (other sources say to call it [ud]):
devalias ud /pci@f2000000/usb@1/disk@1
Then I write a boot command, with a number addressing the partition... in my case its 1s3, or the third partition file on disk 1 (the device):
boot ud:3,\System\Library\CoreServices\BootX
This is how you boot a PPC like mine off of USB. I've done this before, so I know it can be done. I question whether the Mac OSX Extended (journaled) format on the external drive I am booting from is the culprit that prevents me from performing the above operation, because of what I understand about Open Firmware.
My understanding is that Open Firmware is the most pared down shell you can reach on your Mac. Ostensibly, it's for the CPU to know where all it's computer components are. Unlike Single User Mode, you can't even get a command list by typing [help], or anything else for that matter. I read somewhere that it is or resembles the programming language Forth. I can't be sure, because these are things i read elsewhere on the web, so I am unwilling to commit to the truth of that.
Anyhow, my understanding is that journaling is just complicated enough of an extra bit of format code, that it can obscure the Open Firmware's attempts to "see" the device. I don't know if this is necessarily true, as you suggest it may not be, but I've heard it surmised otherwise.
I suppose I could test by formatting a usb Flash stick in MOSXE(j), and see if Open Firmware has trouble with that. I know it can read the flash stick fine in nonjournaled MOSXE. Unfortunately, backup to flash stick failed. I've looked around, and have read that it's because a flash stick is too "slow" a form of memory storage for a mac attempting the boot process.
I'll attempt that soon, and get back to you all...
what R. Proffitt says below is intriguing to me, because I am using a 1TB drive which Open Firmware is failing to detect, and when I succeeded at the USB external boot, it was using an actual 160GB mac formatted desktop drive in an swappable drive chassis. He suggested it was a formatting problem, but perhaps it's the drive size, like some information technology equivalent of mismatched gear ratios?
I'm using an older drive with an earlier OSX for now, but I'd like to fix this issue at some point, so I can put the newer, larger drive back into my laptop. I'd also like to see if I can solve this problem with what I have, rather than use the Open Wallet command.
Was this reply helpful? (0) (0)
Staff pick