Depending on your process and your system, there are a couple things to check:
1) Partition table type.
Depending on your system, it may need to use an Apple Partition Table rather than a DOS-compatible one. From your livecd, check the output of fdisk -l in a terminal/console. If it shows partitions, then you're using an MSDOS pattition map. Another partiion tool named apple-fdisk or similar is used to create Apple/ppc partition types and one of the "partitions" would show as an Apple Bootstrap partition. If you used the wrong type during install, your system may not be able to boot from it even if it can use that partition scheme for other, non-root drives. The wrong partitioning type would mean you wouldn't see any bootloader or even start to the boot process beyond.
The last I used Linux on a ppc the bootloader named yaboot was used and it may still be the current bootloader of choice for ppc based distros. On some hardware it had difficulty or "quirks" during bootloader installation. You may have to look for a work around if you're not able to even get a bootloader screen. I had systems were bootloader vs kernel "named" the device differently so I had to edit configs accordingly to make it boot (old Apple X Serve).
Assuming you get past the bootloader and your kernel is loaded (scrolling messages and/or logo during boot) but then it fails, it could be a driver problem. If you were seeing scrolling messages, you'll usually see output of disk partitions it can "see" or use if it fails to mount your / (root) partition during boot. I'm assuming you didn't configure your kernel manually, but if you were seeing scrolling kernel mesages and it shows no partitions in the output, it means it does not have a driver that works for your IDE/SATA drive controller.
If arch (your distro) is using an older IDE-based driver for your hardware, it may be listed as hdX rather than sdX in /dev. I wouldn't expect this if the version is new, but it's possible. You should be able to use dmesg | less in your rescue shell (or dmesg |grep sd then dmesg |grep hd ) to search for how your drive is named.
Simply not seeing a /dev/disk or /dev/ doesn't mean much during boot. In your rescue shell udev wouldn't be running yet likely and you would only have static device nodes included in the initial ramdisk layout. You could try mount proc and look for partitions there: mount -n -t proc none / where your mount point could be any directory, even /tmp if you find yourself in read-only mode. The -n will prevent it from trying to write to /etc/mtab, just in case you can't write to the filesytem in the rescue shell. After proc is mounted, look cat /proc/partitions or /proc/disks or /proc/diskstats for entries about your drive.
4) Provide more information
If none of the above were helpful, provide more information including version of your linux distribution.