Try running the following command from the rescue system: fw_setenv setenv usb_init ''by Jeff - Debian
The delay time is configurable. You can adjust it with the 'fw_setenv' command. The below example will make the kernel wait 20 seconds for the device to spin up. fw_setenv usb_rootdelay 20by Jeff - Rescue System
You can try booting from your old drive. All you have to do is plug it in and restart the Dockstar. You do NOT need to change any uBoot variables. If it doesn't work, or if you can't figure out how to repair it, you can just run the Debian installer from the Rescue Environment. -- Jeffby Jeff - Debian
It looks like your Pogoplug install is corrupt. We're going to manually reset your uBoot environment and then manually install the Rescue System from uBoot. Follow these steps very, very carefully. If something does not work or does not match, do not keep going: Download the following images to the root of your ext2 formatted thumb drive, and verify their m5dsum http://jeff.doozby Jeff - Debian
Assuming that TFTP is timing out because the TFTP host is not active, you can just add a 'ping check' before the TFTP commands. See the netconsole thread for an example of how to do this. -- Jeffby Jeff - uBoot
It sounds like your tftp server isn't working. New approach. From your laptop, format a flash drive as ext2, copy that uboot-original file to it, and the connect the drive to your Dockstar. Then run the following from uBoot: usb start ext2load usb 0:1 0x800000 /uboot-original-mtd0.kwb go 0x800200by Jeff - Debian
It sounds like you configured netconsole on uBoot, so once it sees that 192.168.2.4 is alive, it's sending all the console output to that device. Unplug the ethernet cable and reboot the dockstar and you be able to interrupt the boot process with the serial device. After that, connect the ethernet cable and then continue with the tftp commands I posted earlier.by Jeff - Debian
You've got serverip and ipaddr settings in your environment, so I'm assuming that you've got a TFTP server set up on 192.168.2.4. If not, you need to setup a TFTP server someplace on your network and the adjust those variables accordingly. Download uboot-original-mtd0.kwb and put it on your tftp server. From uBoot: tftp 0x800000 uboot-original-mtd0.kwb go 0x800200 -- Jby Jeff - Debian
The instructions you are trying to follow cause more problems than they solve. Flashing partitions from uBoot isn't a very good idea, it's possible that you could brick you device (it's not bricked yet, you've just misconfigured the uBoot environment.) Having said that, did you stop when you found out that 'resetenv' didn't work, or did you keep going? If sby Jeff - Debian
The error message tells you how to fix your problem. You can run the Debian installer from the rescue system.by Jeff - Rescue System
You're correct, thanks for reporting that. I've updated the script with the correct definition for FW_CONFIG_URL. -- Jeffby Jeff - Rescue System
Don't use sudo on your make commands, either.by Jeff - Debian
Try running the codesourcery script without using 'sudo' -- the script sets up some environment variables and you want those to be in your account, not the root account.by Jeff - Debian
FW_PRINTENV_URL is defined at the top of the script. If it's not, you should re-download the script.by Jeff - Rescue System
Don't worry about blparam, the script is expecting it to fail unless you're using the original Pogoplug environment. It looks like your error is related to unpacking the debootstrap package. Delete /tmp/debootstrap and try again. -- Jeffby Jeff - Rescue System
I've updated the installer to fix this. -- Jeffby Jeff - uBoot
CONFIG_BOOTCMD only sets the default bootcmd value. If 'bootcmd' is already defined in your environment, then uBoot will use that instead. -- Jeffby Jeff - uBoot
Good catch! I copied the wrong pkgdetails into the rescue image. I've fixed this and uploaded a new image. -- Jeffby Jeff - Rescue System
I've updated the Recovery System with a few good suggestions from Johannes plus better support for the Debian install scripts. The system now uses zcip to handle IP addresses. If will attempt DHCP, and if that fails it will use 169.254.9.17 and if that's already in use, it will get an availble IP address in the 169.254.0.0/255.255.0.0 subnet. The LED status lights have also changby Jeff - Rescue System
You probably have a bad mke2fs from the old script. If you delete it, the new script will find your good mke2fs and you'll be fine.by Jeff - Rescue System
It turns out there were still a few commands at the end of the script that were failing. The script has been fixed now. Just ignore the blparam errors. -- Jeffby Jeff - Rescue System
The Debian squeeze installer has been updated to work with the rescue system. If you tried to run the broken script, you will need to remove /usr/sbin/debootstrap and the directory /usr/share/debootstrap before running the new script. -- Jeffby Jeff - Rescue System
It's just when the ubifs system is rw. I mentioned it in the 'Debian on NAND' instructions, but hopefully this thread brings a little more visibility to the issue. This concern is also one of the motivating factors behind the new recovery system. If you have a rw ubifs on mtd3 and it fails, your system will fall over to the new recovery system on mtd2, which has support for clby Jeff - Debian
I think we should be using proper machine IDs rather than overloading sheevaplug. I plan to add support a configurable arcNumber variable in the uBoot environment so we can make the transition a little easier. -- Jeffby Jeff - Debian
Great suggestions, Johannes. There are probably many other parts of the system that need cleaned up, as the current version has very little polish. I pushed it out as soon as it was usable to let all of the smart people on the forum get a look at it and make their own contributions. I think a few days of everyone working on it is worth much more than another week of me working on it alone.by Jeff - Rescue System
The Debian install script has not yet been ported to the rescue system. In addition to the mke2fs error, I need to recompile pkgdetails for the rescue system. -- Jeffby Jeff - Rescue System
@thomas If the uBoot config didn't get set for some reason (odd, since that "# Rescue System installation has completed successfully." message is printed AFTER they run) you can configure them by hand: fw_setenv set_bootargs_rescue 'setenv bootargs console=$console ubi.mtd=2 root=ubi0:rootfs ro rootfstype=ubifs $mtdparts' fw_setenv bootcmd_rescue 'run set_boby Jeff - Rescue System
I've updated the images with a few changes: decompression support in tar, support for xfs and reiserfs, and zcip. zcip isn't configured, but if anyone wants to test it out, I'd be happy to add configs for it. My uBoot installer has been updated to run from this rescue system. The Debian installer still needs to be ported. The new list of binaries is: rescue:~# ls /bin /sby Jeff - Rescue System
It looks like those errors may just be bad blocks. Have you tried booting the rescue system? You mount the partition from Debian with the following commands just to see if it was created properly: ubiattach /dev/ubi_ctrl -m 2 mount -t ubifs ubi0:rootfs /mnt -- Jeffby Jeff - Rescue System
Everything looks fine, you're safe to reboot.by Jeff - Rescue System