HR10-250: Problemu upgrading from 6.3a to 6.3e

[dtle]

I've followed instructions like here (http://www.dealdatabase.com/forum/showthread.php?t=51121) to upgrade my HR10-250 from 6.3a to 6.3e.

The 6.3a is running on partition 4. My problem is that even though I did the dd to copy the partition 3 kernel to partition 6 kernel after the upgrade:

dd if=/dev/hda3 of=/dev/hda6 bs=1024k

It doesn't seem like the dd took affect. After rebooting, the tivo is stuck on the "few minute" screen.

I took the drive out and looked at the files. All of the modifed files are deleted, including tivoapp, that's why it was stuck at the screen.

I also tried to re-killhdinitrd the kernel, but that didn't work either.

I'm still using the 3.1.5 kernel. Do I need to buy the new upgraded kernel?

[wundernaut]

I am seeing the same exact problem here when trying to upgrade from 6.3a to 6.3e using the same 6.3a upgrade post as my template. The dd from the old kernel partition to the new kernel partition didn't seem to install the proper kernel so the chainload stuff never happens (I see initrd being loaded into the ramdisk). Yet when I set my bootpage back to my old root partition and boot, it chainloads correctly and boots up as before indicating the kernel on my old kernel partition is still the one I want. I even grabbed my old PTVupgrade disk with the 3.1.5 kernel and dd it to my new kernel partition with the same result. It seems the kernel from my new kernel partition is not being loaded, but I have to wonder where it's coming from then...

BTW, dtle, if you want to try reinstalling the 6.3e slices again there's a variable in the installSw.itcl script (EMERGENCY_REINSTALL, I think, but check the script for it). If you "export EMERGENCY_REINSTALL=1" you can then run the installSw.tcl script again if you want.

Blast CBS! :-(

[wundernaut]

Well, my problem was that I thought I would "improve" test.conf a bit after copying it over. Instead it failed to execute, with the result that the chainloading stuff never happened. The tipoff was that chainload/check.sh puts a message of one kind or another out onto the console no matter what it decides to do, and I was never seeing it.

Now I'm seeing the monte'd kernel panic, but no doubt that's another case of operator error I need to sort out. Let's just say I'm a lot more confident that the 6.3a to 6.3e upgrade is possible when things are done right... ;-)