On Saturday 27 December 2008 22:14:39 Adam wrote: > Chris Knadle wrote: > >> Attempt #1 to install Ubuntu 8.04, using a burned copy of the ISO I > >> downloaded, got partway (27%?) through "copying files," then stopped > >> with an error message along the lines of "Your CD is defective or your > >> CD drive is defective or your HD is defective." > > > > Another friend of mine ran into this problem, and he found the > > same issue with the CD passing the validity check too. I think > > the reason the validity check passes but the installation fails > > is that it's far easier for a CD drive to make a linear pass of > > all the data rather than having to random seek for files. So for > > all we know this has to do with the Linux kernel on the newer > > Ubuntu Live CD timing out on file seeks by default earlier than > > previous versions. > > > > If I remember correctly, the trick that worked for him was to burn > > the Ubuntu CD again but at a slower burn rate. The only other > > suggestion I have is to try replacing the CD drive in the target > > machine, or try cleaning it with a CD cleaner. The "CD cleaning > > fluid" is just isopropyl alcohol. > > Thanks, everybody, for your suggestions. Chris nailed it, although I > didn't see his message until afterward.
Maybe, but to be honest I thought Mike Muller had the best suggestion. I'm just glad to be part of the team effort on this stuff. These computers are complicated, I tell ya. > As part of my strategy of "try everything," I booted up Windows on > my parents' computer, and used Nero to make a copy of the pressed > Ubuntu 8.04.1 CD at the default speed, which was either 40x or 48x. > Then I rebooted from the new copy, and Ubuntu installed with no > problems at all, and runs. (Although I didn't get to check out > whether the installation can access the optical drive.) > > So Ubuntu's installed, but I don't understand /why/ it worked. How > would the copy I burned be any different from the one that Ubuntu > pressed? By definition they should be the same, and in fact if you run a test I bet you will find that they are. Yes, I know that one works and the other doesn't, but that doesn't mean that what's written on them WHEN THERE ARE NO READ ERRORS is different. I think all that's going on is that one CD is causing the CD drive to have read errors, where the other one isn't, and only on random accesses, but I don't know why. What you can do, and should do to prove this for yourself, is to run 'md5sum' and/or 'sha1sum' on the downloaded ISO file, and then on the CD/DVDs you burnt directly. i.e. first 'md5sum unbutu-8.10-desktop-i386.iso', and then a checkum on the CD device with the CD in the drive, such as 'md5sum /dev/hda'. The checksums should come out the same, and most likely will, since the Ubuntu media check passed. You'll see that in terms of the data written, Nero did the same thing that the program you used in Ubuntu did. Nero, K3b, and all other programs will burn the ISO image the same as far as I know, with the exception of the meta information that's not in the data area, i.e. "Burnt by Nero version X" kind of thing. They MUST burn the data the same, because an ISO image is a filesystem within a file, which is why you can checkum the ISO image file and then the ISO image that's burnt to CD and have the checksums come out the same. It's just a direct burn onto the media. There's no clear way to take a single instance of a bad burn and come to a conclusion of why burning the ISO image using Ubuntu didn't work where burning from Nero from Windows did. It could just be a bad or mismanufactured CD, maybe it was an old or dusty one, maybe there was a tiny brief data underrun in one instance rather than the other... There are lots of reasons why a single CD burn can come out bad, especially when you're trying to burn fast like at 40x or 48x. Maybe you'll eventually figure it out, maybe it won't happen again, maybe it'll be inconsistently inconsistent. Who knows. All I know is that experimentally I found burning slower yields more reliable results than trying to burn at super-fast rates. And since I'm usually not in a hurry for the burn, I typically burn CDs at 8x and DVDs at 4x. > And if there is a difference, why didn't the 8.04 that I > burned on my own machine work there? BTW most of my parents' > machine is from about 2004 if that makes any difference. I make consistently good burns from a box from 2001. Shouldn't matter. -- Chris -- Chris Knadle [email protected] _______________________________________________ Mid-Hudson Valley Linux Users Group http://mhvlug.org http://mhvlug.org/cgi-bin/mailman/listinfo/mhvlug Upcoming Meetings (6pm - 8pm) MHVLS Auditorium Dec 3 - Lightning Talks & Swap Session
