[Discuss] Ways to Get New Laptops to Boot a LiveCD

Alan W. Irwin irwin at beluga.phys.uvic.ca
Thu Jan 14 16:42:32 PST 2010


On 2010-01-14 13:11-0800 Joan McIlmoyl Cleghorn wrote:

> Finally, the
> suggestion was made that I try putting it on a USB thumb drive. So I did,
> using unetbootin. Well, that wasn't successful either. I'm still unsure
> whether it was the fact I had other things on the drive or why exactly it
> didn't work. Finally, last night, I decided to download the iso again and do
> another burn. This time I used a CD instead of a DVD (a case of trying
> anything different). To my shock, it worked!!!

Hi Joan:

I am glad you finally solved your issue.

You have speculated there might be something Linux unfriendly about your DVD
driver, but another good possibility to consider is you had a bad DVD iso
download or a bad DVD burn.  I suggest it is always a good idea to check the
downloaded iso against the published md5sum (regardless of whether it is a
DVD or CD iso).  Do that like this:

md5sum downloaded.iso

which will give you a result you should be able to compare with an md5sum
value published on the original download page (usually via a file containing
md5sum values there).

You should also check the final burned result with md5sum as well. Note that
some drives will write garbage after an iso that is burned to them so they
never satisfy the automatic checksum tests in (say) k3b.  This happened to
me just the other day with k3b for a Debian distro CD.  The solution was to do
the following:

ls -l downloaded.iso

which gives the number of bytes in the iso you have downloaded (and which
you know is correct because of the md5sum check done above).  Then make
sure the burned disk created by k3b is in the drive (but no need to mount
it) and run

head --bytes=number_of_those_bytes /dev/scd0 |md5sum

to obtain the md5sum value right off the burned disk to compare with the one
for downloaded.iso. The initial "head ..." step makes sure there is no
interference from the trailing garbage.

(replace /dev/scd0 with wherever your DVD of CD drive resides in /dev.  In
my case, I obtained /dev/scd0 from my /etc/fstab file so that is probably
a good place to look initially for that information.)

I don't know whether the trailing garbage is produced on the drive by the
burn step or by some issue when reading a burned disk.  But as far as I can
tell it makes no practical difference at all (the result mounts without
issues, for example) except in interfering with the md5sum of the entire
disk at once unless you use head as above to get rid of the trailing
garbage.

So to summarize this

* Always check md5sum for downloaded iso to be sure it is correct.

* Always check md5sum of the burned disk, but if k3b gives you an error on
that, it may still be fine and you are just subject to the trailing garbage
issue. OTOH, if you still get an incorrect md5sum even with trailing garbage
trimmed, then it is time to transform that disk into a coaster and try
several more times with a variety of disk media. If all those trys fail, I
suppose you could be the victim of a kernel issue or a lemon drive.  In
either case, perhaps the simplest solution is to buy a new DVD/CD burning
drive making the deal beforehand that you get a free replacement if it
doesn't work on Linux.

* If md5sum tells you that you have a good burn (with or without trailing
garbage) then that disk should work unless the original from the download
site was screwed up by the distro.  But this is extremely rare, and bad
download or bad burn issues are encountered much more often.

Alan
__________________________
Alan W. Irwin

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________


More information about the Discuss mailing list