Hello Troy,
Very well seen: i am currently downloading ( only the
first ) RH9 CD from Redhat. I will report you the
result.
To answer to your ""What classifies your cd's
as'certified'? Which test did you run on them?"":
nothing except the manufacturer's, TDK, "certified
for" mention... and the fact that i am using this
brand, and others, for years without _any_ problem.
Just let me write how i burn my (precious) datas _and_
operating system CD ( i am not so carefull with
softwares ):
1/ shut down any overclocking, even for the graphic
card, and return to nominal speeds and voltages.
2/ shut down network interfaces, services, softwares.
3/ flag the iso as 'read only'
4/ md5sum check
5/ mount the iso and scan against virus
6/ unmount
6.a/ play guitar while the computer is burning CD
7/ check for the cd-burner software report
7.a) buffer overrun OK
7.b) burning OK
7.c) comparaison iso / burned ( nero & K3b ) OK
7/ eject the disc and re-insert it on the second CD
player ( in fact a DVD player )
8/ mount the iso image and check via windiff/diff that
the content is equal to the content of the CD disc...
9/ ...and check for no IO errors in the system log
10/ read randomly some files
I think this is rather 'paranoid', but i allow myself
to say that my CD are _perfectly_ burned.
So, that is why i am surprised with the failure of the
CD verification feature during the install.
An explanation may be related to my motherboard ( asus
P4P800 ) which already had a problem with Fedora Core
( 2 or 3, i don't remember ): the install procedure
destroyed the master boot record. Clearly this
motherboard may have some problem with a Linux release
not perfect regarding the IO bios section.
I think i also will download centos to verify this
point, and i will report you the results.
--- Troy Dawson <[log in to unmask]> wrote:
> hilare wrote:
> > Thank you, Connie and Troy, for answering so fast.
> I
> > am currently downloading the first iso image to
> solve
> > my md5sum unfortunate experience.
> >
> >
> >
> > Regarding the second point ( no problem with CDrom
> > capabalities up to 800Mo write/read, either with
> > W2K/Nero or various SL/Xcdroast/K3b. Currently i
> am
> > using 700Mo certified - TDK - CDroms ):
> >
>
> What classifies your cd's as 'certified'? Which
> test did you run on them?
>
> > 1/ i am aware that the autocheck disk is an option
> (
> > but a strong one: the only way to skip it is to
> press
> > a very well named button ""skip CD verification"",
> > which is more than 'incentive' regarding laws and
> > professional responsibility )
> >
> > 2/ i am sure this is a good feature ( i used it
> since
> > RH9 from ftp.redhat.com )
> >
> > 3/ but this never had reported OK with SL for me (
> RH9
> > always was OK, not on the same hardware than SL,
> but i
> > do not have this hardware anymore, nor i have the
> RH9
> > CD ). Do you mean that you, or anybody, had
> success
> > with it ?
> >
>
> Yes
> I knew I had already tested all of the 3.0.4 i386
> CD's, but I have
> reburned another batch, and the 4.0's. Every one of
> 3.0.5 and 4.0 i386
> CD's have passed that test.
>
> I don't know what else to say.
>
> > Anyway, i am ready to provide you any information
> if
> > you need them, in order to help you on this topic.
> >
> > In particular, you guess that i am ready to burn
> and
> > install SL305/i386 32bits (from W2K/Nero), and
> very
> > soon SL4.1 or 4.rolling (from SL305/K3b), so, if
> you
> > want to ask me anything for a testing purpose,
> feel
> > free to do so, i will spent the needed time on it.
> I
> > don't know if this is related, but i never had
> > obtained the original md5sum from an iso burned (
> and
> > previously checked md5sum ok ), and then
> re-extracted
> > as an image. Maybe is this an other way to say
> what
> > Troy stated as ""Not all CD media is created
> equal""
> > ?, but this is not explaining why RH9 CD
> verification
> > was ok and SL never (for me).
> >
>
> You could try burning the RH9 cd's, on the same
> media, in the same way,
> and see if they all pass the test. That's the only
> test I can think of.
>
> Troy
> --
> __________________________________________________
> Troy Dawson [log in to unmask] (630)840-6468
> Fermilab ComputingDivision/CSS CSI Group
> __________________________________________________
>
____________________________________________________
Start your day with Yahoo! - make it your home page
http://www.yahoo.com/r/hs
|