[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: 2° Weekly "wheezy" from jigdo file md5sum errors



On Wednesday 04 April 2012 16:29:20 Alex wrote:

Hi,

> This issue is being presented this week:
> 
> # md5sum -c MD5SUMS.large
> debian-testing-amd64-DVD-1.iso: OK
> debian-testing-amd64-DVD-2.iso: NON RIUSCITO
> debian-testing-amd64-DVD-3.iso: NON RIUSCITO
> debian-testing-amd64-DVD-4.iso: OK
> debian-testing-amd64-DVD-5.iso: OK
> debian-testing-amd64-DVD-6.iso: OK
> debian-testing-amd64-DVD-7.iso: OK
> debian-testing-amd64-DVD-8.iso: OK
> debian-testing-amd64-DVD-9.iso: OK
> debian-testing-amd64-DVD-10.iso: OK
> md5sum: ATTENZIONE: 2 di 10 codici di controllo calcolati NON corrispondono

Hm... this must be the first run, right? DVD-1 is fine, DVD-2 and DVD-3 failed.

> # md5sum -c MD5SUMS.large
> debian-testing-i386-DVD-1.iso: NON RIUSCITO
> debian-testing-i386-DVD-2.iso: OK
> debian-testing-i386-DVD-3.iso: OK
> debian-testing-i386-DVD-4.iso: OK
> debian-testing-i386-DVD-5.iso: OK
> debian-testing-i386-DVD-6.iso: OK
> debian-testing-i386-DVD-7.iso: OK
> debian-testing-i386-DVD-8.iso: OK
> debian-testing-i386-DVD-9.iso: OK
> debian-testing-i386-DVD-10.iso: OK
> md5sum: ATTENZIONE: 1 di 10 codici di controllo calcolati NON corrispondono

Now DVD-2 and DVD-3 ISO images are just fine... DVD-1 failed.
Do I understand it correctly that these are two subsequent runs of 'md5sum -c' 
which produces different results: the first run failed at the second and the 
third ISO, while the second run failed at the first ISO image. My best guess is 
that your md5sum is not working properly, due to hardware issues, namely RAM 
(memory) fault.

For your own test, you can also try running several subsequent times:

md5sum debian-testing-i386-DVD-1.iso

and you will see the different hash sum results.

> Last week and this week are different errors on DVD.
> I changed the mirrors from previous week for jido-lite in:
> 
> mi.mirror.garr.it/mirrors/debian/
> giano.com.dist.unige.it/debian-non-US/
> 
> I don't any "script" in jigdo-lite man page, sorry.

This is the script utility from the bsdutils package. You want to read 
script(1) man-page and use it like:

#script

now call the jigdo command;
when done with this session type 'exit' to quit the forked shell

the log file of this session is by default called 'typescript'.

> At the end of jigdo-lite command the message is for complete retrive all
> files, for example:
> 
> --2012-04-02 17:55:52--
> ftp://mi.mirror.garr.it/mirrors/debian/pool/main/g/gmtp/gmtp_1.3.1-2_amd64.
> deb =>
> "debian-testing-amd64-DVD-5.iso.tmpdir/mi.mirror.garr.it/mirrors/debian/poo
> l/main/g/gmtp/gmtp_1.3.1-2_amd64.deb" Connessione a
> mi.mirror.garr.it|193.206.139.34|:21... connesso.
> Accesso come utente anonymous ... Login eseguito!
> ==> SYST ... fatto.   ==> PWD ... fatto.
> ==> TYPE I ... fatto.  ==> CWD (1) /mirrors/debian/pool/main/g/gmtp ...
> fatto.
> ==> SIZE gmtp_1.3.1-2_amd64.deb ... 132598
> ==> PASV ... fatto.   ==> RETR gmtp_1.3.1-2_amd64.deb ... fatto.
> Lunghezza: 132598 (129K) (non autorevole)
> 
> 100%[======================================================================
> ===========================================================================
> ==========>] 132.598      391K/s   in 0,3s
> 
> 2012-04-02 17:55:53 (391 KB/s) -
> "debian-testing-amd64-DVD-5.iso.tmpdir/mi.mirror.garr.it/mirrors/debian/poo
> l/main/g/gmtp/gmtp_1.3.1-2_amd64.deb" salvato [132598]
> 
> TERMINATO --2012-04-02 17:55:53--
> Scaricati: 9 file, 7,5M in 16s (468 KB/s)
> Found 9 of the 9 files required by the template
> Successfully created `debian-testing-amd64-DVD-5.iso'

Using LANC=C would be more helpful, but translate.google.com helped a bit with 
the Italian words... All of them look just fine. A possible suspect was 
TERMINATO, but it should mean 'completed', rathen than 'terminated', possibly 
abruptly. So, all is fine with.

> -----------------------------------------------------------------
> Finished!
> 
> OK: Checksums match, image is good!
> #

Jigdo said it is fine.... and indeed I tried the second DVD via:
debian-testing-amd64-DVD-2.jigdo. The ISO image was constructed properly,
and the output of md5sum, sha1sum, sha256sum, sha512sum commands produced the 
same hash sums which are also found in the jigdo file itself, these are:

$ zgrep Image debian-testing-amd64-DVD-2.jigdo 
[Image]
# Image Hex MD5Sum d9b78e08b53e13a4b9b2a5b3ff38d260
# Image Hex SHA1Sum 8b80ea3630dd122a2ac360925faf9a98f1094013
# Image Hex SHA256Sum 
395cbcb42147ed7fee30a354b21547cfbf924a31729574d3ee11b83dab7c85c0
# Image Hex SHA512Sum 
ab590990521231aab9e358ec6e6b65faece54d15bc1a1e67ee87a81af514179c99d770de289609549874262dabb6dbfd50b476a13fee9399c5d14deb9303e2c9


And I can also see the same hash sums at the cdimage server too... the files:
 MD5SUMS.large
 SHA1SUMS.large
 SHA256SUMS.large   
 SHA512SUMS.large    

> Of course, for .iso DVD incorrect, change only last row in error and image
> is not good.

Sorry, I can't parse that... what has changed or rather what should be 
changed, and where is the error to be seen?

-- 
pub 4096R/0E4BD0AB <people.fccf.net/danchev/key pgp.mit.edu>


Reply to: