Now that I've been pointed out the command "scu" to diagnose (and
apparently magically solve) my problem with the CD drive ... I've actuallt
encountered one more (or again the same) problem.
I post it to the list because I still suspect may have something to do
with Alpha hardware, although not all problems are Alpha-related.
1) I got from a colleague two data CDs which I had to copy and give back
to him (the CDs are scientific data and are produced by the European
Space Agency)
My colleague told he had problems with the filenames reading the CDs
on his Sun.
2) When I mount the ORIGINAL CDs on my Alpha (DU 3.2) with my usual
mount -r -t cdfs -o rrip,noversion /dev/rz4c /cdrom
it gives a warning : no RRIP on disk, defaulting to ISO-9660
Said that, I'm able to work perfectly (an ls sees long lowercase
names, while HTML pages on the CD, which contain URLs in UPPERCASE
pointing to files on the same CD, are also (surprisingly) resolved.
(This was the part not working for my colleague on Sun)
3) For this reason I wanted to make a physical image (with readcd) and
then burn a CD (with cdrecord) to preserve exactly the same type of
filesystem (whatever it is). This I have to do on a Sun (the burner
is there)
I could not do it however because readcd systematically fails on the
last 20-30 blocks of each CD (out of several ten thousands).
4) So I ended up mounting the CD's and making a Rock Ridge physical image
with mkisofs and burning with cdrecord (any other combination had
problem with filenames)
When I verified the CDs ON MY ALPHA it's when I had the (hardware ?)
problems I told to the list (and which this morning seemed to have
disappeared)
5) first (marginal) question. If I mount the (Rock Ridge) COPIES on
my Alpha an ls sees long lowercase names, but the UPPERCASE URLs
are NOT resolved any more. I tried mount without specifying -o rrip
and even without specifying -o version, but the result is even
worse (one either sees *short* lowercase ISO names or short UPPERCASE
ISO names.
There is no way to obtain the effect of "no RRIP on disk, defaulting to
ISO-9660" shown by the original CDs with the freshly burned copies.
I suppose this is a feature of the CDFS driver of my rather old DU 3.2
6) Then I went back to compare the original and the copies. Now the effect
is that, of 2 originals and 2 copies three are OK while
- ONE of the originals gives i/o error during mount on MY Alpha
(while other 3 disks are mounted ok)
- the same disk is mounted ok on ANOTHER Alpha
If I try a scu tur on my Alpha I get this
scu: 'test unit ready' failed, EIO (5) - I/O error
scu: Sense Key = 0x3 = MEDIUM ERROR - Nonrecoverable medium error,
Sense Code/Qualifier = (0x57, 0) = Unable to recover
table-of-contents
The only difference is that the other Alpha has the factory RRD45
drive, while mine has a RRD47 recently replaced by Compaq.
The scu show device is identical except for these lines
Product Identification: RRD47 (C) DEC
Firmware Revision Level: 1206
Product Identification: RRD45 (C) DEC
Firmware Revision Level: 1645
What is exactly the difference between an RRD47 and an RRD45 ?
Could it be that one is more sensible to poor recording (and is
poor recording possible for CDs as it was for magtapes ?) ?
----------------------------------------------------------------------------
Visitate il sito / visit the site
http://www.9-11peace.org/peace.php3
An eye for an eye leaves us all blind. Occhio per occhio ci fa tutti ciechi.
----------------------------------------------------------------------------
Lucio Chiappetti - IFCTR/CNR - via Bassini 15 - I-20133 Milano (Italy)
For more info :
http://www.ifctr.mi.cnr.it/~lucio/personal.html
----------------------------------------------------------------------------
Received on Tue Oct 09 2001 - 11:38:02 NZDT