Re: growisofs, LG GSA-4082B, DVD+RW : SK=4h/ASC=A0h/ACQ=80h
> > What does it mean when i get this message from a single evil DVD+RW media :
> >
> > :-[ WRITE@LBA=26770h failed with SK=4h/ASC=A0h/ACQ=80h]: Input/output error
>
> ASC=A0h is from reserved range, meaning that vendors are free to assign
> own meanings to these error codes, consequently meaning that only LG
> will be able to say what does this particular code mean exactly.
Oh. This explains why it is not on your list.
> > #2 : fails after 300 MB
> > 322306048/3946780672 ( 8.2%) @2.2x, remaining 17:14
> > 322666496/3946780672 ( 8.2%) @0.1x, remaining 17:47
> > :-[ WRITE@LBA=26770h failed with SK=4h/ASC=A0h/ACQ=80h]: Input/output error
> > builtin_dd: 157552*2KB out @ average 2.4x1385KBps
> > :-( write failed: Input/output error
> > /dev/sr0: flushing cache
> > :-[ FLUSH CACHE failed with SK=4h/ASC=09h/ACQ=01h]: Input/output error
> >
> > Now the whole system hangs.
> > First i believed that DMA and/or growisofs version had an influence but
> > now i think that the particular media just gets more and more evil.
>
> It apparently does. It does sound like a media defect of some sort. I
> bet both unit and kernel could have handled it in more graceful way,
> better than system lock-up. But they didn't and that's life...
I am spoiled by the overall stability of my Linux systems, i fear.
> sense code with RICOHJPN/W01 mediaid. After my reply to "if problem
> persists over several brands, then unit is defective and you should talk
> to your vendor," the user never heard back. So I suppose same is
> applicable in your case.
At least you hear from me. :))
Current state : 19 of 20 Fujis work well. The media that only made 2x
is now at 3.9x like all the others. I do not know what i watched there but
now it is not reproducible any more. About 45 write jobs were successful
up to now.
The single media which caused the crashes is now labeled with a skull
and stored safely so it cannot get into the DVD writer. It still
worked ill as reported when i tested it a few days ago.
The slowdowns are reduced in number and slowliness. If they occur now,
they can be explained by heavy disk activity due to other processes.
(That was not the case two weeks ago ! I am sure.)
Is it (under lax superstition filters) possible that the DVD writer
needed a few tries to get in working shape ? All problems were with
media that were used at a very early moment in its life.
I fiddled with the system meanwhile but mainly at the graphics card
and not IDE devices. A connection to the DVD writer is not obvious.
Nevertheless, i will pre-test all newly bought DVD media at times
when a system hangup is only annoying and not a big damage.
> > Both media which do not fail show intermediate states of slowliness
> > while getting written (regardless of DMA) :
> > 1573978112/4700372992 (33.5%) @3.2x, remaining 11:59
> > 1574961152/4700372992 (33.5%) @0.2x, remaining 12:04
> > 1574993920/4700372992 (33.5%) @0.0x, remaining 12:10
> > 1581416448/4700372992 (33.6%) @1.4x, remaining 12:13
> > 1595932672/4700372992 (34.0%) @3.1x, remaining 12:09
> > I found this : http://lists.debian.org/cdwrite/2004/03/msg00126.html
>
> But what you should try to figure out is why 3x? I suppose expected
> speed was 4x. As mentioned earlier 3x sound very much like lack of DMA...
Usually it is 3.9 :
/dev/sr0: "Current Write Speed" is 4.1x1385KBps.
25231360/4700372992 ( 0.5%) @5.3x, remaining 15:26
43712512/4700372992 ( 0.9%) @3.9x, remaining 14:12
...
4688183296/4700372992 (99.7%) @3.9x, remaining 0:02
builtin_dd: 2293840*2KB out @ average 3.9x1385KBps
hdparm -d1 and hdparm -d0 did not make a reproducible difference.
> > This is from a gracefully ended try with growisofs-5.12 :
>
> This is odd. Kernel complained as hell and growisofs didn't say a thing
> and completed gracefully? Or do you mean "gracefully" as "it complained,
> but sysytem didn't lock-up"?
A non-lock-up unsuccessful end of growisofs. With ":-[" and all.
In contrast to a sudden freeze without an end.
The media is reliably ill with unrealiable number of writeable bytes.
Mostly about 300 MB. (Most tests made with a 4480 MB random pattern)
> Well, one way or another, just discard that
> particular media and keep trying. If problem persists over several
> medias and brands, and if you're positive it's not some interference
> with another program such as automounting factility, then talk to your
> vendor. A.
I will at least try to get a translation of the error codes from LG.
Next i will buy some DVD-RW and look how good they work.
Some day i will even learn about writing DVD-RAM and try one. Some day ...
Thanks for your attention.
Have a nice day :)
Thomas
Reply to: