Blade100, Sol9, A07 Pro-DVD
From: anon (anon_at_nowhere.net)
Date: 05/21/04
- Next message: milagro_at_beanfield.com: "Trying to Compile Image Magick on Solaris 8"
- Previous message: joe_at_invalid.address: "Re: Shared libraries used by an Unix process"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Thu, 20 May 2004 18:25:05 -0400
I have done limited testing of Pro-DVD 2.01a27 on the following
configuration:
Sun Blade 100
Solaris 9 8/03 w/ rec. patch cluster dated around 9/03
Pioneer DVR-A07U (1.09 and 1.13 fw)
cd-r, cd-rw, and dvd-rw (1x) media
This system is one that has IDE support built in on the motherboard. I
tested configurations where the data was located on a (ide) hard drive that
was on the same cable as the dvd writer as well as that where the data was
located on a drive that was on the other cable.
When they are on different cables the software seems to work great.
When they are on the same cable the s/w seems to work well but the system
may hang for a while and a message to /var/adm/messages saying "dada:
WARNING: /pci@1F,0/ide@d/dad@0,0 (dada0): ATA transport failed: reason
'timeout: retrying command" always occurs during finilizaton. Occasionally
this will be followed by a message saying "dada: WARNING:
/pci@1F,0/ide@d/dad@0,0 (dada0): ATA transport failed: reason 'timeout:
giving up". The latter is a bit disturbing, but I will have to leave it to
others to say if it really represents a problem or not.
JS seems to tend toward the position (If I am understanding him correctly -
something that may not be the case) that this hardware configuration (ide hd
and dvd writer on the same cable) is not "correct" (I agree that it is not
optimal, but am puzzled by anything beyond this), and that the problem is a
result of Solaris not liking anything tying up the ide interface for more
than 20 seconds. Seems to me that any application that tied up the bus for
that long could equally be considered as misbehaving... but JS is much more
of an expert in the area then I am so you probably should listen to him on
this.
JS says that -immed will probably cause the transport failed messages to go
away, but I have not tested this on a blade 100. I am testing it on a
blade150, but that will be covered in a different post. You probably want
to read the man page regarding the -immed switch... it is experimental.
Note that in the above test setup none of JS's other software (mkisofs, scg
driver, etc) are loaded on this system, nor are they used in this test.
Also, I think the test configuration is using sun's USCSI driver, which,
according to sun, is "unsupported" and should not be used except for
development. I have not discussed this with JS, so I am not going to take a
position on this.
It is of interest is that the cdrw(1) contained in a recently released patch
for Solaris 9 (adds dvd support to Solaris 9) has problems finalizing the
dvd in either of the two hardware configurations, so it may be having
trouble with the same issue.
Your results may vary.
a johnston
- Next message: milagro_at_beanfield.com: "Trying to Compile Image Magick on Solaris 8"
- Previous message: joe_at_invalid.address: "Re: Shared libraries used by an Unix process"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
