Linux-Hardware Digest #919, Volume #10 Tue, 3 Aug 99 15:13:29 EDT
Contents:
Re: SCSI vs. IDE (Mark Hahn)
Re: SCSI vs. IDE (Mark Hahn)
Re: SCSI vs. IDE (Mark Hahn)
Re: Soundblaster Live - SuSE 6.0 (Michael Perry)
US Robotics Sportster 33.6 ISA modem supported? ("A. Giachino")
Re: bttv and wintv problems (Paul 'Tok' Kiela)
Problem with DE-528CT and RH6.0 ("Terrance")
ATI PC2TV NTSC output ("t.j")
SONY DVD-ROM DDU220E PROBLEM !! ("Tsui Kwok-wai")
Re: help need a 3dfx vododo (Frank v Waveren)
Re: 3 button mouse (Craig Ruff)
Need kernel compile help ([EMAIL PROTECTED])
Re: Configuring network speed for Netgear FA310-TX NIC ("Andrew J. Norman")
UPS for multiple linux machines? (crux)
Re: more applications for Linux ("jacobite")
Re: SCSI emulation layer & CDRW (Thomas Zajic)
Re: Diamond Viper 770 at 24-bit??? (Michael Perry)
Re: Apache and ASP (Student)
----------------------------------------------------------------------------
From: Mark Hahn <[EMAIL PROTECTED]>
Subject: Re: SCSI vs. IDE
Date: 3 Aug 1999 15:59:37 GMT
> : this is mostly an old wive's tale. ide and scsi are both equally well
> : suited to "workstation" vs "server" workloads.
> no, its NOT a wives tale. scsi still multitasks better since it
> optimized i/o as requested by multiple quasi-simultaneous users.
no, you are wrong. Linux does head-scheduling for IDE; this eliminates
the supposedly different access pattern you're talking about. and reality
is that servers and workstations have quite similar patterns, since there's
precious little difference between them.
> : the main reasons for going
> : scsi are number of devices, cable length, and sometimes, if you really
> : need more than 20 MB/s sustained, speed.
> again, wrong. both ultra2 (80meg) and udma66 (66meg) far exceed the
> b/w of any single drive. the channel b/w is not the issue currently.
non sequitur. don't resort to expensive scsi unless you need it.
you need it if you have more devices than ide handles, need longer
cables, or if you really have a use for >40 MB/s sustained (which
two channels of cheap, off-the-shelf udma33 will deliver.)
> and while cpu util on udma drives has gotton better, the multitasking has not.
but "multitasking" is irrelevant when the OS is smart enough to sort
the command queue. Linux is.
------------------------------
From: Mark Hahn <[EMAIL PROTECTED]>
Subject: Re: SCSI vs. IDE
Date: 3 Aug 1999 16:01:55 GMT
>> How much of a difference is there between the two interfaces? I'm mostly
>> concerned with hard drives, but also CDROMs, CDRWs, and tape/zip/orb (is orb
>> supported by linux?) Hardware Central ( www.hardwarecentral.com ) claims that
>> the difference between scsi and ultra-66 ide is miniscule, and not worth the
>> added cost. I'd like to hear other opinions.
> The numbers look the same until you read the fine print about
> ultra33/66 being "burst mode", while scsi is continuous at
> high rate.
this is gibberish: SCSI bandwidth (for instance, 80 MB/s u2) is burst.
the burst transfer rate is basically irrelevant until you have enough
or fast enough devices to saturate it. since current fast disks run
20-30 MB/s, udma33 is fine. similarly, putting several fast disks on
a u2 channel delivers higher bandwidth, but at a fancy cost.
------------------------------
From: Mark Hahn <[EMAIL PROTECTED]>
Subject: Re: SCSI vs. IDE
Date: 3 Aug 1999 16:19:22 GMT
> Really depends on the use. A typical server will benefit from SCSI's
> concurrent access, queueing, etc. I use IDE for our server because our
think about this for a minute. the specific feature you're talking
about is tagged command queueing. it's pretty cool, but:
- disks have relatively small queues, as in ~100 entries.
often, SCSI systems will actually have TCQ disabled, or else
run with small (~20) entry queues.
- Linux sorts the IDE request queue. this is quite comparable
to how a SCSI disk optimizes it's queue, except that Linux
assumes that all seeks are monotonic in cost. thus the disk
wins when you're doing a seek that happens to lie at a lower block
address within the same physical track. whoopie! what a win!
- IDE controller and protocol latency has traditionally been
significantly lower than SCSI.
> fine and costs less. In a more typical office environment, where you
> have a lot of small files and many hits on the server at once, a SCSI
> will outperform IDE any day.
prove it. sounds like you're actually making a seek-time argument,
nothing to do with the interface. it's quite true that current high-end
SCSI disks claim ~5ms seeks, versus ~9ms for current IDE's. but don't
you expect your fileserver to have most of such small files cached?
> On an IDE chain, you can only access one drive (either slave or master)
> at a time. I'm not sure if this is true for the two channels on an IDE
> controller;
it's not. which is why one of the good reasons to go SCSI is if you
need more than 2-4 disks. these days, even 2 disks on the free/builtin
IDE controllers means you have up to 50 GB. tossing in another $20 UDMA33
controller gets you up to at least 100 GB.
in short, get SCSI when you:
- need support for many devices. more than one disk per IDE channel
is often a bad idea.
- need longer cables, at least with U2 (since SE SCSI has severe limits).
IDE is always limited to 18" or less.
- really do need more than 40 MB/s sustained, which a pair of cheap
UDMA disks will deliver on any current motherboard. single high-end
SCSI disks are clearly faster than current IDE's: pushing 30 MB/s,
versus around 20 MB/s. they also require serious thought about power
and case airflow.
- really do lots of small, random seeks. modern, 10K RPM SCSI's have
significantly lower seek times. this is a very odd workload, though;
I'd be interested in hearing about it if you have it.
- need hotplug
- need devices not supported by IDE (tape changers, etc).
- have enough devices to be worried about IDE's typical 500K hour MTBF,
versus SCSI's typical 1000K hour rating.
it's a travesty to see a "workstation" with a $200 SCSI controller,
and *single* $50/GB SCSI disk. the system would be much better off
with a free UDMA controller and $15/GB UDMA disk.
regards, mark hahn.
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Crossposted-To: comp.os.linux.setup,comp.os.linux.x
Subject: Re: Soundblaster Live - SuSE 6.0
Date: Sun, 01 Aug 1999 21:37:02 GMT
Reply-To: [EMAIL PROTECTED]
On 25 Jul 1999 16:54:41 GMT, chad evan peiper <[EMAIL PROTECTED]> wrote:
>HI there. I am running SuSE 6.0 and I have been struggling with getting my
>Soundblaster live to work. Could anyone, (if at all possible) send me the
>procedure to get drivers working.
>
>I recompile the kernel with modular sound
>i copy the sblive.2.0xxxx to the /lib/modules/2.x.y/misc/sbliv.o
>
>and then i do the insmod -f sblive and it does not work.
>i get the following:
>
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol unregister_sound_dsp
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol unregister_sound_midi
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol register_sound_dsp
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol register_sound_mixer
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol
>unregister_sound_mixer
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol
>unregister_sound_special
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol
>register_sound_special
>/lib/modules/2.0.36/misc/sblive.o: unresolved symbol register_sound_midi
>
>
>any help would be GREATLY appreciated. I have been trying for months.
>thanks
>Chad
>
>
>--
>||____________|O___I____,.::"|____}___I___Chad Evan Peiper__K9NOR_______
>||_o___|%.____|____I___|"____|___{____I____Violinist-of-all-trades___________
>||_o___|______|____I___|____O|.___}[EMAIL PROTECTED]
>||____O|___________I__O|______________I______University of Illinois, UC____
I believe that the sblive module is based on two kernel levels especially
made for redhat 2.2.5-15 and 2.2.9 or something like that. Is there one for
the 2.0 kernels? I have only seen the one for the 2.2.x kernels from their
ftp site.
--
Michael Perry - | No one can give you wiser advice
[EMAIL PROTECTED] | than yourself -Cicero
------------------------------
From: "A. Giachino" <[EMAIL PROTECTED]>
Subject: US Robotics Sportster 33.6 ISA modem supported?
Date: Tue, 03 Aug 1999 07:18:13 -0700
Hello,
I have a US Robotics Sportster 33.6 ISA modem which I am
having trouble configuring.
Is it supported?
It has jumpers for IRQ/COM and/or PNP mode.
With jumpers set, for example COM3, IRQ 5, wvdial fails to find
the modem. I used setserial to make sure the IRQ matches.
The phone line is plugged in.
With jumpers not set (PNP), isapnp reports an IRQ of 0, which is
bizzarre.
Anyone have any ideas on something else to try?
thank you,
AG
--
------------------------------
From: [EMAIL PROTECTED] (Paul 'Tok' Kiela)
Subject: Re: bttv and wintv problems
Date: 3 Aug 1999 13:36:30 -0400
In article <[EMAIL PROTECTED]>,
James Taggart <[EMAIL PROTECTED]> wrote:
>I am trying to get the bttv-6.0.3-17.i386.rpm to install, but it
>complains that there are deps needed, all of which happen to be
>installed. I then tryied bttv-0.5.22-3.i386.rpm which installs but bombs
>out during bootup. Any ideas? I am running this on a P2 333 with 320ram
>and redhat 6.0.
>
If you are sure that all dependancies will be met, just
rpm -ivh --force --nodeps bttv-filename.rpm ..
Good luck,
Paul.
--
-- Paul Kiela - Linux, IRIX, Solaris, NextStep
-- mailto: echo tokvgeminifphysicsfmcmasterfca | sed s/v/@/ | tr f .
"My title isn't Public Relations. If it was, I would have said something nice"
------------------------------
From: "Terrance" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.networking,linux.redhat.install
Subject: Problem with DE-528CT and RH6.0
Date: Tue, 3 Aug 1999 12:33:41 -0500
Just installed RH6.0. I have a D-Link DE-528CT PCI ethernet card that
gets
detected as a ne2k-pci clone during the installation, but the ne2k
driver
fails to initialize the DE-528CT card during boot. Does anyone have a
solution to this?
Please reply to "tlockett at ti dot com".
------------------------------
From: "t.j" <[EMAIL PROTECTED]>
Subject: ATI PC2TV NTSC output
Date: Tue, 3 Aug 1999 11:36:55 -0400
Hello All,
Does anyone know if it is possible to use the NTSC or S-Video ("TV") output
of an ATI PC2TV card with Linux? preferably with X.@ 640x480. I'm guessing
that if I setup the card under win95 and set the TV output on, then boot to
linux it may work, however I don't have win95 installed on that system, and
I'd hate to have to boot to 95 every time the machine is powered down just
to enable the TV output.
Any help would be appreciated.
Regards,
Tom
[EMAIL PROTECTED]
------------------------------
From: "Tsui Kwok-wai" <[EMAIL PROTECTED]>
Subject: SONY DVD-ROM DDU220E PROBLEM !!
Date: Tue, 3 Aug 1999 22:21:38 +0800
I USE RED HAT LINUX 6.0. IT DON'T DETECT SONY DVD-ROM DDU220E. ON THE
OTHER HAND, IT DON'T ASSUME IT TO CD-ROM. IT IS A BIG PROBLEM.
------------------------------
From: [EMAIL PROTECTED] (Frank v Waveren)
Subject: Re: help need a 3dfx vododo
Date: Tue, 03 Aug 1999 15:23:36 GMT
Isn't the vododo extinct?
--
Frank v Waveren
[EMAIL PROTECTED]
ICQ# 10074100
------------------------------
From: [EMAIL PROTECTED] (Craig Ruff)
Subject: Re: 3 button mouse
Date: 3 Aug 1999 08:22:36 -0600
In article <[EMAIL PROTECTED]>,
R.K.Aa <[EMAIL PROTECTED]> wrote:
>I find it uncomfortable to use but it works. IBM and SGI
>make excellent and "light touch" 3 button mice, intended for UNIX'es.
I second the use of an SGI mouse. I immediately tossed the Microsoft mouse
and the Dell keyboard that came with the Dell system I inherited for my
Linux box at work. Put an SGI keyboard and mouse on it and am much happier.
--
Craig Ruff NCAR [EMAIL PROTECTED]
(303) 497-1211 P.O. Box 3000
Boulder, CO 80307 Amateur Call KI0NO
------------------------------
From: [EMAIL PROTECTED]
Subject: Need kernel compile help
Date: Tue, 03 Aug 1999 17:07:22 GMT
Hi,
HELP! I must be doing something wrong here.
I have red hat 6.0 and a dual processor computer. I can not
get my modules to work after I recompile a new kernel.
When I boot to the new kernel, it will stop at the first instance of a
module. I can not boot at all if I compile my scsi controller as a
module.
If I compile all my critical stuff as non-modules, the non-critical
stuff won't work (like my sound card). Changing these modules
to non-modules and re-compiling fixes the problem.
I have attached a note I wrote myself on compiling a
new kernel. What am doing wrong?
Many thanks,
--Tony
[EMAIL PROTECTED]
Please email me as well as respond to the news group (my nntp server
only gets back about 1 in 10 replies and I have to dig them out out
dejanews).
================== Note on compiling a kernel:============
1) login as root
2) cd /usr/src/linux
3) startx
4) Note: Red Hat 5.2 and lower: If you are compiling an SMP
(multiple processors) kernel, edit "Makefile" and set SMP=1
(note that "1" stands for true not the number of processors)
5) make xconfig
Make desired changes
Red Hat 6.0 and higher: for SMP support is enabled in "Processor
Type and Features". You also must turn off APM support in "General"
AND enable "Enhanced Real Time Support" in Character Devices
6) make dep; make clean
7) "make zImage" or make "bzImage" if the kernel is large
8) make modules
Then rename the old modules direcory, example:
mv /lib/modules/2.2.5-15 /lib/modules/2.2.5-15-old
make modules_install
9) cp /usr/src/linux/arch/i386/boot/zImage /boot/vmlinuz.smp
or
cp /usr/src/linux/arch/i386/boot/bzImage /boot/vmlinuz.smp
Note: change vmlinuz.smp's name as needed. NEVER overwrite the
original kernel.
10) edit /etc/lilo.conf as needed. Example:
boot=/dev/sda
map=/boot/map
install=/boot/boot.b
default=linux.smp
#default=linux
prompt
timeout=150
image=/boot/vmlinuz.smp
label=linux.smp
image=/boot/vmlinuz-2.0.32
label=linux
root=/dev/sda3
initrd=/boot/initrd-2.0.32.img
read-only
other=/dev/sda1
label=dos
table=/dev/sda
11) fix initrd image
run the following command to creat a new initrd in the boot
directory. Make sure that the lilo.conf entry points to it.
/sbin/mkinitrd /boot/newinitrd-image 2.2.5-15
Where newinitrd-image is the name of the image in lilo.conf
and 2.2.5-15 is the current version of the modules you are
using (see "ls -alg /lib/modules")
12) lilo
Fix any error that pop up with 10 & 11
13) shutdown -r now
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: "Andrew J. Norman" <[EMAIL PROTECTED]>
Subject: Re: Configuring network speed for Netgear FA310-TX NIC
Date: Tue, 03 Aug 1999 14:36:52 GMT
=====BEGIN PGP SIGNED MESSAGE=====
To force the tulip driver into a specific mode you should included an
options string for when the modules is loaded as an example (in
/etc/conf.modules):
alias eth1 tulip
options tulip options=0 debug=1
Possible values for the options are 0-15 correspoding to:
0 - Auto
1 - 10base2
2 - AUI
3 - 100baseTX
4 - 100baseTX (Full Duplex)
6 - 100baseT4
7 - 100baseFX
8 - 100baseFX (Full Duplex)
9 - MII 10baseT
10 - MII 10baseT (Full Duplex)
11 - MII (auto)
12 - 10baseT (no auto)
13 - MII 100baseTX
14 - MII 100baseTX (Full Duplex)
15 - MII 100baseT4
Debug options are 0++ and correspond to more verbose output
Andrew J. Norman
______________________________________________________________
Dept. of Physics Phone: 757-221-3571
College of William & Mary [EMAIL PROTECTED]
"It is only with the heart that one can see rightly;
what is essential is invisible to the eye" -The Little Prince
______________________________________________________________
On Mon, 2 Aug 1999, Dragon wrote:
> I am in the process of setting up a number of computers for testing DSL
> modem hardware. I am using multiple Netgear FA310-TX NIC cards in each
> computer. Some of these cards connect to hubs or single speed interfaces
> so the line speed autodetect feature on the card works fine with them.
> Others are cabled directly between computers using cross-over cables,
> the line speed autodetect feature gets confused when I try to do this so
> I need to disable it.
>
> Does anyone know how to force the NIC driver to set a specific speed for
> each card? More specifically, I have four NIC cards in each machine,
> three of the cards need to be set to 10Base-T full-duplex and the fourth
> to 100Base-TX full duplex.
>
> The conf.modules file currently contains the following lines:
>
> alias eth0 tulip
> alias eth1 tulip
> alias eth2 tulip
> alias eth3 tulip
>
> What do I need to add to set each card speed? Or is this possible? I am
> almost to the point of going into the driver to hard-code it but I
> thought I would ask first.
>
>
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
>
>
=====BEGIN PGP SIGNATURE=====
Version: 2.6.3i
Charset: noconv
iQCVAwUBN6b+gyRqxyk0Hh01AQGEzAP/afhYCmXoxk9KCUiGWEoKJMv22ap7KlNA
xxArcDa+5oVSc5YUdR/rGI+A+WtBH7MIxn1SLDigu7Nmv9BORgyRwcQPZWN27kHs
j9jmV/qL+F3cklHhcC+erLJJkzLa4WGEZJecGYTfeH1+jOClj0qHF/sbzl4/Ibr7
r4A/fqRVEpk=
=QYlk
=====END PGP SIGNATURE=====
------------------------------
From: crux <[EMAIL PROTECTED]>
Subject: UPS for multiple linux machines?
Date: Tue, 03 Aug 1999 17:36:35 GMT
I need an UPS that works with Linux and will handle 2 PCs, not
necessarily with a monitor. It would be nice if it could support *one*
monitor as well if the UPS that can provide enough power for it (along
with the 2 PCs), but this is not mandatory. I also want the power
daemon on the Linux server connected to the UPS to use the network to
power down the other machine when power runs low.
I'm looking at an APC Back-UPS Pro 650---can it do what I just listed
above? Or is there any other UPS that can meet these demands? I am
aware that the Best UPS's are the best ones for Linux, but the price
seems a little steep. If there's a Best UPS that can do all of the
above and stay within my price range, please let me know. My maximum
price is about $300.
Would it be better to buy, say, two cheap little Back-UPS Pro 280's?
Last question is, do the APC UPSs come with a cable or is that
purchased separately?
--
++crux
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: "jacobite" <[EMAIL PROTECTED]>
Subject: Re: more applications for Linux
Date: Tue, 3 Aug 1999 11:48:07 -0700
[EMAIL PROTECTED] wrote in message
<[EMAIL PROTECTED]>...
> Look at this web page to learn more about the best new applications
>
<snip>
Perhaps YOU should search it for some decent newsreading software.
------------------------------
From: [EMAIL PROTECTED] (Thomas Zajic)
Crossposted-To: comp.os.linux.setup
Subject: Re: SCSI emulation layer & CDRW
Reply-To: [EMAIL PROTECTED]
Date: Tue, 03 Aug 1999 17:57:31 GMT
On Tue, 03 Aug 1999 18:22:23 +0200, Kees Schrama wrote:
> [ ... ]
> Furthermore, make a few symlinks from your SCSI-emulated (!) cdrom player
> /dev/scd0 or similar. Thus ln -s /dev/cdrom /dev/scd0 or something. Then all
^^^^^^^^^^^^^^^^^^^^^^^^^^
> applications that use /dev/cdrom will should work again.
Good idea, bad implementation. ;-) This would create a link
called /dev/scd0 pointing to /dev/cdrom. Instead, better try
'ln -sf /dev/scd0 /dev/cdrom'. On a side note, my CD-RW gets
recognized as /dev/sr0 on bootup. The original poster might
want to check his bootup messages to find the proper device
(dmesg | grep 'scsi CD-ROM').
> Change the permissions of the device with chmod +w /dev/scd0.
That won't help, since a CD-ROM _is_ not writeable. Note that
/dev/scd* and /dev/sr* are SCSI CD devices, non of which
is writeable. CD writing software uses the generic /dev/sg*
SCSI devices for writing. Instead, add '-o ro' to the mount
command, or add 'ro' to the options in /etc/fstab.
HTH,
Thomas
--
=--- Thomas Zajic aka ZlatkO ThE GoDFatheR, Vienna/Austria ---=
=-- "It is not easy to cut through a human head with a hacksaw." M.C. --=
=-- Posted with Free Agent 1.11/32 running on Linux 2.0.37/Wine-990731 --=
=--- Spam-proof e-mail: thomas(DOT)zajic(AT)teleweb(DOT)at ---=
------------------------------
From: [EMAIL PROTECTED] (Michael Perry)
Crossposted-To: comp.os.linux.x
Subject: Re: Diamond Viper 770 at 24-bit???
Date: Sun, 01 Aug 1999 21:27:52 GMT
Reply-To: [EMAIL PROTECTED]
On Sat, 31 Jul 1999 17:26:33 GMT, megasurg <[EMAIL PROTECTED]> wrote:
>Is there a way to make the Diamond Viper 770 display at 24-bit color? I
>mean one would think that a 32 Mb video card could handle such a thing
>but apparently do to the SVGA server it uses it's stuck at 16-bit
>color. Is there a way to tweak this so it can display higher colors?
>That's sad if not. Please help.
Have you tried the later RivaTNT drivers? We use that card at work on
several Dell systems and get some good resolutions and color depths.
The new driver is a binary replacement which gives us higher depths and
resolutions.
--
Michael Perry - | No one can give you wiser advice
[EMAIL PROTECTED] | than yourself -Cicero
------------------------------
From: Student <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.misc,comp.os.linux.setup,comp.os.linux.x,linux.dev.config,no.it.os.unix.linux.diverse,no.linux,uk.comp.os.linux
Subject: Re: Apache and ASP
Date: Tue, 03 Aug 1999 18:07:35 +0200
i know that there are asp plugs for apache but can they suport vb script ??
Ron G wrote:
> I've heard of a few approaches. One was already posted in response to your
> question. Here's another: http://www.chilisoft.com
>
> Joffer <[EMAIL PROTECTED]> wrote in message
> news:7o3s2q$s3e$[EMAIL PROTECTED]...
> > Does Apache support ASP?
> >
> >
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.hardware) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Hardware Digest
******************************
