Linux-Hardware Digest #537, Volume #10 Sun, 20 Jun 99 06:13:34 EDT
Contents:
Linux and Near-Line storage (Maisel)
DIVX is now dead --- Circuit City gives up after a year of brainwashing (DIVX SUCKS)
Re: linux memory size problems (David E. Fox)
Re: Creative Graphics exxtreme driver? (paches)
Re: linux on 386 (Andy Adams)
Re: Linux Backup Solution. (David E. Fox)
SMP on Dell 4300 server ("chris - newnews")
Re: Coax-Ethernet Problems ("Gene Heskett")
Help:Install linux boot on RAID. (building farms of servers) (Abe Lin)
problems with 17Gb disk (Bob Berman)
how to configure on board Ethernet chip? (Bob Berman)
Re: Tape question ("Gene Heskett")
Re: Strange problem with my RH6.0 box... ("William Willis")
----------------------------------------------------------------------------
From: Maisel <[EMAIL PROTECTED]>
Subject: Linux and Near-Line storage
Date: Sun, 20 Jun 1999 00:24:45 -0600
Hi!
Does anyone know anything regarding "near-line" storage using Linux?
I have a 41GB optical Juke that works reat with automount and the jukebox
module, but I would love to find some near-line stuff.... I'ts hell trying
to keep track of where everything is!!!
Thanks!
Lee
------------------------------
From: [EMAIL PROTECTED] (DIVX SUCKS)
Crossposted-To: comp.os.linux.misc,comp.os.linux.advocacy
Subject: DIVX is now dead --- Circuit City gives up after a year of brainwashing
Date: Thu, 17 Jun 1999 03:37:18 GMT
As of today 6/16/99, Circuit City's evil greedy pay
per view DVD format known as DIVX is out of business. They have
discontinued operations and will be offering some refunds to
customers. I suggest that if you bought one of their (BETA) DIVX
machines that you try to get your money back. And all your DIVX
encrypted discs that can not play on any other machine that you try to
get a refund for them as well. In the next few days much info will
be posted on the anti-divx sites on how to ask for your refund and
what you should do now.
If you own DVD then today is a great day for you. From now on
their should be no DIVX exclusive titles. If a movie studio wants
your money they are going to have to let you own your movies and not
just rent them. Sorry Circuit City. Movies like Titanic are now
available on DVD and hopefully some of the good ones we have all been
waiting for will soon be on DVD.
More Information:
http://www.fightdivx.com/
http://www.divx.com/
http://www.zdnet.com/zdnn/stories/news/0,4586,2276806,00.html
DVD Deals
http://www.fightdivx.com/dvd-deals.htm
------------------------------
From: [EMAIL PROTECTED] (David E. Fox)
Subject: Re: linux memory size problems
Reply-To: [EMAIL PROTECTED]
Date: 20 Jun 1999 07:35:24 GMT
In article <7jpdm8$up6$[EMAIL PROTECTED]>, Andy wrote:
>My motherboard has 4 memory slots. (2 banks of 2 slots) When I installed
>my machine I started out with 2 8M sticks in bank 0. Every thing worked
Same here. I'm presently suffering from a similar problem.
>fine installing and linux runs great. I noticed, though, that While Bios
>reports the correct ammount of memory, linux only sees 13520 (K?).
If you mean that it reports 13520 in the 'free' line, this is normal,
because the free command is only reporting memory that is in
userspace. The kernel and some 384K (more or less) will occupy the
remainder.
>So I went about trying to get it to recognize my memory by putting
>"linux mem=16M" at the linux boot prompt. But whenever I do this the
>boot process croaks with a general protection fault at the point where
>the system starts to do a Partition check on hda. I've tried various
This is happening to me, but for a different reason. I started, like
you, with two sticks in bank 0, for a total of 16. I later added
two more sticks of 16 meg each for a total of 48 meg, and that
worked for a while. But recently I decided I needed even more
memory, so I obtained (by mail order) two 32meg sticks, and put
those in bank one.
Theoretically, I should have 96 megs total, but the BIOS only
reports / counts up to 64. And, the two simms I did get from
the mail order place are physically distinct: one has two rows
of chips (one on the front, one on the back) and the other one
does not -- leading me to believe that one of the simms is 16
megs.
Nevertheless, if I boot up linux normally (I have an append line
in /etc/lilo.conf setting it to use 96 megs) I get the same
exact error you do, a kernel panic just as it does the partition
check. (Why there, and not somewhere else, I have no idea.) If
I tell it to use only 64megs, it runs just fine.
>Andy
--
========================================================================
David E. Fox Tax Thanks for letting me
[EMAIL PROTECTED] the change magnetic patterns
[EMAIL PROTECTED] churches on your hard disk.
=======================================================================
------------------------------
From: paches <[EMAIL PROTECTED]>
Subject: Re: Creative Graphics exxtreme driver?
Date: Fri, 18 Jun 1999 22:39:16 +0000
Yeah, it works great on the XWindows server. I got the one with 8meg
ram. A friend installed
the driver for me, so I don't know the name of it (sorry). But, it's
definately possible.
-Paches
"bernard.deleplanque" wrote:
> Is it possible to install a graphics exxtreme card with Linux??
>
> [EMAIL PROTECTED]
------------------------------
From: Andy Adams <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: linux on 386
Date: Sun, 20 Jun 1999 01:38:23 -0500
Stefano Ghirlanda wrote:
> Hi everyone,
> I happen to have an 386 with which I would like to do something useful,
> but mostly play with :-)
>
> Now, I still have to open it to see what's inside (got it from a friend
> for free) but I would like to know if there is some info about minimum
> hardware requirements to run a simple linux system.
>
> My plan is to use it as a loghost for a small network (5-6 machines) so
> that it should only allow syslogd or simlar connections and from these
> machines only. There doesn't need to be any X or fancy stuff or even user
> programs beyond what's needed to examine the logfiles.
>
> Any suggestions or pointers are welcome...
> thanks a lot,
> Stefano
>
> --
> Stefano Ghirlanda, Zoologiska Institutionen, Stockholms Universitet
> Office: D554, Arrheniusv. 14, S-106 91 Stockholm, Sweden
> Phone: +46 8 164055, Fax: +46 8 167715, Email: [EMAIL PROTECTED]
> Support Free Science, look at: http://rerumnatura.zool.su.se
I've run linux on a 386/25 with 4mb of ram. It was painfully slow and I
ended up scraping the machine. You'd probably want to recompile
your kernel and try to keep things lean. The biggest problem I had
was with disk space. I was trying to install redhat5.1 on a 119MB drive.
For some reason the install script HAD to install perl and several other
things that I didn't want. I think the absolute minimum is 2mb of ram.
I would guess that compiling your drivers into the kernel would speed
up the time to boot.
Hope this helps. Good luck.
Andy
------------------------------
From: [EMAIL PROTECTED] (David E. Fox)
Subject: Re: Linux Backup Solution.
Reply-To: [EMAIL PROTECTED]
Date: 20 Jun 1999 07:18:26 GMT
In article <7k8c6b$ljn$[EMAIL PROTECTED]>, killbill wrote:
>I don't understand why the multi-volume option and the compress option
>are mutually exclusive...
Well, to put it simply, tar if used with compression will compress
a media's worth of data and then write it. I first noticed this
long ago when attempting to back up to floppies. I, of course, wished
to save on the # of floppies needed, and at first glance it looked
like tar w/compression and w/multi-volume would be enough to do the
job. But it isn't, and it's because tar doesn't know how much space
is *really* available and usable if it were to do compression. The
compression is essentially done as a separate process, and tar just
knows that floppies can hold 1.4 megs of data.
It's not that the two options are necessarily compatible, but if you
use them it won't buy you anything. tar will happily compress a
floppies worth of data, write it, and then ask you for the next
disk. Meanwhile, much of the disk is unused, and you end up using
the same number of disks either way.
>As a response to that problem, I wrote a GPL package called backburner
>(see www.freshmeat.net) that will chop up, fixate, and recombine
>streams. It is a pretty simple collection of Perl scripts, that could
Which probably addresses the issue. The key thing here is that you
need to compress, and then watch the size of your compressed file, and
when it approaches your limit (e.g., 1.4 meg for a floppy) you stop the
compression and write the chunk out to floppy.
>Bil Kilgallon ([EMAIL PROTECTED])
--
========================================================================
David E. Fox Tax Thanks for letting me
[EMAIL PROTECTED] the change magnetic patterns
[EMAIL PROTECTED] churches on your hard disk.
=======================================================================
------------------------------
Reply-To: "chris - newnews" <[EMAIL PROTECTED]>
From: "chris - newnews" <[EMAIL PROTECTED]>
Subject: SMP on Dell 4300 server
Date: Sun, 20 Jun 1999 10:08:13 +0100
Has anyone got this to work relyably ?
I have tried 2.2.3 and 2.2.9 kernels and both have caused
intermitant panics (about one per day) when using both CPU's.
Tthe same kernel's with SMP disabled have been solid for a week now.
any pointers as to what works on this board would be helpfull
Chris Cain
------------------------------
Date: 20 Jun 99 04:43:39 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Re: Coax-Ethernet Problems
Gene Heskett sends Greetings to Eric Wick;
EW> Hello,
EW> just using 4 Linux-Computers onto a 26m Coaxline. Very often the
EW> troughput breaks down to 400kB/s instead of full 1100kB/s.
EW> Turning the T-Adapters sometimes help, i think the contacts will
EW> be not the best material. Any idea to get better contact on this
EW> pieces?
EW> I hate RPC-Timeouts and such things:-/
EW> One line end is such grounded, both ends have 50 Ohms
EW> Terminators. What about the Cables? Some of them have AL-Foliet
EW> and Copperweb, some other habe only the copperweb. Will this give
EW> a badly shielding?
I'd suspect the mix of cable types might have something to do with it,
and poor installation of the connectors might account for something too.
I haven't tried to look this up in 50 ohm stuffs, but in 75 ohm video
cabling, I'm a huge fan of Clark Wire & Cables type 7558, which has both
the foil shielding AND a healthy copper braid overlay. And 'Kings' or
'Amp' crimpon connectors. The cable is extremely flexible, and
connector retention is such that I get a year's service out of them in
the studio, hooking up monitors and moving them around, and as prompter
cables, which get drug around as the cameras move.
I'd think, that if Clark makes a 50 ohm twin to that cable, and your tee's
were decently silverplated (I've had lots of trouble with the 'chrome'
ones), it might make a difference.
BTW, that foil shielding? By itself, copper braid with '98%' coverage
will give shielding effectiveness in the 50-55 db range. Add the foil
inside the braid, and it goes up another 60-70 db! 60 db is a million to
1 reduction. In this case, you aren't worried about leakage signal
getting out, but other garbage getting in.
I closed this, and then noted your .de address, so Clark isn't known
there. But there should be a supplier of such that is accessable to
you.
Cheers, Gene
--
Gene Heskett, CET, UHK |Amiga A2k Zeus040 50 megs fast/2 megs chip
Ch. Eng. @ WDTV-5 |A2091,GuruRom,1g Seagate,CDROM,Multiface III
|Buddha + 4 gig WDC drive, 525 meg tape
|Stylus Pro, EnPrint, Picasso-II, 17" vga
RC5-Moo! 690kkeys/sec isn't much, but it all helps
email gene underscore heskett at iolinc dot net
--
------------------------------
From: md5�ϼs�[EMAIL PROTECTED] (Abe Lin)
Subject: Help:Install linux boot on RAID. (building farms of servers)
Date: Sat, 19 Jun 1999 18:11:35 GMT
Hi, guys,
We're gonna build a farm of servers. The config is like this:
PIII 400, Asus 100 Mhz MB with SCSI, Mylex AcceleRAID, 5x 9 G HD
(seagate baracuda). cheapo video card. 512MB ram.
Here's the help-wanted issues:
a.Can we configure AcceleRAID so Linux sees the 45G as a big HD, and
still be able to boot from it?
b.The HD of our choice is not suuported on the
ftp://ftp.mylex.com/pub/dac960/diskcomp.html for AcceleRAID, but it is
for those DAC960PL/PDU/PG. Can we install on those disks still?
Those are Seagate Baracuda ST19171W.
c.I'm not sure about the CPU issue, I do hope this setup would be
better than what we had:
Ultrasparc 143Mhz, 196Meg RAM. 2xSun 2.1G HD, 1xSeagate 39173. total
13Gig. Cannot believe that we packed like 200 sites on it.
------------------------------
From: Bob Berman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: problems with 17Gb disk
Date: Sat, 19 Jun 1999 14:22:43 -0500
I recently bought a new 17Gb disk and tried to fdisk it and mke2fs it. I have
the Large Disk Howto and the Upgrading Mini Howto and am using them extensively.
My motherboard is an ASUS P2B-LS - brand new - with the latest BIOS as far as I
can tell.
I want the disk partitioned into 5Gb - Win98, 500 Mb - swap, 6Gb Linux,
6 Gb. Linux. I am running Linux 2.2.9, Pentium II 450, Maxtor Quasar 17.2 Gb
Ultra DMA disk.
My problem is that when I go to make a new fs on the Linux partition, I get all
sorts of error messages.:
Can any one help me? What is this all about? I have the disk fdisk'd (Linux
version) as follows:
58-> fdisk /dev/hdc
The number of cylinders for this disk is set to 33483.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)
Command (m for help): p
Disk /dev/hdc: 16 heads, 63 sectors, 33483 cylinders
Units = cylinders of 1008 * 512 bytes
Device Boot Start End Blocks Id System
/dev/hdc1 1 10159 5120104+ c Win95 FAT32 (LBA)
/dev/hdc2 10160 11175 512064 82 Linux swap
/dev/hdc3 11176 23366 6144264 83 Linux native
/dev/hdc4 23367 33483 5098968 83 Linux native
When I then run mke2fs, it goes ok for a while and then prints many error
messages:
mke2fs 1.06, 7-Oct-96 for EXT2 FS 0.5b, 95/08/09
warning: 263 blocks unused.
Linux ext2 filesystem format
Filesystem label=
1542000 inodes, 6144001 blocks
307213 blocks (5.00%) reserved for the super user
First data block=1
Block size=1024 (log=0)
Fragment size=1024 (log=0)
750 block groups
8192 blocks per group, 8192 fragments per group
2056 inodes per group
Superblock backups stored on blocks:
8193, 16385, 24577, 32769, 40961, 49153, 57345, 65537, 73729,
81921, 90113, 98305, 106497, 114689, 122881, 131073, 139265, 147457,
155649, 163841, 172033, 180225, 188417, 196609, 204801, 212993, 221185,
229377, 237569, 245761, 253953, 262145, 270337, 278529, 286721, 294913,
...............................................................
6053889, 6062081, 6070273, 6078465, 6086657, 6094849, 6103041,
6111233, 6127617, 6135809
Writing inode tables: 12/ 750
26/ 750
41/ 750
.................
256/ 750
Warning: could not write 8 blocks starting at 2097180 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097188 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097196 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097204 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097212 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097220 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097228 for inode table: No such
file or directory
Warning: could not write 8 blocks starting at 2097236 for inode table: No such
file or directory
Writing superblocks and filesystem accounting information: done
If I now try to mount and copy files onto the partition, it takes a long time
to mount and many copys fail with "Not enough space" - even though there are
6-7Gb free in the partition.
Help! please.
Bob
------------------------------
From: Bob Berman <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: how to configure on board Ethernet chip?
Date: Sat, 19 Jun 1999 14:42:59 -0500
I believe that I need to enable support for my on board Ethernet chipset in the
kernel somewhere. My motherboard - ASUS P2B-LS has what it says is an Intel
82558 Ethernet 10/100 chipset. I can not find any information about how to
enable this in the kernel (2.2.?). I selected "EISA,PCI,VLB and on board
controllers", but none of the controllers listed below that option resemble the
82558. How do I get this Ethernet working?
------------------------------
Date: 20 Jun 99 05:06:08 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Re: Tape question
Gene Heskett sends Greetings to Zoltan Kocsi;
I have had similar problems in the past with poorly terminated cabling,
and by a badly written device driver. In that case I sent the drive
back to Seagate, whom I'm beginning to think likes to see their products
over the fall holidays as they seemed to fail about that time of the
year.
When it came back, it still didn't work, so I changed the driver and
card back to a much slower combo here. That worked till the next
holiday season. Since that was the 3rd year in a row, I changed the
brandname on the drive nameplate to TecMar.
Its been 2 years now, and still going great.
Exabyte has an excellent reputation, however the NS8 looks like a travan
design, and while I have 3 of them that work great with HP labels on
them, the basic design has been panned here before. As we keep a
multilevel backup, I don't know if we have 20 full passes on one of
those TR-4 cartridges yet. We have had some failures, but that was
traced to our software, and the tapes have since been re-used.
So, check your cabling, terminations and see if there is an updated
driver for the device available. Also, if you've been using a cleaning
tape, good luck, those things eat heads for lunch. I much prefer to
gain access to the heads, and wipe them with a q-tip soaked in alcohol,
paint thinner type, not rubbing, which is worthless for such uses, way
too much water in it.
ZK> I have an Exabyte Eagle NS8 tape drive, with some Travan tapes.
ZK> After using a tape a *little* while, when reading from the tape
ZK> I get I/O errors and /var/log/messages has the following message:
ZK> Jun 14 21:50:42 tade kernel: st0: Error with sense data: Current
ZK> error st09:00: sense key Medium Error
ZK> Jun 14 21:50:42 tade kernel: Additional sense indicates
ZK> Unrecovered read error
This could be a symptom of a card driver thats not fully compliant with
the 'scsi direct' operating modes.
ZK> Since I'd like to use this drive and the tapes for system backup
ZK> purposes, a tape going unusable after about 20 rewrite cycles
ZK> doesn't give me much faith ...
ZK> The tape was not mishandled, touched, dropped, magnetised,
ZK> whatever. Actually when it broke it was sitting in the drive,
ZK> which read its previous content (success), wrote new one
ZK> (success) and tried to verify it (error). The tape is completely
ZK> unusable now.
ZK>
ZK> Are tapes indeed that unreliable or I have bad luck with this
ZK> particular tape ?
ZK> Thanks,
ZK> Zoltan
Cheers, Gene
--
Gene Heskett, CET, UHK |Amiga A2k Zeus040 50 megs fast/2 megs chip
Ch. Eng. @ WDTV-5 |A2091,GuruRom,1g Seagate,CDROM,Multiface III
|Buddha + 4 gig WDC drive, 525 meg tape
|Stylus Pro, EnPrint, Picasso-II, 17" vga
RC5-Moo! 690kkeys/sec isn't much, but it all helps
email gene underscore heskett at iolinc dot net
--
------------------------------
From: "William Willis" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.help,comp.os.linux
Subject: Re: Strange problem with my RH6.0 box...
Date: Sun, 20 Jun 1999 10:09:59 +0100
The problem you have is an invalid boot sector, put in a bootable dos disk
with fdisk on and type "fdisk /mbr" which will sort that out.
then get a boot floppy for linux and try to boot linux from it (using the
old rawrite to get boot and root floppies from dos) then, when you get back
in to linux re - run lilo.
------------------------------
** 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
******************************
