From:     Digestifier <Linux-Admin-Request@senator-bedfellow.mit.edu>
To:       Linux-Admin@senator-bedfellow.mit.edu
Reply-To: Linux-Admin@senator-bedfellow.mit.edu
Date:     Thu, 2 Dec 93 21:13:49 EST
Subject:  Linux-Admin Digest #199

Linux-Admin Digest #199, Volume #1                Thu, 2 Dec 93 21:13:49 EST

Contents:
  Re: sz problems (Gert Doering)
  Re: Difference between cua and ttyS (Gert Doering)
  Network interface lockup problems with Slackware (Nov 8th) (Seng-Poh Lee, Speedy)
  Re: SLIP server for MSDOG (David A. Nedved)
  Re: ld.so 1.3 needed (Kevin Cummings)
  Re: Difference between cua and ttyS (JAMES LEWIS NANCE)
  ftape with summit drive (Michael Peek)
  Re: Disk troubles (Gary Anderson)
  Re: FTP-Problems with LINUX and NCSA (Tom Murray)
  Re: Once again, using tape drives under Linux (Karsten Steffens)
  linux with NFS root filesystem?? (Brian J. Murrell)
  SCSI problem (Dr. Raimund K. Ege)
  How to set timezone information? (Maxim Matveev)
  Re: Once again, using tape drives under Linux (Piercarlo Grandi)
  Network Gateway... (Gareth Bult)

----------------------------------------------------------------------------

From: gert@greenie.muc.de (Gert Doering)
Subject: Re: sz problems
Date: Thu, 2 Dec 1993 10:45:32 GMT

hm@seneca.ix.de (Harald Milz) writes:

>Michael Lipka (lipka@lip.hanse.de) wrote:

>: > These problems come up again and  again.  You can't use the cua-kermit
>: > with the ttyS-mgetty for instance :-(( mgetty doesn't notice kermit on
>: > the corresponding cua-device and eats all characters from the modem.

>If you are talking about Gert Doering's mgetty+sendfax package, you're
>definitely wrong. This mgetty is _designed_ to work on ttySx solely. 
   ^^^^^^^^^^
Not quite. Michael just said that using kermit on cua won't work if
mgetty sits on ttyS - and there, he's right. Solution: remove mgetty -
*or* force kermit to use /dev/ttyS*. Unfortunately, some version of
kermit seem to insist on using /dev/cua... (but then, just symlink it
to ttyS).

>I have mgetty installed on ttyS1 and play UUCP, dial-up BBS, kermit, and
>all that on ttyS1. 

Well, I use ttyS0, but apart from that, same here.

gert
-- 
I've got a signature breakdown! Anybody got a spare one?

Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-3243328                         gert.doering@physik.tu-muenchen.de

------------------------------

From: gert@greenie.muc.de (Gert Doering)
Subject: Re: Difference between cua and ttyS
Date: Thu, 2 Dec 1993 10:47:57 GMT

pfau@cnj.digex.com (Thomas Pfau) writes:

>Use /dev/ttyS<n> for incoming (getty).  Use /dev/cua<n> for outgoing
>(kermit, lpd).

Naaaaaaah!

That's true for simple, stupid gettys. But it's plain false for
mgetty+sendfax.

For mgetty+sendfax, you have to use the *SAME* device for incoming and
outgoing calls. (mgetty will take care so that no collisions occur). Won't
work otherwise. Believe me. Or RTSL.

Anyway, the original poster asked for the difference between /dev/ttyS* and
/dev/cua* - and your answer described only the different use...

gert
-- 
I've got a signature breakdown! Anybody got a spare one?

Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-3243328                         gert.doering@physik.tu-muenchen.de

------------------------------

From: splee@pd.org (Seng-Poh Lee, Speedy)
Crossposted-To: comp.os.linux.help
Subject: Network interface lockup problems with Slackware (Nov 8th)
Date: 2 Dec 1993 15:43:43 GMT

I recently installed Slackware (dated 22nd Nov).  Everything went
smoothly.  Good job by the Slackware people! I installed from the
network via NFS and it went real fast.  5 minutes network configuration
and the system was up on the network!

The machine is a 486SX33 with a 3COM 3C503 card (IRQ 5, 0x300, 0xc800). 
Its configured to use DNS via a server.  Most everything works fine, but
the network interface locks up very often.  When this happens, you can't
ping it, but arp appears to work, cos the MAC address appears on my
Suns' cache.  On the machine itself, it runs fine except the network
interface appears to be dead.  You can telnet to itself fine (via
loopback), but nothing via eth0 works. 

NFS and the rpc stuff is enabled, and our network is pretty busy.  Could
that be a clue? I've also had occasions where portmap suddenly spawns
dozens of zombie processes for no reason.  This ties up all the
processes, and I get a 'can't fork' message and can't do anything. 
*sigh*

I've tried installing the PL14 kernel, but it still locks up. The next
step is to disable all the rpc and NFS stuff, which I'll try tomorrow.
[update]
I did this and the network interface STILL dies.

I used to run SLS 1.0.2, and got that to the point where the network
stuff was reliable. The current networking stuff looks pretty solid, as
long as I can find out what causes this lockup.

Well, while I've got your attention, some other networking related stuff
I've encountered;

I use remote printers and use printcap to point to them. They work,
but lpq always give a message of 'connection to xxx is down', where xxx
is the print server. lpr actually works ok, although the status file
in the spool directly is always left at 'sending file to xxx'. It
never changes to 'file sent to xxx'. The file IS printed out. lpc
reports that the queues are all up.

The rwhod always appears to die right away. Any ideas?

Having not made any changes in about 8 months, this upgrade was very
painless.  And the benefits are great! Kudos to all involved! To be able
to install via NFS, configure, and bring up networking and X in about 30
minutes is astounding!!

--
Seng-Poh Lee    <splee@pd.org>

------------------------------

From: dnedved@hubcap.clemson.edu (David A. Nedved)
Subject: Re: SLIP server for MSDOG
Date: Wed, 1 Dec 1993 16:21:35 GMT

In article <2ddv1m$nce@orion.cc.andrews.edu> aldy@orion.cc.andrews.edu (Aldy Hernandez) writes:
>Does anyone know of any SLIP server for msdos-- commercial or free
>(preferably free)?
>
>Please respond by e-mail
>
>Thanks,
>Aldy
>(aldy@andrews.edu)

Would anyone who replies also reply on the net also please?  There are
quite a few of us out here who would be interested.

Thanks

-- 
David Nedved
Department of Computer Science, Clemson University
dnedved@hubcap.clemson.edu
===============================================================================

------------------------------

From: cummings@hammer.CV.COM (Kevin Cummings)
Subject: Re: ld.so 1.3 needed
Date: Wed, 1 Dec 1993 21:21:20 GMT

In article <CGut40.5MC@eskimo.com>, scottw@eskimo.com (Scott Wolfe) writes:
> I need to have ld.so in order to run XFree86 2.0 (ver 1.3). Where can I 
> get this? I am running pl13 and decieded to upgrade XFree86 this morning.
> Any help would be great.

You should be able to find it wherever you get you GCC from.  I got mine
from the GCC directory on tsx-11.mit.edu.  It should be added to the
distribution packages when they add XFree86-2.0 since X needs it.

-- 
Kevin J. Cummings                       ComputerVision Services
cummings@kjc386.framingham.ma.us        cummings@primerd.cv.com

------------------------------

From: jlnance@eos.ncsu.edu (JAMES LEWIS NANCE)
Subject: Re: Difference between cua and ttyS
Reply-To: jlnance@eos.ncsu.edu (JAMES LEWIS NANCE)
Date: Thu, 2 Dec 1993 14:43:04 GMT


In article <CHELzy.EnA@greenie.muc.de>, gert@greenie.muc.de (Gert Doering) writes:
> 
> Anyway, the original poster asked for the difference between /dev/ttyS* and
> /dev/cua* - and your answer described only the different use...
> 

As I understand it (which may be flawed):

When a process opens /dev/cuaN /dev/ttySN blocks, that is whatever is trying
to read from /dev/ttySN will not receive any chars.  This allows programs
like getty to listen on /dev/ttySN and not get confused when a comm program
opens /dev/cua1.

Also:

/dev/cuaN does not receive job control signals.  That is if you say:

/bin/sh </dev/cua1 >/dev/cua1

You will end up with a shell running on /dev/cua1, but it will ignore things
like ^C and ^Z that come through /dev/cua1.

Hope this helps,

Jim Nance

------------------------------

From: peek@duncan.cs.utk.edu (Michael Peek)
Subject: ftape with summit drive
Date: 2 Dec 1993 17:51:37 GMT

        I've got a summit 250meg drive and I'm using kernel 13 with ftape.
If I copy a small file, nothing get's written to the tape.  If I copy a
large file, sometimes the file is truncated and the end is never written
to the tape.
        I've gone through the code and watched the traces.  Everything
seems to be correct - but the tape drive ignores the last write.  Has
anyone run across this problem before?  Does anyone have a solution or
a clue to what's going on?

        Sincerely,
                "Off my reel" Peek

                                peek@duncan.cs.utk.edu


------------------------------

From: anderson@sdd.comsat.com (Gary Anderson)
Subject: Re: Disk troubles
Date: 2 Dec 1993 17:45:29 GMT

Shekhar Mahadevan (mahadev@Erc.MsState.Edu) wrote:

: I am running Linux 0.99pl13r on my 486. A few days ago, my filesystem
: started to misbehave. Whatever space was freed was not becoming
: available for use. ie. No matter how much space I free, I don't have a
: block to write on. I tried to fix this problem using e2fsck (I even
: tried the latest version available from tytso@mit.edu.) No luck.

: Can somebody tell me how to fix this problem without having to go to
: extreme measures like backing up and/or reinstalling?

: Thanks

: Shekhar
: --

Shekar,
I had a similar problem just last night.  Here's what happened and what
I did to fix it.  There are probably other methods better than this, and
what I did may have only worked by a stroke of luck, or 'divine
intervention' (grin).  I had just, a couple of days ago, installed what
I thought was a new syslogd program.  Turned out it wasn't.  When I ran
'netstat -x', things started going haywire in a hurry.  The end result
was that after a couple of hours of the console being hung, the system
being non-responsive, and the disk drive light flickering about once a
second, I ended up with a /usr/log/kernel file that filled all remaining
space on my hard drive.  Deleting that file DID NOT reclaim the disk
space. 'df' still showed nothing available.  At this point I synced, did
a 'shutdown -r now' and let e2fsck do its thing.  Note: e2fsck MUST be
run on a filesystem which is mounted READ-ONLY !  Of course the fsck run
showed a couple of errors, put a massive file out in /lost+found, etc.
I deleted that file from the /lost+found subdirectory, and ran went
through the sync/shutdown/fsck routine a couple more times, and I know
have my disk space back, and no _apparent_ problems as a result (Yet).

To other reading this group, I'm relatively new to Unix/Linux (as if you
hadn't gathered that already), and the reason I posted this to the
group, instead of replying in mail, is I'd like some feedback as to
whether or not I did 'the right thing(tm)'.


--
|----------------------------------------------------------------------|
| Gary Anderson               | Internet: anderson@sdd.comsat.com      |
| Phone: 301-214-3367         |           ganderson@clark.net          |
| Fax:   301-214-7100         | Amateur Packet: N3KCD@WB3V.MD.USA.NA   |
| Comsat World Systems, Bethesda, MD                                   |
|----------------------------------------------------------------------|


------------------------------

From: tmurr@hw22.ma30.bull.com (Tom Murray)
Subject: Re: FTP-Problems with LINUX and NCSA
Date: Thu, 2 Dec 1993 18:59:08 GMT

In article <1993Dec1.205856.3142@eng.cam.ac.uk>, rda@eng.cam.ac.uk (R.D. Auchterlounie) writes:
|> hessenberger@maulaf.ISD.Uni-Stuttgart.DE (Klaus Hessenberger) writes:
|> 
|> >Hi,
|> 
|> >last month we installed some linux-machines on our network. One of the more serious 
|> >problems that we can't solve, was the file transfer from linux to our DOS-PC's.
|> 
|> >  Configuration:  DOS-PC - NCSA-Telnet 2.3
|> >                  Linux  - pl13 and pl14 
|> >                           ftpd from new-net2
|> >                           diku-ftpd
|> 
|> >Whenever I try to 'get' a file from a linux-machine that ist greater than 1023 bytes
|> >ftp freezes. Same happens with 'dir' in big directories.
|> >A workaround is to configure the DOS-PC as a telnet server (telnet -s). Then it is
|> >possible to log in from the linux-machines an 'put' and 'get' files.
|> >Ftp from DOS to other Unix-machines works.
|> 
|> >Ftp from other Unix-machines (SGI,Apollo) works well with the diku-ftpd.
|> 
|> >Is this a bug or a configuration-problem?
|> [...]
|> 
|> Bug, methinks - happens exactly as you describe right here too. 
|> What I haven't tried yet is newer than pl13 kernels - this may be a
|> good move as rumour has it that a few net-2 bugs have been fixed in
|> the interim.  It could also be NCSA - it has some problems with messages
|> but not quite the same symptoms...

        I'm having the same problem.  So far I have found that NCSA
can send to the linux box Much better than it can receive from the linux
box.  That is "put at the ncsa keyboard is far less likely to hang, than
get is.  Linux can put or get to NCSA (at least I haven't seen it hang).
        when it hangs, netstat -n reveals that the SEND_Q is usually stuck
with 4 to 5 k (characters?), and sits there forever.  netstat by itself
segmentation faults (number too large ?).  
        furthermore, both machines ethernet interfaces are ok.  you
can ping the NCSA box from linux.
        Linux to linux (just got the second machine up) seems to work
flawlessly.  Also, I'm using slackware 1.1.0, which I think has the latest
99.13.x, but not 99.14.
        Don't have answers yet, but maybe the data will shed some light.

|> 
|> Also I have had problems with FTP to a Novell 3.11 server - I always get
|> the error "Bad directory components" whenever I try to get or put anything.
|> Other un*x ftp clients connect to this server ok. 
|> Looks like Linux ftp has some problems...
|> 
|> -ray
|> <rda@eng.cam.ac.uk>
|> "Linux installation complete. Would you like your DOS disks boiled, fried
|>  or lightly grilled sir ?" 
|> 

-- 
== Tom Murray                   Zenith Data Systems              ==
== (508) 294-2285               300 Concord Road    MA30/MS853A  ==
== T.Murray@ma30.bull.com       Billerica, MA       01821-4186   ==

------------------------------

Crossposted-To: comp.os.linux.help
From: karsten@kshome.ruhr.de (Karsten Steffens)
Subject: Re: Once again, using tape drives under Linux
Date: Thu, 2 Dec 1993 05:15:20 GMT

Curtis Varner (cvarner@ucrengr.ucr.edu) wrote:
: In article <hua.754642748@base.cs.ucla.edu>,

:       Seems to work pretty well.  My thanks and appreciation to the
: author.

Hi,

        my thanks as well to the author. But did you ever make
a 'tar -dvf < /dev/ftape' after tarring to tape? I get lots of comparison
errors afterwards... That is due to the (still) missing bad block handling
of ftape. As nice as the driver is at times, and as much as I honour the
fine (and hopefully ongoing) work of Bas (the author), its clear that one
cannot rely on the backups. Hope its devellopping. And, what about including
ftape into the new kernel versions? I think its time now, due to the fact
that one cannot upgrade to a new kernel without putting lots of efforts
in adopting ftape.

Regards,
        Karsten

: Curtis Varner
: Student Programmer/Tech., Bio-Ag Library, Phy-Sci Library
: University of CA, Riverside
: cvarner@ucrengr.ucr.edu


-- 
==================>         Karsten Steffens        <=====================
   karsten@kshome.ruhr.de          |      steffens@ikp.uni-muenster.de
Marl - close to Recklinghausen     |         Institut fuer Kernphysik
  North of the Ruhrgebiet          |   Westf.Wilhelms-Universitaet Muenster

------------------------------

Crossposted-To: comp.os.linux.help
From: brian@ilinx.wimsey.com (Brian J. Murrell)
Subject: linux with NFS root filesystem??
Date: Thu, 2 Dec 93 19:51:16 GMT

What do you suppose is the feasability of a Linux system with a root
filesystem??

What I want to do is boot from a kernel image `zdisk' and have the 
system mount an NFS filesystem from our server as it's root filesystem.

This way I'd be able to take a floppy to any of our systems in house and
boot Linux and also have a half decent sized system.

I realize that I could just create a "ramdisk" root filesystem, but I
don't really want to use up such a precious resource (well it is in the 
systems in our office) as RAM for a root filesystem.  Besides having to
mount all of the subdirectories onto that root would be a real pain.

I also realize that I could use a floppy based root filesystem, but again
I don't want to use a slow floppy for my root filesystem, and I want to
be able to use that drive after the system is up.

Comments??

b.

-- 
Brian J. Murrell                                        brian@ilinx.wimsey.com
InterLinx Support Services, Inc.                      uunet!van-bc!ilinx!brian
North Vancouver, B.C.                                             604 984 2545
        Platform and Brand Independent UNIX Support - R3.2 - R4 - BSD

------------------------------

Crossposted-To: comp.os.linux.development
From: ege@cerl.fiu.edu (Dr. Raimund K. Ege)
Subject: SCSI problem
Date: Thu, 2 Dec 1993 20:04:43 GMT

after installing a UltraStor 34F SCSI card and a SEAGATE 1.05GB disk
I get the following warning upon reboot:

        Uhhuh. NMI received. dazed and confused, but trying to continue
        You probably have a hardware problem with you RAM chips

Linux continues to boot, recognizes the disk, and runs fine as usual.
I am running .99pl14. At first the SCSI card did not work: it
used IRQ 14, which I changed to IRQ 11. (all other setting are
still the factory default: BIOS c8000, IO base 330H).

What does the warning indicate ? Should I change some other settings ?
Do I have to worry ? Should I remove the card ?

-- 
Raimund K. Ege                             School of Computer Science
                                             Florida Int'l University
ege@scs.fiu.edu           (305) 348-3381              University Park
ege@servax.bitnet     FAX (305) 348-3549              Miami, FL 33199

------------------------------

From: mmatveev@boi.hp.com (Maxim Matveev)
Subject: How to set timezone information?
Date: Thu, 2 Dec 1993 23:48:04 GMT

Hi,

Can anybody explain me, how to set timezone information on Linux?
First, I simple set TZ environment variable - no effect,
after I try to play with settimeofday. Sure it set desired time,
but not timezone.

Really this is the honor for me to live near London, but
please bring me back to Idaho :)

Max
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Max Matveev                            (208) 396-7900 (work)
mmatveev@hpbs669.boi.hp.com            (208) 385-9103 (home)

------------------------------

Crossposted-To: comp.os.linux.help
From: pcg@aber.ac.uk (Piercarlo Grandi)
Subject: Re: Once again, using tape drives under Linux
Reply-To: pcg@aber.ac.uk (Piercarlo Grandi)
Date: Fri, 3 Dec 1993 00:42:55 GMT

>>> On Thu, 2 Dec 1993 05:15:20 GMT, karsten@kshome.ruhr.de (Karsten Steffens) said:

  [ ... on the floppy tape support for Linux ... ]

Karsten> my thanks as well to the author. But did you ever make a 'tar
Karsten> -dvf < /dev/ftape' after tarring to tape? I get lots of
Karsten> comparison errors afterwards... That is due to the (still)
Karsten> missing bad block handling of ftape.

Floppy tapes are a baaad idea. However for masochists that want to use
them, recovery from the unreliability of the medium should not be done
in the driver, but at the user level. There are a few filters that add
ECC; for example there is an ecc package on every GNU FTP site. Pipe tar
into that...

------------------------------

From: gareth@gblinux.demon.co.uk (Gareth Bult)
Subject: Network Gateway...
Date: Sat, 20 Nov 1993 12:46:43 GMT

I have a SLIP connection to Internet via a Linux box that is also on a LAN
using thin ethernet. I can't seem to get packets to go from another node on
the LAN thru the internet connection (& back)... I've tried all manner of
route command.. Anyone know how this is done???

TIA
Gareth.                                                                     

------------------------------


** 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: Linux-Admin-Request@NEWS-DIGESTS.MIT.EDU

You can send mail to the entire list (and comp.os.linux.admin) via:

    Internet: Linux-Admin@NEWS-DIGESTS.MIT.EDU

Linux may be obtained via one of these FTP sites:
    nic.funet.fi				pub/OS/Linux
    tsx-11.mit.edu				pub/linux
    sunsite.unc.edu				pub/Linux

End of Linux-Admin Digest
******************************