Linux-Networking Digest #219, Volume #11 Thu, 20 May 99 21:13:52 EDT
Contents:
Re: mix hub and bnc (Mircea)
sendmail (Nick)
Segmentation Fault. What to do? (Denis)
Re: FTP doesn't work through masquerading (James Myers)
Re: NFS service not responding (Johannes Nix)
Re: NIS database (Johannes Nix)
Re: 100Mbps Ethernet trouble (John Kim)
Re: PPP Help Please... (Vilmos Soti)
Re: General.NFS.Questions ("Christopher R. Thompson")
Intel ISDN voice/date card for Linux (Nelson Yan)
Re: ftp server problems- Unable to load Interpreter, ([EMAIL PROTECTED])
Re: LINUX ability to handle large beowulf clusters (Brian Vicente)
How to make my Linux box a terminal to talk on serial port ttySx... ("Brian")
Re: How to use smbmount (Piotr Lukasz Bychowski)
Re: General.NFS.Questions ("Christopher R. Thompson")
----------------------------------------------------------------------------
From: Mircea <[EMAIL PROTECTED]>
Subject: Re: mix hub and bnc
Date: Thu, 20 May 1999 20:02:17 -0400
What you need is a hub that has a BNC (10b2) port, beside the RJ45
(10bT) ones. Check www.pricewatch.com
MST
peter wrote:
>
> is this possible ?
>
------------------------------
From: Nick <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: sendmail
Date: Thu, 20 May 1999 23:23:16 GMT
I have a linux box running RedHat 6.0. How do I set up Sendmail to
become a POP3 server. I have heard about how it is supposed to work,
but every time I try to connect, I get an error message. I am trying to
connect with the Netscape Communicator mail service.
--Nick
------------------------------
From: Denis <[EMAIL PROTECTED]>
Subject: Segmentation Fault. What to do?
Date: Thu, 20 May 1999 02:45:48 -0400
I installed rvplayer the other day, enjoied it for a while.
Today I can't run it, get a segmentation fault after I enter "rvplayer".
Anybody knows what to do when you get a segmentation fault?
Thanks.
Denis
------------------------------
From: [EMAIL PROTECTED] (James Myers)
Subject: Re: FTP doesn't work through masquerading
Date: Thu, 20 May 1999 23:46:38 GMT
I had the same problem, until I set my FTP program to use passive
transfer mode. I believe most FTP clients have this option somewhere.
I didn't modify the masquerading box at all (didn't load the ftp
module).
James
On Mon, 17 May 1999 16:12:59 GMT, "David Murray" <[EMAIL PROTECTED]>
wrote:
>I just got my new Masquerading box setup. It is all working fine (at least
>I thought) and is masquerading for 2 Linux Boxes, 2 Windows95, and a
>Macintosh. However, I noticed yesterday that if I use FTP, I am able to
>connect and even do things like change to binary or cd to a directory. but
>when I do a simple "ls" to list the files, it locks up the FTP program.
>Sometimes it comes up with an error that there is no route to host. If I
>go directly to the IP Masquerading box and run FTP from there (or telnet in
>and run FTP) it works fine.. just not on any of the client machines.
>Otherwise, HTTP and POP3 and such work just fine. I am even able to use
>RealAudio for streaming data on a client machine.. So why can't I do an
>"ls" inside FTP?!
> Just for info, my masquerading machine is RedHat 5.2, running the 2.0.36
>kernel.
>
>--DavidM
------------------------------
From: Johannes Nix <[EMAIL PROTECTED]>
Subject: Re: NFS service not responding
Date: 21 May 1999 02:27:11 +0200
Bill Clark <[EMAIL PROTECTED]> writes:
I am not sure if this is relevant. NFS uses rpc and portmap, both go
trough the tcpd and name lookup via NIS uses portmap also.
That means that if you use domain names for rpc/nfs/portmap services
in hosts.allow, this could cause name lookup via NIS, which could
cause portmap requests, which would cause name lookup, and so on.
Therefore, hosts allow should contain something like
portmap: 123.456.789
and perhaps also
rpc.nfsd: 123.456.789
rpc.mountd: 123.456.789
please correct me if the service names or anything else are wrong.
> >
> > # hosts.allow This file describes the names of the hosts which are
> > # allowed to use the local INET services, as decided
> > # by the '/usr/sbin/tcpd' server.
> >
> > ALL:bell.xxxx.xxx
> >
> > ###This is the only machine that needs to see this machine!
> >
> > # cat hosts.deny
> >
> > # hosts.deny This file describes the names of the hosts which are
> > # *not* allowed to use the local INET services, as decided
> > # by the '/usr/sbin/tcpd' server.
> >
> > # The portmap line is redundant, but it is left to remind you that
> > # the new secure portmap uses hosts.deny and hosts.allow. In particular
> > # you should know that NFS uses portmap!
> >
> > ALL:ALL
> >
==================================================
[EMAIL PROTECTED]
==================================================
------------------------------
From: Johannes Nix <[EMAIL PROTECTED]>
Subject: Re: NIS database
Date: 21 May 1999 02:33:33 +0200
Do a
#cd /var/yp
#make
and show us the result.
You may need to add NIS users to
/etc/passwd
or create a NIS passwd file, e.g. in
/etc/NIS/passwd
and adjust the file name in
/var/yp/Makefile
Johannes
==================================================
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (John Kim)
Subject: Re: 100Mbps Ethernet trouble
Date: 21 May 1999 00:36:11 GMT
Frank Bauer ([EMAIL PROTECTED]) wrote:
: > Now, I'm trying out Fast Ethernet. I just got a generic FastEthernet card
: > for my Win98 machine and connected it to a 100/10 autoswitching hub. The
: > new hub seems to work fine for all machines except for the Linux box. All
: ...
: normally you need a switch not a hub for connection at different speeds and
: normally an autowitchting hub means, that it would detect the speed
: automatically for the whole hub (NOT FOR EVERY SINGLE PORT!!! - this does
: only a switch)
: HTH
: Frank
>From reading the manual, it seems to be able to combine both 10bT and
100bT networks on the same hub. But even if what Frank is saying is true,
it still doesn't explain why it's only the Linux machine that has trouble
when it's hooked up to the 10/100 autoswitching hub.
--
=====================================================signature begins here
John D. Kim [DK] FC @ MMF [EMAIL PROTECTED] ICQ # 4736158
donotfearthepenguins...donotfearthepenguins...walkintothelight...
=======================================================signature ends here
------------------------------
Date: Thu, 20 May 1999 09:35:12 +0000
From: Vilmos Soti <[EMAIL PROTECTED]>
Subject: Re: PPP Help Please...
Roy wrote:
>
> Greg Aeschliman wrote:
>
> > I'm starting to go a bit crazy trying to get connected to my ISP here.
> > I've read every FAQ How-to I've gotten my hands on and still can't get it
> > going. Here's a bit from the last time I logged on. I'm a complete
> > Linux newbie so please dumb it down for me. Here goes the log:
> >
> > May 20 18:25:45 localhost pppd[6343]: pppd 2.3.5 started by root, uid 0
> >
> > May 20 18:25:45 localhost pppd[6343]: Using interface ppp0
> >
> > May 20 18:25:45 localhost pppd[6343]: Connect: ppp0 <--> /dev/cua2
> >
> > May 20 18:26:16 localhost pppd[6343]: LCP: timeout sending Config-
> > Requests
> >
> > May 20 18:26:16 localhost pppd[6343]: Connection terminated.
> >
> > May 20 18:26:16 localhost pppd[6343]: Receive serial link is not 8-bit
> > clean:
> >
> > May 20 18:26:16 localhost pppd[6343]: Problem: all had bit 7 set to 0
> >
> > May 20 18:26:26 localhost pppd[6343]: Exit.
> >
> > It will connect but then pppd will terminate. This is as close as I've
> > gotten and it's taken me some time to get this far. Any ideas?
> >
> > Thanks,
> >
> > Greg Aeschliman
>
> sorry/testing nr
Hi,
Your problem is "Receive serial link is not 8-bit clean.
Check this out.
Vilmos
------------------------------
From: "Christopher R. Thompson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.protocols.nfs,comp.os.linux.help,alt.linux,alt.os.linux,comp.os.linux.misc,comp.os.linux.questions
Subject: Re: General.NFS.Questions
Date: Thu, 20 May 1999 01:03:43 -0700
Christopher R. Thompson wrote:
> Christopher R. Thompson wrote:
>
> > Christopher wrote:
> >
> > > In article <[EMAIL PROTECTED]>,
> > > Christopher <[EMAIL PROTECTED]> wrote:
> > >
> > > > I have recently been attempting to build a Linux Beowulf type cluster
> > > using
> > > > the Red Hat distibution Linux-Pro 5.4. The Linux kernel version is
> > > 2.0.33 using
> > > > rpc.nfsd Universal NFS Server Version 2.2beta16 and rpc.mountd
> > > Universal NFS
> > > > Server Version 2.2beta37.
> > > >
> > >
> > > Questions:
> > >
> 4. What do all of these logging messages mean? I don't know and I don't
> care!
Now "I know" why they suggest that you "NOT" install nfs-root clients in
Linux.
After perusing many many megabytes of rpc.nfsd -d fhcache messages it would
seem that a file handle is some number generated by the host. When a client
requests to see a file that resides on the host it uses it's own file name
initially. The host looks up in the cache to see if the file name already has
a handle-number attached to it. If it does and it is belongs to another client
it is deleted and a new handle-number is generated and passed to the new
client. When the old client tries to access it's file using the old
handle-number... I don't know what happens exactly but the result is "STALE
FILE HANDLE" operation aborted. That can be a boot-up operation, a shutdown
operation, or any operation in between when two or more clients reference
filenames similarly.
We have 3 clients 192.168.50.3, 192.168.50.4, and 192.168.50.5 labled
respectively 3, 4, and 5. I am currently typing on 5.
It would also appear that NFS-FILE-HANDLES are defaulted to be removed after 3
minutes of inactivity if unused. That is how I can be typing now on client 5
while clients 3 and 4 are also active.
If two or more clients attempt to run "BOOTPD" or "BOOTPS" or "INIT" or "TOP"
or who knows how many other programs at the same time concurrently on this
system these programs will fail. If you are patient however... 3 minutes at
least, depending on your rpc.nfsd timers, it is possible to get them all up
and running.
Here's the configuration:
Host 192.168.50.1 has hard disk with directories in / named
/tftpboot/192.168.50.3/
/tftbboot/192.168.50.4/
/tftpboot/192.168.50.5/
each of the above directories has the following subdirectories:
"/bin /boot /dev /etc /home /initrd /lib /mnt /net /proc /root /sbin /tmp /usr
/var".
Only the directories "/proc /usr /home" are empty. The others have useful
programs, control files, system specific device files, libraries and other
such phemomena required by a running Linux operating system.
The problem (it appears) is that when two or more of the clients require the
same file name "/sbin/init" at the same (within 3 minutes give or take) time a
"STALE-FILE-HANDLE" situation occurs. This can give rise to a few different
symptoms.
1. On "shutdown -whatever (now/or whenever)" you may get "INIT: PANIC:"
"SEGMENTAION VIOLATION! GIVING UP..".
2. On Boot-up... you may... or... may not! With various disconcerting and
spurious results identified in subsections below.
a). Modules will not load therefore access to File Systems ETH0, FAT, VFAT,
MSDOS, ISO9660, ...etc will not occur.
I). Normally this will only occur following a Kernel compile and a make
modules install thus having the distracting effect of making you think that
somehting went wrong with the Kernel or module compilation. When in reality is
was simply that you booted and ran init up to point where libraries were
linked and the stale (recent) usage by another computer 3 or 4 caused 5 to
miss something. If only I could be patient for just 3 minutes.
b). Various NFS mounts such as "/bin /home /lib /sbin /usr" fail to occur
even though they previously appeared to mount on certain occasions.
I). Normally this will only occur following some modifications to
"/etc/hosts, /etc/hosts.allow, /etc/exports, /etc/services, /etc/fstab" or
some other inocuous file that affects the NFS access system. Thus allowing you
to percieve that the fault must lie in the changes that you have just
incurred. Beware... the real reason was because "/etc/rc.d/init.d/network"
initaliaztion script... access's a file called
"/etc/sysconfig/network-scripts/ifcfg-eth0" among others and for some reason
yet unknown a stale (recent) file handle got deleted causing the network
initialization to fail "somewhat" and all of the ARP cache is now gone
forever!.
3. On running various programs that require files in the
/tftpboot/192.168.50.x/ file-directory system you can get simultaneous
"STALE-NFS-FILE-HANDLES". Even though they are not stale at all. They are
actually quite recent and have probably just been deleted.
This all seems to occur because RPC.NFSD cannot distingush the difference
between "/var/run/utmp" and "/tftpboot/192.168.50.5/var/run/utmp" and
"/tftpboot/192.168.50.4/var/run/utmp" and
"/tftpboot/192.168.50.3/var/run/utmp". They all look like "/var/run/utmp" to
RPC.NFSD.
I have recently been trying to find WWW.NFS.NET to see if maybe there is some
source code to be had or a patch for this particular problem but "WWW.NFS.NET"
does not now seem to be alive. Does anyone know if there is a mirror site
around?
Regards,
Most Curious George, and I don't know why?
------------------------------
Date: Thu, 20 May 1999 16:20:15 +0800
From: Nelson Yan <[EMAIL PROTECTED]>
Subject: Intel ISDN voice/date card for Linux
==============5D6E131659D050468EEB5DE0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Can any body tell where I can find driver of Intel ISDN card for Linux.
Thanks,
NY
--
Nelson Yan
Zi Corporation (H.K) Limited
Room 509-510, Lippo Centre Tower II, 89 Queensway, HongKong SAR
Tel: +852 2525-5283 Fax: +852 2530-3526
http://www.zicorp.com.hk/
==============5D6E131659D050468EEB5DE0
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: 7bit
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Can any body tell where I can find driver of Intel ISDN card for Linux.
<p>Thanks,
<p>NY
<pre>--
Nelson Yan
Zi Corporation (H.K) Limited
Room 509-510, Lippo Centre Tower II, 89 Queensway, HongKong SAR
Tel: +852 2525-5283 Fax: +852 2530-3526
<A HREF="http://www.zicorp.com.hk/">http://www.zicorp.com.hk/</A></pre>
</html>
==============5D6E131659D050468EEB5DE0==
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: ftp server problems- Unable to load Interpreter,
Date: Fri, 21 May 1999 00:33:27 GMT
Sorry to keep posting snippets, but I just noticed more information
that is displayed if I try a few hundred connections at once. Stuff
like:
pam[4660]: _pam_init_handlers: no default config /etc/pam.d/other
pam[4687]: unable to dlopen(/lib/security/pam_listfile.so)
This is a totally fresh copy of RedHat 5.2 installed using the "Server"
automated method.
Thanks,
Will
[EMAIL PROTECTED]
--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---
------------------------------
From: [EMAIL PROTECTED] (Brian Vicente)
Crossposted-To: comp.os.linux.development.system,comp.os.linux.advocacy
Subject: Re: LINUX ability to handle large beowulf clusters
Date: Fri, 21 May 1999 00:47:04 GMT
What are the reasons for Linux not being able to support more than
300?
I can't think of anything that has that limitation.
Jon Roberts <[EMAIL PROTECTED]> wrote:
>I was talking with a friend who is researching a supercomputer design
>that may become one of the top 10 machines in the world. I mentioned
>Beaowulf Linux clustering and the response was that Linux is good but
>can't support more than about 300 nodes in a cluster (and they plan on
>being larger than that). Is this true? If there IS a way around this
>limitation I'd like to know so I can make further recommendations for
>Linux to him.
------------------------------
From: "Brian" <[EMAIL PROTECTED]>
Subject: How to make my Linux box a terminal to talk on serial port ttySx...
Date: Thu, 20 May 1999 08:34:19 GMT
Hi:
Very simple.
I want to talk to my pm-11 with my Linux box on ttySx
(unused serial port).
How do I do it - should be real simple.
Best regards,
Brian
------------------------------
From: Piotr Lukasz Bychowski <[EMAIL PROTECTED]>
Subject: Re: How to use smbmount
Date: Thu, 20 May 1999 10:42:02 +0200
On Tue, 18 May 1999, Takashi Masuda wrote:
> I want to mount a Windows shared folder to a Linux mount point.
> I have setup smb.conf. Linux directries can be accessed from Windows boxes.
> I tried following commands.
> smbmount \\\\(NetBIOS Name)\\(Shared Folder) (Linux mount-point)
> smbmount \\(NetBIOS Name)\(Shared Folder) (Linux mount-point)
> smbmount //(NetBIOS Name)/(Shared Folder) (Linux mount-point)
> smbmount /(NetBIOS Name)/(Shared Folder) (Linux mount-point)
Perhaps You should try
smbmount ////(NetBIOS Name)//(Shared Folder) (Linux mount-point)
;-))
But seriously, have You tried to display resources by command
smbclient -L (NetBIOSName)
to check if Your linux can really see it?
And what about the system You are using? Because I've tried those upper on
Red Hat 5.1 (5.2) and it works. On RH 6.0 You have to enter
smbmount ////(NetBIOS Name)//(Shared Folder) -c 'mount mount-point'
And the most important: read the manual (of course if you haven't done it
already)
> But those were all failed. What should I do?
> Thanks in advance.
>
> Takashi Masuda
> K.K.Rocky Japan
>
>
>
------------------------------
From: "Christopher R. Thompson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.protocols.nfs,comp.os.linux.help,alt.linux,alt.os.linux,comp.os.linux.misc,comp.os.linux.questions
Subject: Re: General.NFS.Questions
Date: Thu, 20 May 1999 01:44:35 -0700
This is a multi-part message in MIME format.
==============98D78A7D805504464C15A267
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Christopher R. Thompson wrote:
> Christopher R. Thompson wrote:
>
> > Christopher R. Thompson wrote:
> >
> > > Christopher wrote:
> > >
> > > > In article <[EMAIL PROTECTED]>,
> > > > Christopher <[EMAIL PROTECTED]> wrote:
> > > >
> > > > > I have recently been attempting to build a Linux Beowulf type cluster
> > > > using
> > > > > the Red Hat distibution Linux-Pro 5.4. The Linux kernel version is
> > > > 2.0.33 using
> > > > > rpc.nfsd Universal NFS Server Version 2.2beta16 and rpc.mountd
> > > > Universal NFS
> > > > > Server Version 2.2beta37.
> > > > >
> > > >
> > > > Questions:
> > > >
>
==============98D78A7D805504464C15A267
Content-Type: text/plain; charset=us-ascii; name="chris.nfs"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="chris.nfs"
I should have stated in my previous posting the following:
Aside from the problems mentioned... everything else seems to be working very well.
This only occurs for NFS-ROOT mounted clients. Once a host directory such as/lib is
mounted to /tftpboot/192.168.50.x/lib... the "STALE-FILE-HANDLE" problem disappears.
I am also only aware of this problem occuring on a Linux Host running the previously
mentioned revisions of RedHat and rpc.nfsd.
The client generates the handle "/var/run/utmp" and passes it to the host which looks
for the file in the clients mount tree and assigns a handle number. This handle number
is then used by the client in subsequent file i/o operations to acces the file. When a
second client attempts to access a different "/var/run/utmp" in a different mount tree
(from the hosts point of view) the host gets confused and delete's the prior clients
file handle. This is probably the correct thing to do when two clients are accessing
the same physical "/var/run/utmp" in the hosts directory tree.
Curiously George, and I don't know why?
==============98D78A7D805504464C15A267==
------------------------------
** 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.networking) 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-Networking Digest
******************************
