Linux-Networking Digest #715, Volume #11 Tue, 29 Jun 99 07:13:39 EDT
Contents:
tcp_timestamps=1 gives lousy networking on 2.2.x !!!
about firewall (Patrick)
Ypbind errors and sloowww shutdown (Dan Alderman)
ethernet (Daniel Herkes)
fetchmail configuration ("Daniel Wagner")
Re: ethernet ("Frank Bauer")
Re: Two user accounts w/ different ISPs (Andrew Williams)
Re: DHCP/DNS/Behind FW => slow internet
Tool for Accounting on a DNS - Server (Uwe Peissig)
Re: Very big problem with RCP ("Graham Holt")
Re: Printing to Win NT ("Graham Holt")
Linux windozeNT network problem ("Peter De Zutter")
QOS traffic shaping kernel 2.2 (Lalot Dominique)
Re: Two user accounts w/ different ISPs (Bill Unruh)
rsh defunct processes? (Bill Comisky)
Re: Why not C++ (Nathan Myers)
Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was: (John Imrie)
Modprobe cannot load eth0??? (Frank Waarsenburg)
Re: caching only nameserver (Bernd Eckenfels)
Re: denying arp-requests (Bernd Eckenfels)
Re: Connecting Linux box to internet thru Sygate ("Al @Work")
----------------------------------------------------------------------------
From: <[EMAIL PROTECTED]>
Crossposted-To: redhat.networking.general
Subject: tcp_timestamps=1 gives lousy networking on 2.2.x !!!
Date: 29 Jun 1999 07:30:40 GMT
Hi,
The TCP timestamping cause my network to be extremely
slow.
When turned off everythings works fine!!
I used this ugly fix:
echo \000 -n > /proc/sys/net/ipv4/tcp_timestamps
I guess it should be in '/etc/rc.d/rc.local' or something
similar to be active at boot time.
I don't know the exact reason but it seems like packages
are fragmented and I also belive there are probelms
with sequence numbering. I haven't analysed the problem
in detail because I am no Linux guru (in other words I
don't know what to do).
Is this a well known problem/bug?
I am on kernel 2.2.5 (SuSE 6.1/RedHat 6.0).
/Christer
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: [EMAIL PROTECTED] (Patrick)
Subject: about firewall
Date: 29 Jun 1999 06:56:43 GMT
An Orielly book (Managing Internet Information Services p505) said:
"To telnet to external hosts, the user must first telnet to the bastion
host(where tn-gw runs), receive the proxt prompt and type connect
remotehostname. The FTP proxy is similar."
if i built a firewall using TIS FWTK,
then can we use the ordinary FTP program(e.g. WS_FTP)
behind the firewall?
--
------------------------------
From: Dan Alderman <[EMAIL PROTECTED]>
Subject: Ypbind errors and sloowww shutdown
Date: Tue, 29 Jun 1999 08:31:07 +0100
Hi,
I have a small problem with ypbind. Whenever the server get a shutdown
(Compaq Proliant 1600, Rh6.0) I get the following error.
ypbind forgot to set AF_INET in udp sendmsg. Fix It!
Anyone know what this is?
Also the Linux clients that run of this box take ages to shut down.
They're okay until they get to removing eth0 and the kernel/system
loggers. Is this a common problem? Any ideas?
Many thanks for any help in advance.
Regards,
Dan.
--
======================================================================
Dan Alderman Director WebUKs Ltd.
WebUKs Ltd.
10 Milner Road Phone: +44 20 8239 1924
Thornton Heath Fax: +44 20 8239 1924
Croydon Email: [EMAIL PROTECTED]
Surrey CR7 8JQ Email: [EMAIL PROTECTED]
UK
------------------------------
From: Daniel Herkes <[EMAIL PROTECTED]>
Crossposted-To: redhat.networking.general
Subject: ethernet
Date: 29 Jun 1999 08:30:42 GMT
Hi,
In my continuing struggle to adapt ancient computers to modern devices,
I have run into another snag. I have a small LAN setup at home, and the
etherpro 100 cards are working with the "other O.S." But while I have
gotten then to work with /usr/bin/netcfg, and can turn them off and on
with /usr/bin/usernet, I cannot get the NFS to work. I set the net
address's to 255.255.255.127 and 128 respectively. There is another
computer but not constantly connected to the LAN. I also disabled the
DNS entry for my ppp, and in fact the ppp during LAN sessions. But when I
attempt to invoke the NFS, to transfer Word Perfect files, I get a "can't
find the domain server" message.
Any ideas?
Thanks
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: "Daniel Wagner" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.setup
Subject: fetchmail configuration
Date: Tue, 29 Jun 1999 09:35:31 +0200
Hello everybody,
I've some probs with my fetchmail.
I use the following fetchmailrc:
poll pop.netway.at protocol pop3 username myuser password ushouldnotknow
When calling fetchmail in verbosemode I see that it's retrieving that =
messages from the remotehost, but I don't get them locally deliverd.
Can anybody tell me what I'm doing wrong.
Thanks Daniel.
PS: in sendmail's virtualusers, global... Files I've already aliases for =
my global mail addresses.
--=20
Java rulz! Linux rulz!
E-Mail: mailto:[EMAIL PROTECTED]
ICQ: 41472160
WWW: http://www.computer.privateweb.at/daniel.wagner/
------------------------------
From: "Frank Bauer" <[EMAIL PROTECTED]>
Crossposted-To: redhat.networking.general
Subject: Re: ethernet
Date: Tue, 29 Jun 1999 10:35:38 +0200
> DNS entry for my ppp, and in fact the ppp during LAN sessions. But when I
> attempt to invoke the NFS, to transfer Word Perfect files, I get a "can't
> find the domain server" message.
>
> Any ideas?
added the entry in hosts for the NFS-Host????
Frank
------------------------------
From: Andrew Williams <[EMAIL PROTECTED]>
Subject: Re: Two user accounts w/ different ISPs
Date: Tue, 29 Jun 1999 11:34:27 +0200
Reply-To: [EMAIL PROTECTED]
I do the whole lot with kppp and use 1 linux user-account for each ISP.
dkselich wrote:
> I have one linux box with two users and two different ISPs. I have no
> problem setting up one ISP and PPP0 and PPP1.But I have not found any
> documentation on how to handle the DNS of the other ISP.
>
> Dennis
--
Mielipiteet omiani - Opinions personal, facts suspect, especially on my
http://www.germanynet.de/teilnehmer/101/69082/samba.html
Simple Samba Solutions web page. ICQ 1722461
__________________________________________________________
| Fight Spam! Join EuroCAUCE: http://www.euro.cauce.org/ |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
------------------------------
From: <[EMAIL PROTECTED]>
Subject: Re: DHCP/DNS/Behind FW => slow internet
Date: 29 Jun 1999 07:30:42 GMT
Hi,
The problem was due to the TCP timestamping mechanism in
Linux. When it was turned off everything works fine!!
I used this fix:
echo \000 -n > /proc/sys/net/ipv4/tcp_timestamps
I guess it should be in '/etc/rc.d/rc.local' or something
similar to be active at boot time.
The TCP timestamp caused the network to be extremely slow.
I don't know the exact reason but it seems like packages
where fragmented and I also belive there where probelms
with sequence numbering. I haven't analysed the problem
in detail because I am no Linux guru.
Is this a well known problem/bug?
/Christer
Christer Olsson wrote:
> Hi,
> I have a networking problem with my Linux (SuSE 6.1)
> installation.
>
> At my company we reach the internet through a proxy
> server and I've got my IP address through a DHCP server.
> We also use aDNS.
>
> I can use all services inside our domain. Ftp, rlogin,
> telnet, e-mail and reach our intranet through Netscape 4.51
> (or some other browser like Arena). That is, I can reach
> everything that is not using the proxy.
>
> But as soon as I try to access some address outside our
> domain from a WWW browser everything gets *very* slow.
> When hit 'go' the proxy is contacted and the first
> info from the page pops up quite fast but then it all hangs.
> Sometimes forever and sometimes after a *long* wait the
> rest is displayed (it is a bit different depending on the address,
> some pages are totally stone dead).
>
> I have dual-boot on my PC, Linux/NT, and on Windows NT with
> exactly the same hardware, Netscape settings and DHCP server
> everything works fine. I also have a Sun Solaris 2.6 machine which
> also works perfect with the same Netscape settings.
>
> I can do 'nslookup' on our proxy and I have tried with a fixed
> IP address instead of DHCP but the same problem is still there.
>
> As far as I can understand my '/etc/resolv.conf', '/etc/rc.config'
> and '/etc/route.conf' looks fine. But of course I'am not an
> networking expert. Any hints that could lead me forward would
> by very much appreciated. Below I have attached some data and
> I can mentioned that I have used the SuSE configuration tool
> 'yast' to configure my network settings.
>
> thanks
> / Christer
>
> My machine name:
> ----------------
> mypc218 (mypc218.ks.andersson.se)
>
> Dynamically allocated IP address (DHCP)
> --------------------------------------
> 151.133.44.43
>
> /etc/resolv.conf
> ----------------
> search ks.andersson.se
> nameserver 151.133.40.28
> nameserver 151.133.40.14
>
> /etc/route.conf
> ---------------
> 151.133.44.0 0.0.0.0 255.255.255.0 eth0
> default 151.133.46.1
>
> 'netstat -rnv' gives:
> ---------------------
> Kernel IP routing table
> Destination Gateway Genmask Flags MSS Window irtt Iface
> 151.133.44.43 0.0.0.0 255.255.255.255 UH 0 0 0 dummy0
>
> 151.133.44.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0
> 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo
> 0.0.0.0 151.133.44.1 0.0.0.0 UG 0 0 0 eth0
>
> 'route' gives:
> --------------
> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref Use
> Iface
> mypc218.ks.ande * 255.255.255.255 UH 1 0 0
> dummy0
> ksnet44.ks.ander * 255.255.255.0 U 0 0 0
> eth0
> loopback * 255.0.0.0 U 0
> 0 0 lo
> default ksrsk6da.ks.and 0.0.0.0 UG 0
> 0 0 eth0
>
> 'cat /etc/hosts' gives:
> -----------------------
> 127.0.0.1 localhost
> 151.133.40.14 solstam.ks.andersson.se solstam
> 151.133.40.28 ms.ks.andersson.se ms
> 151.133.44.43 mypc218.ks.andersson.se mypc218
> 151.88.253.11 www-proxy.andersson.se www-proxy
>
>
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: [EMAIL PROTECTED] (Uwe Peissig)
Subject: Tool for Accounting on a DNS - Server
Date: Tue, 29 Jun 1999 11:43:12 +0200
Hi!!!!!!!!
I`m lokking for a tool who helps me with the Accounting under SUse 6.1.
We want to reach, that , when more than 5 requests /sec. are coming from
a single User, we get an Alarm or an Alert-Message or something like
that.
Is there a tool which offers this or a tool which makes the accounting
more comfortable.??????
Thanks
UP
------------------------------
From: "Graham Holt" <[EMAIL PROTECTED]>
Subject: Re: Very big problem with RCP
Date: Fri, 25 Jun 1999 10:11:58 +0100
You have the same users on all machines. Are the uid and gid s the same on
all machines for all users.
Graham
Giacomo Pasinetti wrote in message <[EMAIL PROTECTED]>...
>I have 3 different server in 3 different subnetworks connected by
>routers.
>The hosts file is ok.
>I have the /etc/hosts.equiv file with :
>SERVER USERNAME
>
>I have the same user on all servers.
>File and directory permission are ok ; but I can do rcp: It says
>permission denied.
>HElp !!!!! :-((
>
>Giacomo.
>
------------------------------
From: "Graham Holt" <[EMAIL PROTECTED]>
Subject: Re: Printing to Win NT
Date: Fri, 25 Jun 1999 10:19:20 +0100
I do this by sharing the NT printer and then set up from linux as an smb
printer. I t a user name and password on the NT for the linux box to
authenticate to.
Graham
jay nospam beatty wrote in message ...
>I'm trying to print from my RedHat 6.0 ,2.2.10, box to a postscript printer
>attached to a winNT workstation.
>
>The Printing HOWTO says there's a Print to Windows mini-HOWTO, but I can't
>find it.
>
>This must be done all the time. How do you do it? Where is the secret
>manual?
>
>Thanks.
>
>Jay
>
>
------------------------------
From: "Peter De Zutter" <[EMAIL PROTECTED]>
Subject: Linux windozeNT network problem
Date: Tue, 29 Jun 1999 11:55:26 +0200
Hey,
i have a problem to access a linux server(RH5.2+samba) and an openVMS aplha
server on a NT workstation 4.0 (sp4)
I can ping the machines with the IP adress, but not when you try to ping
with the machinename .
Can anyone help
I think it is a combination of serveral parameters in smb.conf and the
settings of the NT station.
Thanks in advance
Peter De Zutter
------------------------------
From: [EMAIL PROTECTED] (Lalot Dominique)
Subject: QOS traffic shaping kernel 2.2
Date: 29 Jun 1999 08:06:59 GMT
Hello,
I am looking for some documentation about QOS in the new kernel.
Alas, there is good programmers but nobody to write a single page about.
No man..nothing..
Does somebody find something?. I have no time to read the source and
the RFC as it is suggested by alexei.
Thanks
Dom
------------------------------
From: [EMAIL PROTECTED] (Bill Unruh)
Subject: Re: Two user accounts w/ different ISPs
Date: 29 Jun 1999 07:51:22 GMT
In <[EMAIL PROTECTED]> dkselich <[EMAIL PROTECTED]> writes:
>I have one linux box with two users and two different ISPs. I have no
>problem setting up one ISP and PPP0 and PPP1.But I have not found any
>documentation on how to handle the DNS of the other ISP.
a) Put all the DNS addresses into /etc/resolv.conf. It will try them
all. The only problem may be that it will be slow as it times out on the
first before trying the second.
Have a script to change the /etc/resolv.conf when youstart up each users
ppp.
ln -sf ~/resolv.conf /etc/resolv.conf.
This would mean that you would have to make /etc/writeable by everyone
(not a great idea).
make a subdirectory
etc/resolver
into which you put a file /etc/resolver/resolv.conf
make this directory writeable by everyone.
Then do
( as root)
ln -s /etc/resolver/resolv.conf /etc/resolv.conf
and as above put
ln -sf ~/resolv.conf /etc/resolver/resolv.conf
(Note each user then has his/her own resolv.conf in the home directory)
------------------------------
From: Bill Comisky <[EMAIL PROTECTED]>
Subject: rsh defunct processes?
Date: Tue, 29 Jun 1999 07:58:10 GMT
I have a program which reads and writes to/from stdin and stdout.
When I execute this program on a remote host using rsh (with input and
output redirection) I always get 2 rsh processes, one of which shows up
as <defunct>.
syntax:
bash$ rsh remote_host program < input > output
ps shows 2 rsh processes with one being defunct.
This is on a closed cluster where all of the nodes trust each other, so
there is no password required.
The PPID of the <defunct> rsh process is the PID of the non-defunct rsh
process. Both end when the remote program finishes executing. This
wouldn't bother me, except that I am using the PID's to keep track of
what jobs are running where, and spawning two processes with every rsh
command is making it messy.
I'm running the redhat 6.0 distribution.. anybody get the same results
or have an explanation/fix/workaround?
thanks,
bill
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.development.system,comp.lang.c++
Subject: Re: Why not C++
Date: 29 Jun 1999 00:47:32 -0700
Isaac <[EMAIL PROTECTED]> wrote:
>Nathan Myers <[EMAIL PROTECTED]> wrote:
>>Some people -- notably, the author of STL -- have no use for classes
>>(inheriting or otherwise), and promote templates above all.
>>
>I'd consider this opinion valid, but if your goal is object oriented
>programming, then why wouldn't a list which emphasizes those aspects
>be reasonable?
Right. He doesn't do object-oriented programming.
> My list is probably pretty close to the order of
>presentation in my learning material which probably does make
>the list suspect : }.
Precisely. The books grew, chapter by chapter, as the language
did. They give most space to earlier features, and treat the
later, probably more important, features as afterthoughts.
>Agreed. I draw it pretty much where you do, but I note operator
>overloading didn't make the expert's top five either. If I
>were writing a matrix library or a complex number handling
>package, I might have a more liberal opinion.
Right. Each feature that actually "got in" was very important
to some significant group of users.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: John Imrie <[EMAIL PROTECTED]>
Crossposted-To:
omp.os.ms-windows.nt.advocacy,comp.os.linux.advocacy,comp.infosystems.www.servers.unix,comp.os.linux.misc
Subject: Re: Could Microsoft Cheat On The New Mindcraft Benchmark? (was:
Date: Tue, 29 Jun 1999 11:52:45 +0100
Philip Brown wrote:
> On Sat, 26 Jun 1999 15:11:20 -0700, [EMAIL PROTECTED] wrote:
> >On Sat, 26 Jun 1999 18:15:46 GMT, Anthony Ord <[EMAIL PROTECTED]>
>wrote:
> >>On Sat, 26 Jun 1999 11:56:07 +0100, Robin Becker
> >>>a bit off topic, but an article in my paper, the Independent, states
> >>>that M$'s encarta has different versions for different countries. If M$
> >>>can claim in the US that Edison (October 1879) invented the electric
> >>>light bulb before Swan (February 1879) then a few adjustments to
> >>>benchmark results seem minor. Apparently the M$ mouthpiece says these
> >>>sort of 'facts' aren't always black and white etc etc.
> >>
> >>It's just to appease the American public. Just like the
> >>Second World War went from 1941 (when the Americans joined)
> >>to 1945. What was it before that? A bun fight?
> >
> > Does Encarta say that? American public school textbooks
> > certainly don't. Ours even covered the concentration camps.
>
> you mean, the american-run concentration camps?
>
Or the British run concentration camps
The Pilgrim
=============================
I have the one thing Mohamid al Fiad wants
a British Passport
------------------------------
From: Frank Waarsenburg <[EMAIL PROTECTED]>
Subject: Modprobe cannot load eth0???
Date: Tue, 29 Jun 1999 09:58:22 +0200
Hi,
I'm running RH6.0 without any problems. I thought. When I looked at the
syslog, I found that every time Linux starts, it writes to the log:
modprobe: cannot load module eth0:0
modprobe: cannot load module eth0:1
.
modprobe: cannot load module eth0:49
Yes, 50 lines all the way from 0 thru 49. And the same for lo!!!
Why? I've compiled my ne2000 clone into the kernel, the box runs OK
(connected to my local LAN, ISDN, runs DHCP, Samba and everything
works!!)
What causes these lines to appear?
Frank
------------------------------
From: Bernd Eckenfels <[EMAIL PROTECTED]>
Subject: Re: caching only nameserver
Date: 27 Jun 1999 02:10:06 GMT
Herbert Sauerer <[EMAIL PROTECTED]> wrote:
> is it somehow possible to get some information about which hosts my caching
> only nameserver already konws about? where is the data stored?
read the man page of your nameserver, it will tell u something like:
"ndc dump sends a signal to the nameserver to dump the data into /tmp/"
Greetings
Bernd
------------------------------
From: Bernd Eckenfels <[EMAIL PROTECTED]>
Subject: Re: denying arp-requests
Date: 27 Jun 1999 02:08:36 GMT
In comp.os.linux.networking teinmahl <[EMAIL PROTECTED]> wrote:
> like to make the packet-filter invisible for everyone.So I would like to
> make the filter to answer only certain arp-requests.
ifconfig eth0 -arp
and then use a user mode programm like arpsend to answer the queries.
Greetings
Bernd
------------------------------
From: "Al @Work" <[EMAIL PROTECTED]>
Subject: Re: Connecting Linux box to internet thru Sygate
Date: Tue, 29 Jun 1999 07:04:10 -0400
Brian,
I just did this last night. I run SyGate 3.0 and share my @Home service
among four PCs. The IP address you note for the LINUX box, 192.168.0.1, is
the same as the internal IP address on your SyGate box. It should be
DIFFERENT...My inner boxes have addresses in the range of:
192.168.0.1 SyGate gateway (Inner private LAN NIC)
192.168.0.2 Wife's machine
192.168.0.3 Laptop
192.168.0.4 LINUX RH 6.0 machine
I also setup routing to the 24.3.0.0 network (@Home in the Baltimore area)
through the 192.168.0.1 gateway address, and to the 192.168.0.0 network
through the 192.168.0.4 address. I also used a netmask of 255.255.255.0...My
/etc/resolv.conf file only has entries for the two @Home name servers and
the default name search path is set to "hosts, dns". You setup all of this
by running "linuxconf" as root.
Al
Brian Devlin <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> I have two computers (one is my family's). We have a cable modem hooked
> up to that box. It is running Win98 and uses sygate to share the cable
> modem. My computer is dual booting linux and Win98. I can access the
> internet using win98, but I am not sure how I should configure linux to
> access the internet.
>
> Here is a list of the settings.
>
> Sygate computer:
>
> Internal IP - 192.168.0.1
> SUBNET - 255.255.255.0
> DNS - 24.0.240.33
> 24.0.240.34
> HOSTNAME CX500477-a.shing1.ri.home.com
>
> Linux machine (Win98 settings):
>
> IP - 192.168.0.1
> SUBNET - 255.255.255.0
> GATEWAY - 192.168.0.1
>
> Any help would be appreciated. If you can just point me to a faq that
> would explain to me how to do it, I would appreciate that. You can
> e-mail me at [EMAIL PROTECTED] if you don't want to post.
>
> Thank You
> Patrick Devlin
------------------------------
** 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
******************************
