Linux-Networking Digest #410, Volume #11 Fri, 4 Jun 99 21:13:42 EDT
Contents:
terrible NFS performance with Solaris 2.6 (Robert E. Brown)
Lowdown on using Linux as gateway fo Browsing and Email (Georg Cantor)
Re: Linux: ICMP Redirect, IP Source Routing unterdruecken (Detlef Bosau)
Re: Linux - Windows emulators? ("Cardz")
Re: Why pump instead of dhcpc in RH v6.0? (Iain O'Cain)
Re: 2.0.33->2.2.7 issues (Omri Schwarz)
Re: SOCKS5 ("Curt")
Re: Route (Radovan Brako)
Re: Linux: ICMP Redirect, IP Source Routing unterdruecken ("Robert E.Wijnberg")
base64 - how do I decode? (Tony)
rshd, rlogind, maybe other daemons, in 2.2.7 (Omri Schwarz)
Re: Redhat 6.0 & ipmasqerade problems (Jonas Sextl)
Re: 3c575 pcmcia problem on IBM laptop ("Matt Graham")
REPLICATION between 3 servers .... ("Marcel v.d. Steeg")
3c905b problem with 2.0.36 (bill goodacre)
3C509 NIC and RH 5.2 ("BTRiggs")
Restrict Access via NIC address? (Bob Dusek)
Re: Forwarding of broadcasts possible? (Bob Dusek)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Robert E. Brown)
Subject: terrible NFS performance with Solaris 2.6
Date: 04 Jun 1999 10:54:27 -0400
I have a Linux box running Red Hat 6.0 that mounts via NFS a file system on a
Sun Ultra 2 computer running Solaris 2.6. Network communication between the
two machines seems fine -- FTP transfers use the available bandwidth.
However, NFS write performance from the Linux box to the Sun is really
terrible. Read performance is acceptable.
Does anyone have any recommendations for improving performance? Thanks.
bob
------------------------------
From: [EMAIL PROTECTED] (Georg Cantor)
Subject: Lowdown on using Linux as gateway fo Browsing and Email
Date: Fri, 04 Jun 1999 23:48:30 GMT
Reply-To: [EMAIL PROTECTED]
I have just installed Linux for a workgroup. I now want to setup Email
and browsing using the Linux as a gateway.
Of special interest to me is using Linux to download email from a
virtual server using the company domain name.
Where can I get fairly detailed facts on the whole process?
------------------------------
Date: 03 Jun 1999 23:03:00 +0200
From: [EMAIL PROTECTED] (Detlef Bosau)
Crossposted-To: de.comm.internet.routing
Subject: Re: Linux: ICMP Redirect, IP Source Routing unterdruecken
[EMAIL PROTECTED] meinte am 03.06.99
zum Thema "Re: Linux: ICMP Redirect, IP Source Routing unterdruecken":
> [EMAIL PROTECTED] (Detlef Bosau) writes:
>
> >Ich stimme mit Dir ueberein, dass Routing ausgedient hat.
>
> Jetzt wuerde mich dann doch mal Deine Alternative zu "Routing"
> interessieren.
Ist jetzt ziemlich off topic. Aber ich meine schon seit langem,
dass die Alternative zu einem gerouteten, paketierten, Datenverkehr
in einem Datenverkehr besteht, in dem Peers, die verbindungsorientierten
Datenverkehr moechten (und das ist IMHO doch der haeufigere Fall)
direkt ueber eine virtuelle Verbindung miteinander kommunizieren.
Wie in einem PVC bei FR oder ATM.
Hier wuerden keine paketweisen Routingentscheidungen fallen,
auch die Paketinspektion waere deutlich einfacher als bei IP Packet
Switching. Ich denke schon, linked source routing Verfahren wie
bei ATM oder Frame Relay, bei dem Routing eigentlich ein
irrefuehrendes Wort ist, weil eben keine Wegewahl_entscheidung_
getroffen wird sondern bei Verbindungsaufbau durch Signalisierung
verabredete Verbindungstabellen abgegriffen werden, wuerden die
Kommunikation erheblich erleichtern.
Es gaebe dann schlicht kein TCP mehr. Es erscheint mir persoenlich
seit langem widersinnig, verbindungsorientierte Leitungen auf
Layer 2 kuenstlich auf einem Layer 3 in Pakete zu zerteilen und
auf Layer 4 wieder Verbindungen zu simulieren.
Ein bedarfsmaessiger Verbindungsaufbau zwischen den Datenstationen
mit einem NBMA mit linked source route switching, wie ich
das eigentlich immer nenne, ich finde immer kein vernuenftigeres
Wort dafuer, auch in der Literatur habe ich dazu nichts gefunden,
halte ich persoenlich fuer organischer, fuer verwaltbarer (siehe z.B.
QoS), leichter fehlerkorrigierbar (zwischen den Knotenrechnern
wuerde ein Synchroncode mit Fehlerkorrektur laufen) als
die bisherigen Protokollstapelungen, in denen sich einfach
viele Dinge wiederholen.
Ich persoenlich kann mir gut vorstellen, dass man das ganze zum Client
hin sogar wie IP over Ethernet aussehen laesst. Mehr noch. Ich koennte
mir sogar vorstellen, ueber RSVP oder dergl. so zu tun, als
koennte ich RSVP und wuerde richtig schoenes QoS mit IP machen.
Aber in der Folge wuerde icn nur noch die reinen Nutzdaten ueber
einen PVC tunneln, wobei dieser der Ressourcenanforderung
entsprechend verwaltet waere und die ganze Darstellung in Form
bisheriger Protokollpakete nur noch ein lokales Artefakt waere.
Ueber den PVC ginge nur noch TCP Nutzlast. Kein Paketheader mehr.
Wege_entscheidungen_ waehrend des Transportes wuerden entfallen,
es wuerden vorgefertigte Switchingtables abgefragt.
Nur so als Idee.
Detlef
--
Detlef Bosau [EMAIL PROTECTED]
Bienroder Weg 79 Tel.: +49 531 303383
D2: +49 172 6819937
38106 Braunschweig, Germany Fax: +49 531 303364
>>>> PGP Public Key als Empfangsbestaetigung <<<<
## CrossPoint v3.1 R ##
------------------------------
From: "Cardz" <[EMAIL PROTECTED]>
Subject: Re: Linux - Windows emulators?
Date: Fri, 4 Jun 1999 18:42:27 -0500
Yea if you have enough ram vmware is definietly the way to go.
Alex Meisel wrote in message <[EMAIL PROTECTED]>...
>Jon Finanger wrote:
>>
>> Does such emulator exist?
>> (to run windows programs in linux env.)
>Hi Jon!
>
>Have a look at: http://www.vmware.com
>
>Regards,
> Alexander Meisel
------------------------------
From: [EMAIL PROTECTED] (Iain O'Cain)
Subject: Re: Why pump instead of dhcpc in RH v6.0?
Date: 4 Jun 1999 22:44:56 GMT
Reply-To: [EMAIL PROTECTED]
Steve Snyder <[EMAIL PROTECTED]> (and some others) wrote:
>What is the advantage of pump over dhcpc?
It's smaller, faster, lighter, and tastes great, according to Red Hat.
What I've found is that AFTER making sure I've got their patched copy,
which has the same version number on it as the copy shipped with Red
Hat Linux 6.0, it fails to set IP or any other information. A 5.2
system with dhcpcd on the same network is fine.
In any case, it doesn't look like pump puts the DHCP-provided
information anyplace easy to retrieve. I've got to have an ipchains
script read stuff like the IP address, but it's not clear where I'll
actually get that.
Good luck with it.
- I
------------------------------
From: Omri Schwarz <[EMAIL PROTECTED]>
Subject: Re: 2.0.33->2.2.7 issues
Date: Fri, 04 Jun 1999 20:49:19 -0400
Stefan Monnier wrote:
Stefan Monnier wrote:
>
> >>>>> "Omri" == Omri Schearz <[EMAIL PROTECTED]> writes:
> > copying .config from 2.0.33 and building 2.2.7 gives me a kernel
>
> New config options have appeared, old ones have disappeared or have changed
> names, ...
> I don't know (and haven't look for) a tool that can upgrade your .config
> file, so what you have to do is
>
> cp old.config .config
> gmake menuconfig
> and then go through the whole config steps and fix the few problems.
Okay, thank you!
> > that will not boot and instead loops through getting the errno 111
> > from RPC.
> That might not be a problem with your .config but with something else.
> A good habit is to give a lot more info when posting such a question
> (like in what part of the boot it started giving you those errors).
Will do.
(was unable to do so in this case becausethe messages about
"errno 111" were lost in the quick scrolling).
--
Omri Schwarz ---
Timeless wisdom of biomedical engineering:
"Noise is principally due to the presence of the
patient." -- R.F. Farr
------------------------------
Reply-To: "Curt" <[EMAIL PROTECTED]>
From: "Curt" <[EMAIL PROTECTED]>
Subject: Re: SOCKS5
Date: Fri, 4 Jun 1999 19:17:56 -0500
try downloading the source from http://www.nec.socks.com
and compiling it on your system
Stephen Hicks <[EMAIL PROTECTED]> wrote in message
news:EbF53.244$[EMAIL PROTECTED]...
> I'm running RH6 (w/ kernel 2.2.9) on a Pentium 133 w/ 16 megs ram. I've
got
> ip_masq running but i need a SOCKS5 proxy as well - i downloaded the
> socks5-server-1.0r4-1.rpm, installed it, and copied the
> /usr/doc/socks5-server-1.0r4-1/socks5.conf-single.homed (or whatever the
> file was called) to /etc/socks5.conf and set
> SOCKS5_CONFFILE=/etc/socks5.conf. The problem is that when i run socks5,
i
> get a segmentation fault every time. I tried reinstalling and
redownloading
> it but it still errors. Can anyone help?
>
> --
> Stephen Hicks
> mailto:[EMAIL PROTECTED]
> http://home.tampabay.rr.com/kupopo/
> icq: 5453914; aim: kupopo1
>
>
------------------------------
From: [EMAIL PROTECTED] (Radovan Brako)
Subject: Re: Route
Date: 4 Jun 1999 16:35:58 +0200
In <[EMAIL PROTECTED]> valner <[EMAIL PROTECTED]> writes:
> My dialin is working.. I can listen it dialing, I can see
>the sislog messages and everything is ok... the final of the file is :
>May 31 00:09:07 valner pppd[156]: Using interface ppp0
>>May 31 00:09:07 valner pppd[156]: Connect: ppp0 <--> /dev/ttyS2
>
>however, I'm not connected... and the final words from the ISP didn't
>come... so I've take a look in the route table and I just see the eth0
>interface...
>
> I tryed to type the command line:
>
> route add default ppp0
>
>and I got the error message
>
>SIOCADDRT: No such device
>ppp0: unknown interface
>
> Well... I don't have idea how to setup that... (ifconfig??)
>
> any suggestion is welcome.
Maybe the PPP negotiation was not completed and you did not get
an IP assigned, maybe something else ... What does the output of
"ifconfig" say ? Of "route -n" ? If you add the "debug" option
for pppd, what does the log say ? (Maybe in /var/log/ppp.log, or
elsewhere, may depend on the distribution.)
RB
> Thankx
>Valner
------------------------------
From: "Robert E.Wijnberg" <[EMAIL PROTECTED]>
Subject: Re: Linux: ICMP Redirect, IP Source Routing unterdruecken
Date: Fri, 04 Jun 1999 17:47:36 +0200
Lars Marowsky-Bree wrote:
>
> On 03 Jun 1999 23:03:00 +0200, Detlef Bosau
> <[EMAIL PROTECTED]> wrote:
>
> > Ist jetzt ziemlich off topic. Aber ich meine schon seit langem,
> > dass die Alternative zu einem gerouteten, paketierten, Datenverkehr
> > in einem Datenverkehr besteht, in dem Peers, die verbindungsorientierten
> > Datenverkehr moechten (und das ist IMHO doch der haeufigere Fall)
> > direkt ueber eine virtuelle Verbindung miteinander kommunizieren.
>
Hi Guys,
Please can you move this discussion to a German speaking news group ?
--
\\\///
/ _ _ \
(| (.)(.) |)
=========================.OOOo==()==oOOO.=========================
pub 768/FE7D0209 1998/06/26 Robert E.Wijnberg <[EMAIL PROTECTED]>
Key fingerprint = AA 01 89 69 C0 D1 54 1A EB 36 45 73 A3 12 F4 9A
==================================================================
------------------------------
From: Tony <[EMAIL PROTECTED]>
Subject: base64 - how do I decode?
Date: Fri, 4 Jun 1999 15:44:10 GMT
Dear Linuxers,
I have received some base64 files. I can't find anything apropos base64.
Ho do I decode these?
Thank you all,
Tony
------------------------------
From: Omri Schwarz <[EMAIL PROTECTED]>
Subject: rshd, rlogind, maybe other daemons, in 2.2.7
Date: Fri, 04 Jun 1999 21:01:31 -0400
Hi,
I have two Pentium machines running 2.2.7, with each a DEC Tulip
NIC, and doing from one machine
strace -o /tmp/st -r -f rsh node3 uname -a
shows
461 0.000218 socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 4
461 0.000108 bind(4, {sin_family=AF_INET, sin_port=htons(1023),
sin_addr=inet_addr("0.0.0.0")}, 16) = 0
461 0.000178 fcntl(4, F_SETOWN, 461) = 0
461 0.000085 connect(4, {sin_family=AF_INET, sin_port=htons(514),
sin_addr=inet_addr("192.168.0.3")}, 16) = 0
461 0.000636 socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 5
461 0.000102 bind(5, {sin_family=AF_INET, sin_port=htons(1022),
sin_addr=inet_addr("0.0.0.0")}, 16) = 0
461 0.000118 listen(5, 1) = 0
461 0.000119 write(4, "1022\0", 5) = 5
461 0.000141 select(6, [4 5], NULL, NULL, NULL) = 1 (in [5])
461 54.010760 accept(5, {sin_family=AF_INET, sin_port=htons(1023),
sin_addr=inet_addr("192.168.0.3")}, [16]) = 6
which I assume is the return of STERR's data to the first machine.
I am consistently getting 54 seconds as the duration of that
call.
Has anyone else seen this?
--
Omri Schwarz ---
Timeless wisdom of biomedical engineering:
"Noise is principally due to the presence of the
patient." -- R.F. Farr
------------------------------
From: Jonas Sextl <[EMAIL PROTECTED]>
Crossposted-To: comp.security.firewalls
Subject: Re: Redhat 6.0 & ipmasqerade problems
Date: Fri, 04 Jun 1999 06:55:22 +0000
Hi!
Did you enable ip forwarding?
cat /proc/sys/net/ipv4/ip_forward
should give 0 or 1.
And with the command
echo "1" > /proc/sys/net/ipv4/ip_forward
you can enable forwarding.
cu,
Jonny
------------------------------
From: "Matt Graham" <[EMAIL PROTECTED]>
Subject: Re: 3c575 pcmcia problem on IBM laptop
Date: Fri, 4 Jun 1999 11:57:45 -0400
I had a similar problem. Switched to a Xircom 10/100 adapter and the problem
went away. Seems like a driver issue.
Henrik Malmgren wrote in message <01beae0a$160b5520$01aba8c0@lords>...
>I have a problem with a 3c575 pcmcia card on a IBM thinkpad 600 laptop. I
>run RedHat 6.0 and i have updated the pcmcia modules to 3.0.11, and still
>no go. The card is probing att boot, but when i try to ping, nothing
>happens for several minutes, then it pops up a line "*** irq in use by
>other device maybe", then it takes a few more minutes, and the pinging is
>started. But with 40-50% packet loss. I have tried IRQ 3, 4, 7, 9 and 10,
>but it's still the same. What files should i edit? and what should i type
>into them?
>
>--
>/Henrik!
>
> "We've all heard that a million monkeys banging on a million
> typewriters will eventually reproduce the entire works of
> Shakespeare. Now, thanks to the Internet, we know this is
> not true." -- Robert Wilensky, ILP 1996
>
>
>
>
------------------------------
From: "Marcel v.d. Steeg" <[EMAIL PROTECTED]>
Crossposted-To: at.linux,comp.os.linux.admin
Subject: REPLICATION between 3 servers ....
Date: Fri, 4 Jun 1999 17:42:01 +0200
Hi there,
I've got a question about replication.
Environment:
Linux Redhat 5.2
Question :
Does anybody know how you can replicate a file between 3 servers ???
The file is on all thee servers in the same directory, the only thing that
has to be done is that it's going to be overwritten and that there is
executed a file ....
Is there a program for it ? Or maybe a script somewhere on the WWW ???
Please help me !
Thanx !
Greetz Marcel
--
=====================================================
E-mail: [EMAIL PROTECTED]
------------------------------
From: bill goodacre <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: 3c905b problem with 2.0.36
Date: Fri, 04 Jun 1999 17:42:33 -0700
I need advice.
I had a working RH 4.2 with a NE2000 ISA card configured at irq 10
address 0x300. I then replaced it with a 3com 3c905b 100 base Tx PCI
card. The system would not find the new card. I upgraded to RH 5.2 at
this point. The new os won't find the card either. I have the most
recent (just a few days old) Ethernet HOW-TO and have read it. I
downloaded what I hoped would be the fix from Mr. Becker's NASA site. I
compiled the program using the suggested compiler options. The message I
get from insmod is "value for symbol io not found". This is the same
message I get in the boot sequence. Coincidentally, I was installing
Sybase on this machine (which went fine BTW) and it did a scan of my
hardware and told me I had a 3c905b configured at IRQ 15 and 0x6500.
I've tried passing parameters to the kernel at boot time using the LILO
prompt, to wit "linux ether=" and so on, giving IRQ15 and 0x6500. no go.
Now I've downloaded a two disk set from 3com's ftp site which are said
to be DOS configuration utilities. I would like to boot from floppy into
a DOS environment, then use the configuration utility to set the card at
IRQ10 and 0x300. Then I hope the boot code will find the damn thing and
give me a working network link. I now find I don't have any actual DOS
system left, just Win 95. Well I don't see what else to do so tonight I
will boot w95 from floppy then try to run the configuration utilities.
If anyone wants to take pity on me I will be much obliged.
Also I think 3com is extremely perverse in naming their cards 3c509 and
3c905 and then introducing occasional compatability. As if I didn't have
enought to think about without the nagging fear that I've made some
stupid dyslexic error.
Please reply to the address below. Thanks in advance.
Dave Stevens
[EMAIL PROTECTED]
------------------------------
From: "BTRiggs" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.install
Subject: 3C509 NIC and RH 5.2
Date: Fri, 4 Jun 1999 12:08:57 -0600
I am running a 486 DX P120 with a 3Comm 3c509B NIC. When I set up RH 5.0 it
works fine, if I upgrade to 5.2 or 6.0 it can no longer find the card.
During the install or upgrade process I try to load the 3C509 drivers and it
comes back with "Unable to locate device". During the boot script it reads:
eth0: delaying initialization
eth0...unknown interface
I was running on a split partition, Windoz and Linux. Now it is free from of
any hazards. But I still can't get it to see the card in anything other than
5.0. I also reset it so it is no longer using DHCP. Any assistance is
greatly appreciated.
Thanks,
mail to : [EMAIL PROTECTED]
Brian Riggs
------------------------------
From: Bob Dusek <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Restrict Access via NIC address?
Date: Fri, 04 Jun 1999 10:39:14 -0500
Hello,
We're playing around with a few NFS ideas, here, and one thing that
we're stumped on is: is there any way to restrict access (via
hosts.allow or some other means) to services based upon hardwared
address, rather than IP?
Thanks,
Bob
--
Saint Joseph's College -- [EMAIL PROTECTED] | http://www.saintjoe.edu
=======================================================================
Always the dullness of the fool is the whetstone of the wits.
-- William Shakespeare, "As You Like It"
------------------------------
From: Bob Dusek <[EMAIL PROTECTED]>
Subject: Re: Forwarding of broadcasts possible?
Date: Fri, 04 Jun 1999 10:34:24 -0500
I may be showing my ignorance, here... but, I don't mind:
Couldn't you also set the subnets up with a different netmask (i.e.
255.255.253.0)
Now that I think about it, I'm almost certainly showing my ignorance by
saying that, but I wouldn't mind a short explanation, if someone's
willing to give one.
Bob
------------------------------
** 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
******************************
