Linux-Networking Digest #690, Volume #11 Sun, 27 Jun 99 10:13:22 EDT
Contents:
Re: Setting Linux up as a proxy server ([EMAIL PROTECTED])
SuSE 6.1 Dialup (waco)
Re: MAKE GOOD MONEY $$$$$$$$$$$$$$$$ ("Frederik Meerwaldt")
Re: Trouble networking two linux machines (Steve Cowles)
Re: Why not C++ (Nathan Myers)
Re: Why not C++ (Nathan Myers)
Re: Why not C++ (Nathan Myers)
Re: Why not C++ (Chris Double)
Re: Why not C++ (Bruce Hoult)
Emulating a (telephone) modem via soundcard (Szomraky Stefan)
Re: downloads slow with 2.2 kernel ("Gene Heskett")
Need help sending files win to unix (Ted)
Re: Dial-up problems (Clifford Kite)
Re: Cluster management - software?? (Tobias Anderberg)
Re: tcp_timestamps causing massive network slowdown problems under 2.2.x (John
Gardner)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: redhat.networking.general
Subject: Re: Setting Linux up as a proxy server
Date: Sun, 27 Jun 1999 11:07:16 GMT
In article <[EMAIL PROTECTED]>,
Bob G <[EMAIL PROTECTED]> wrote:
> [EMAIL PROTECTED] wrote:
>
> > I am wanting to set up Redhat 6.0 as a proxy server to serve
cross
> > platform machines (Win9x, and WinNT) on my private network. I want
all the
> > machines to be able to use the cable modem which is hooked up to the
linux
> > box. I also want to be able to use everything on the net and not
just http
> > and ftp.
> >[...]
>
> You can find these on your local system if you installed the
> documentation (look under /usr/doc/HOWTO):
>
> IPCHAINS-HOWTO
> Firewall-HOWTO
>
> Check out Samba if you need to share files and printers using Linux as
> an NT-like server.
>
> - Bob
>
Also check out the IP masquerading howto.. (real effective tool, but
use carefully)
praveen
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: waco <[EMAIL PROTECTED]>
Subject: SuSE 6.1 Dialup
Date: Sun, 27 Jun 1999 07:30:08 -0400
I was wondering if it is possible to use WVDIAL instead of the usual
diald/pppd/script combo to automatically dial the internet?
>From SuSE 5.3 to 6.1, they dropped the old style combo and opted for
WVDIAL, although all the old utilities are still there.
If I DIR the /ETC directory, the WVDIAL.CONF is owned by Dialout.
SuSE 6.1 using 2.2.7 kernel....
Any help would be appreciated...
J,
------------------------------
From: "Frederik Meerwaldt" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.networking.ras,comp.os.ms-windows.networking.win95,comp.os.ms-windows.nt.admin.networking
Subject: Re: MAKE GOOD MONEY $$$$$$$$$$$$$$$$
Date: Sat, 26 Jun 1999 16:51:47 +0200
Does he work at MCdonalds???
Regards,
Freddy
Rick M <[EMAIL PROTECTED]> wrote in message
news:7l2llh$4j$[EMAIL PROTECTED]...
> In article <[EMAIL PROTECTED]>,
> "debee99" <[EMAIL PROTECTED]> wrote:
> > To make money when you are on the internet. You have to visit
> > www.gotoworld.com. When you visit the site you will need to get a pass
> > wordand user name. But when you get that dun at the same time you
> should to
> > enter 1006559531. That goes where it says << Friend's Member ID. When
> you do
> > that every one that use that # gets paid more. And the same for you.
> When I
> > use it you get payed more for useing that #.
> > THANK YOU.
>
> No thank you! If it wasn't for you, I would have had no reason to send
> this post to the gotoworld.com postmaster. The only $$ you'll be
> collecting is from your McDonald's job.
>
> Got fries?
>
> Luser.
>
>
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.
------------------------------
From: Steve Cowles <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Trouble networking two linux machines
Date: Sun, 27 Jun 1999 11:37:52 GMT
bill wrote:
> Hi,
>
> I'm having trouble networking two linux machines. My goal is to setup
> up IP
> Masq so that both computers can use the internet at the same time (my
> bro and
> I have have a tendency to fight for internet time). However, I've
> haven't
> been able to get the two computers to even talk to each other yet.
> Maybe
> someone could point out what I'm doing wrong or what I have neglected
> to do.
> The first of these is running Debian 2.0/2.0.34 with an ISA 3Com
> Etherlink
> card (3c509) and it's configuration is as follows:
>
> NIC card initialization--
> eth0: 3c509 at 0x300 tag 1, 10baseT port, address 00 a0 24 03 b6 ea,
> IRQ 10.
> 3c509.c:1.16 2/3/98 [EMAIL PROTECTED]
> eth0: Setting Rx mode to 1 addresses.
>
> Setting up IP Address--
> eth0 Link encap:Ethernet HWaddr 00:A0:24:03:B6:EA
> inet addr:192.168.1.2 Bcast:192.168.1.255
> Mask:255.255.255.0
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:0 errors:0 dropped:0 overruns:0 frame:0
> TX packets:105 errors:0 dropped:0 overruns:0 carrier:105
> Collisions:0
> Interrupt:10 Base address:0x300
>
> And the routing table--
> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref
> Use Iface
> 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0
> 16 eth0
> 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0
> 14 lo
>
> And for the other computer (aka zeus) I'm running Debian 2.1 with a
> custom
> built 2.2.10 kernel and a PCI 3Com Etherlink card (3c905b).
> Everything looks
> pretty much the same here, too.
>
> eth0: 3Com 3c905B Cyclone 100baseTx at 0xfc00, 00:10:5a:0f:24:ca, IRQ
> 9
> 8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate
> interface.
> MII transceiver found at address 24, status 7849.
> MII transceiver found at address 0, status 7849.
> Enabling bus-master transmits and whole-frame receives.
>
> eth0 Link encap:Ethernet HWaddr 00:10:5A:0F:24:CA
> inet addr:192.168.1.1 Bcast:192.168.1.255
> Mask:255.255.255.0
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:0 errors:0 dropped:0 overruns:0 frame:0
> TX packets:12 errors:0 dropped:0 overruns:0 carrier:12
> Collisions:0
> Interrupt:9 Base address:0xfc00
>
> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref
> Use Iface
> 192.168.1.0 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
>
>
> When I try to ping each of the computers I don't recieve any response.
> One
> thing I did notice I thought was strange was that on each of the
> computers
> arp -a reported--
>
> hermes (192.168.1.2) at <incomplete> on eth0 on the Deb 2.1 machine
> and
> zeus (192.168.1.1) at <incomplete> on eth0 on the 2.0 machine.
>
> From what I understand the ip address on ethernets server as aliases
> for
> the hardware addresses on each of the cards and I thought that maybe
> this
> was my problem.
>
> So, I manually constructed the table for both machines. Then arp -a
> on
> both machines returned completed tables--
> hermes (192.168.1.2) at 00:A0:24:03:B6:EA [ether] PERM on eth0
> zeus (192.168.1.1) at 00:10:5A:0F:24:CA [ether] PERM on eth0
> But, this didn't work either and pinging both computers still failed.
>
> Could I possible be lacking some support in either of my kernels? I'm
> not
> real sure what is supported in the installed kernel of the Debian 2.0
> system
> but with the 2.2.10 kernel that I built I have the following set --
>
> CONFIG_PACKET=y
> CONFIG_NETLINK=y
> CONFIG_RTNETLINK=y
> CONFIG_NETLINK_DEV=y
> CONFIG_UNIX=y
> CONFIG_INET=y
> CONFIG_IP_ROUTER=y
> CONFIG_SYN_COOKIES=y
>
> Or maybe all of this is correct and I'm making some mundane mistake.
>
> Thanks, Bill.
>
> --
> It was one of those perfect summer days -- the sun was shining, a breeze was
>blowing, the birds were singing, and the lawn mower was broken ... --- James Dent
>
>
Bill,
You left out one important piece of information in your post. How the
two computers are "physically" hooked up? i.e.. through a hub/switch or
a cross-over cable between nic's. From what I can tell, it looks like
your problem is NOT related to your protocol (IP) configuration, but at
your networks "physical layer".
The first thing I noticed (from your post) was the 509 is a 10baseT card
and the 905 is a 100baseT card. Also both cards appear to be
transmitting data, just not receiving anything at the other end. (duh!).
Obviously, both cards cannot talk to each other without a hub/switch
that is capable of dual speeds (10/100). If you do not have a hub/switch
such as this, you will need to configure (fix) the 905 to operate in
10BaseT mode through the 3com setup utility. If your trying to hookup
your two Linux boxes using a single cable, 1) make sure it is a
"cross-over" cable (transmit/recieve reversed) and 2) that the 905 card
is set to 10BaseT mode.
Steve Cowles
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 27 Jun 1999 04:42:31 -0700
Linus Torvalds <[EMAIL PROTECTED]> wrote:
>Bruce Hoult <[EMAIL PROTECTED]> wrote:
>
>But what others are saying is "a lot of existing C++ compilers generate
>worse code than a lot of existing C compilers". And they are right too.
>
>Sometimes theory matters. Sometimes it doesn't. The world is not as
>simple as you make it out to be.
Enough generalities. Take for example Egcs. C and C++, same code
generator, same optimizer. The last time you tried g++ was years
and years ago. It's time to look again.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 27 Jun 1999 04:47:46 -0700
Andi Kleen <[EMAIL PROTECTED]> wrote:
>My point was simply:
>In my first tries to use the STL, I had to fight with 10+ lines
>really obscure template error messages. I wouldn't wish that to
>any poor beginner (given that beginning is usually mostly syntax
>error fixing at first)
True, there is a great deal of work yet to be done on improving
error reporting in Egcs, and similar work in Gdb.
OTOH, once you get code past the compiler, it's far less buggy.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 27 Jun 1999 04:18:29 -0700
Paul Jackson <[EMAIL PROTECTED]> wrote:
>Nathan Myers wrote:
>|> [Those saying] it's slower than (e.g.) C are just spreading FUD.
>
>Put it this way -- if I could actually try ten
>different teams of professional but not superstar
>programmers, using different languages to implement
>a real world (multi-year, requirements changing
>for the life of the project, ...) app, right
>now I'd bet that Python team (a delightful language
>that I just learning) would produce the fastest
>working app (because they'd have time to get the
>architecture and algorithms 'right'), and the C++
>team would produce one of the slower apps, with
>a few key loops going fast, but bogged down in a
>slew of other confusions.
Odd choice of example; Python is written in C++.
I will assert that my team using Python with C++ modules would
produce the fastest result, soonest.
See http://www.cantrip.org/realworld.html .
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: Chris Double <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: 28 Jun 1999 00:10:06 +1200
[EMAIL PROTECTED] (Bruce Hoult) writes:
> > In practice, C++ is a much more complex language. As such, C++
> > compilers are much more complex beasts. And that is why some C++
> > compilers do not generate as good code as the equivalent C compilers do.
>
> I'd love to see examples.
[...snip...]
>
> I'm afraid I basically don't use any other compilers. Which C++ compilers
> are bad?
In a recent Windows Developer Journal (June 1999) they printed an
article on a Microsoft Visual C++ 6.0 bug. Some code in C++ mode
caused incorrect assembler to be generated resulting in an infinite
loop. The exact same code in C mode compiled and executed as expected.
Chris.
------------------------------
From: [EMAIL PROTECTED] (Bruce Hoult)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: Re: Why not C++
Date: Mon, 28 Jun 1999 00:32:47 +1200
In article <[EMAIL PROTECTED]>, Chris Double <[EMAIL PROTECTED]> wrote:
> [EMAIL PROTECTED] (Bruce Hoult) writes:
>
> > > In practice, C++ is a much more complex language. As such, C++
> > > compilers are much more complex beasts. And that is why some C++
> > > compilers do not generate as good code as the equivalent C compilers do.
> >
> > I'd love to see examples.
> [...snip...]
> >
> > I'm afraid I basically don't use any other compilers. Which C++ compilers
> > are bad?
>
> In a recent Windows Developer Journal (June 1999) they printed an
> article on a Microsoft Visual C++ 6.0 bug. Some code in C++ mode
> caused incorrect assembler to be generated resulting in an infinite
> loop. The exact same code in C mode compiled and executed as expected.
>
> Chris.
Heh. I could make a really cheap shot there... ;-)
Bad C++ compiler noted, thanks.
-- Bruce
------------------------------
From: Szomraky Stefan <[EMAIL PROTECTED]>
Crossposted-To:
telekabel.linux,at.linux,inode.linux,com.os.linux.misc,ak.os.linux,ahn.tech.linux,alt.linux,alt.os.linux,alt.uu.comp.os.linux.questions,alt.uu.comp.os.linux,be.comp.os.linux,ch.comp.os.linux,chello.linux,cmu.comp.os.linux,comp.os.linux.help,comp.os.linuxmisc,es.comp.os.linux,fido.ger.linux,flensburg.linux,fi.os.linux.networking,fi.os.linux,ger.pc.linux
Subject: Emulating a (telephone) modem via soundcard
Date: Sun, 27 Jun 1999 12:28:26 GMT
Hi!
Sorry for the repost and for the crosspost, but nobody could
helped me, so i=B4ve to ask as many people I can ask....
Is it possible with Linux to emulate a "normal" telephone modem via
the soundcard?
Please help me! I need it realy!
Thank you
PS: please contact me via e-mail, or post the answer in telekabel.linux =
thank you....
------------------------------
Date: 27 Jun 99 09:04:25 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Re: downloads slow with 2.2 kernel
Unrot13 this;
Reply to: <[EMAIL PROTECTED]>
Gene Heskett sends Greetings to greg rudd;
gr> Problem is with ppp performance on RedHat 6.0 system with 2.2.5
gr> kernel compared to performance on RedHat 5.2 system with 2.0.36
gr> kernel. Problem is downloading of large (200 to 400K) JPEG and
gr> GIF files on RedHat 6.0 system causes Netscape to continually
gr> stall and stop receiving data many times during the download. If
gr> you leave Netscape along, eventually the link will begin sending
gr> data again, and the status message will change from stalled to
gr> the amount of data coming over the connection. Do not have this
gr> problem with RedHat 5.2 with 2.0.36 kernel. Using same ppp
gr> scripts and serial port setup on both systems. It seems to me
gr> that the 2.2 series kernels may be having a flow control or some
gr> other problem with large downloads. Has anyone else experienced
gr> similar problems with the 2.2 series kernels?
So am I, same Zoom modem via switchbox is used on this machine, and a
linux box running 2.2.10 and netscape 4.61 us/ca ssl version.
Here, with a 56k connection at 50k, I get 2k-3k speeds total, even with
several parallel conections.
On the linux box, I get 700 cps per connection until I've used up the
bandwidth somewhere around the 5000 cps mark. So it appears to be a
per connection limit.
I've no idea where its at. The linux box is a 400mhz AMD K6-2, 128
megs pc100 memory, DMA is working on all drives, and it *should* be able
to handle 5,000,000 cps without even showing up in a top report.
I did play with the modem init a bit last night, and that seems to have
reduced the re-negotiations, but hasn't helped the speed all that much.
But we need answers, thats why we lurk, and occasionally respond.
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: Ted <[EMAIL PROTECTED]>
Subject: Need help sending files win to unix
Date: Sun, 27 Jun 1999 13:18:51 GMT
Our office has just converted from win3.1 running lantastic to
windows98. The problem is one of the machines sent files to a unix
machine using lantactic tcp/ip.
The batch file to send the files is RCP %1 PLT:/export/home/jobs.
rcp.exe is in the lantastic tcp directory
The lmhost file reads:
89.0.10.20 plot_1 PLOT_1 PLT plt fire1 FIRE1
89.0.10.40 can CAM imark1 IMARK1
Any help on how to do this with win98 would be greatly appreciated.
Thanks
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: Dial-up problems
Date: 27 Jun 1999 08:15:01 -0500
John Tomlinson ([EMAIL PROTECTED]) wrote:
: 1. Minicom is very, very slow.
: 2. I only get as far as sending my login and password then the connection is
: severed.
Sounds like the wrong IRQ might be configured for the device file.
Doing "setserial /dev/ttySx", x=whatever for your modem, will show which
IRQ is configured. The configured IRQ must match the IRQ that the modem
actually uses. The device file is configured by setserial in one of
the boot-up files.
If, after typing AT in minicom terminal mode, it takes on the order
of 19 seconds to get the OK response back from the modem then it is a
misconfigured IRQ.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* My confidence in this answer (X), on a scale of 1 to 10:
|----|----|----|----|----|----|----|----|----|----X
0----1----2----3----4----5----6----7----8----9----10 */
------------------------------
From: [EMAIL PROTECTED] (Tobias Anderberg)
Subject: Re: Cluster management - software??
Reply-To: [EMAIL PROTECTED]
Date: Sun, 27 Jun 1999 15:53:59 +0200
>> >Does anyone know of centralized cluster management software for
>> >a group of networked Linux boxes? We have a number of machines that I
>> >would like to administer from a single centralized point. I've not seen
>> >much in the way of Linux software for this purpose. Any ideas?
>>
>> www.beowulf.org
><rant>Ignore this idiot. Beowulf is a multiple processing cluster, not a
>especially to just put the URL with no explanation...thanks tobias for that
>revelation</rant>
Ouch, harsh words from an obvious expert. Actually I only put that address
there as a reminder to myself to check it out if that was what the orginal
poster wanted and then come back to give more information. Obviously I forgot
to come back to article thus sending off my reminder. Then I though, hey, let
him check it out.. :-) (ie too lazy to send a cancel)
/tobias - notice the irony in the first sentence?
------------------------------
From: John Gardner <[EMAIL PROTECTED]>
Subject: Re: tcp_timestamps causing massive network slowdown problems under 2.2.x
Date: Sun, 27 Jun 1999 13:07:40 +0100
In article <[EMAIL PROTECTED]>, Gary Elmes
<[EMAIL PROTECTED]> writes
>I have just had a couple of fun days trying to work out why my http
>downloads slow to a crawl from certain sites.
>
>I solved the problem by setting /proc/sys/net/ipv4/tcp_timestamps = 0.
>Is this a known problem?
>
>I found that, with timestamps on, packets from problematic sites
>(www.isc.org was the one I tripped over) were showing up as fragmented
>by tcpdump, even though I have the kernel compiled to reassemble
>fragmented packets. In addition, the remote http host took yonks
>(minutes at a time) to respond.
>
>Typically, a 1460-byte packet would be fragmented into 1448 and 12.
>
>Disabling timestamps solved the problem - no fragmenting, no slowdown,
>nice whizzy downloads. I now have:
>
> cat 0 > /proc/sys/net/ipv4/tcp_timestamps
>
>in my rc.local, and everything is fine.
>
>I reproduced the problem under both 2.2.1 and 2.2.10, and with all
>manner of network-related bits both compiled in and left out of the
>kernel (all in an effort to diagnose the problem). I reproduced it with
>both lynx and communicator, as well. My connection to the internet is
>via dial-up ppp; and stuffing with all the available ppp options
>(compression, mru, mtu, etc.) did nothing to solve the problem. Other
>machines - both win95 and linux 2.0.36 - using this box as a gateway to
>the internet did not have the problem (timestamps were not set on their
>packets).
>
>Should somebody be fixing something in the timestamping code, or is my
>"fix" just working around the real problem?
I came across a similar problem a few days ago. When using telnet or
rlogin from Linux and leaving the connection for five minutes without
any traffic on the line, a lockup occurs from the application point of
view.
Examining the trace shows TCP timestamps being used. However, Linux
fails to update the sequence number after the next character is typed
and the remote stack keeps sending the same data. Sometimes Linux
recovers, sometimes the connection is lost.
If timestamps support is removed (in the same way that you did), the
stack works correctly.
The stack employed by the OS I'm responsible for, is developed from
Mentat's streams implementation. Solaris also uses Mentat TCP and this
stack is (presumably) well tried in the field.
In your case, is there retransmission of data, or is it purely a
fragmentation problem?
Certainly, turning off timestamps is only an avoidance.
Regards,
--
John Gardner
------------------------------
** 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
******************************
