[LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread sAwAr
Hi,

my company have just bought new network and I have question about one problem. 
As in topic we must use few completely different dsl's and balance traffic 
between them. 
2M/0,5M 4Mb/0,5M 8M/0,5M
M=Mb/s
I've never done such thing before so I have doubts how it will work. If the 
links are symmetric 2/2 4/4 8/8 there is no problem because with weights I can 
compensate  the difference between them and achieve nice results. But what in 
my situation?
My questions are: how to set load balancing to get all links equally loaded and 
avoid situation when the up load will be full and download almost empty? I 
believe this situation can happen due to fact that load balancing is based on 
flows and for example p2p or smpt/pop3 will eat whole upload. 
If my problem isn't clear I'll try to explain it better later. 


Thanks in advance.
Pozdrawiam
sawar

--
Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8

___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread Jordi Segues

Hello,

I've done this some montsh ago, with a command like:
ip route add default equalize scope global nexthop  via $EXTGW1 dev
$EXTIF1 weight 1 nexthop via $EXTGW2 dev $EXTIF2 weight 1

However, this is not the problem.
While loadbalancing of simple requests worked fine, there where
problems when you worked with connections. I mean HTTPS, of FTP
connection for example.

The problem was fo me that the system trys to send packets of the same
connection throught different gateways, so with different IP source
(each DSL connection was from different ISP). This caused the server
not to understand why the same connection sent packets with 2
different source IP ;)
Well, I hope you understand me.

If you would do real load balancing, and in a proper way, you should
not only do it by link charge, but route packets by connection to.
(routing all packets of the same connection through the same gateway)
This is caused because you must flush the route cache some times (or
packets to a destination will allways take the same route, wich is not
a loadbalance).

So if someone has done it and doesn't have this problem, I'm interested too :)

Thanks!

Jordi Segues

On 22 Jan 2007 09:49:28 +0100, sAwAr [EMAIL PROTECTED] wrote:

Hi,

my company have just bought new network and I have question about one problem.
As in topic we must use few completely different dsl's and balance traffic 
between them.
2M/0,5M 4Mb/0,5M 8M/0,5M
M=Mb/s
I've never done such thing before so I have doubts how it will work. If the 
links are symmetric 2/2 4/4 8/8 there is no problem because with weights I can 
compensate  the difference between them and achieve nice results. But what in 
my situation?
My questions are: how to set load balancing to get all links equally loaded and 
avoid situation when the up load will be full and download almost empty? I 
believe this situation can happen due to fact that load balancing is based on 
flows and for example p2p or smpt/pop3 will eat whole upload.
If my problem isn't clear I'll try to explain it better later.


Thanks in advance.
Pozdrawiam
sawar

--
Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8

___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc




--
Jordi Segués Daina
---
Andorra GSM: (+376) 35 35 68
France GSM: (+33) (0)6 81 88 35 55
[EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
AIM: superjordix
Skype: callto://superjordix
---
http://www.JordiX.com
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread Alex Samad
On Mon, Jan 22, 2007 at 10:03:21AM +0100, Jordi Segues wrote:
 Hello,
 
 I've done this some montsh ago, with a command like:
 ip route add default equalize scope global nexthop  via $EXTGW1 dev
 $EXTIF1 weight 1 nexthop via $EXTGW2 dev $EXTIF2 weight 1
 
 However, this is not the problem.
 While loadbalancing of simple requests worked fine, there where
 problems when you worked with connections. I mean HTTPS, of FTP
 connection for example.
 
 The problem was fo me that the system trys to send packets of the same
 connection throught different gateways, so with different IP source
 (each DSL connection was from different ISP). This caused the server
 not to understand why the same connection sent packets with 2
 different source IP ;)
 Well, I hope you understand me.
 
 If you would do real load balancing, and in a proper way, you should
 not only do it by link charge, but route packets by connection to.
 (routing all packets of the same connection through the same gateway)
 This is caused because you must flush the route cache some times (or
 packets to a destination will allways take the same route, wich is not
 a loadbalance).
 
 So if someone has done it and doesn't have this problem, I'm interested too 
 :)

the above is actually covered in the wiki howto.  Bu tyou need to setup snat on
each interface, then connection tracking takes care of sending each stream out
the right interface, you need to use snat and not MASQ.

Then you need to setup up some ip rule tables for each of the interfaces.


my ip ru looks like this

0:  from all lookup local 
200:from 144.132.145.38 lookup cable 
201:from 60.241.248.86 lookup adsl 
32766:  from all lookup main 
32767:  from all lookup default 


my ip r sh tab default 

default  proto static  metric 5 
nexthop via 144.132.144.1  dev vlan2 weight 1
nexthop via 10.20.20.230  dev ppp0 weight 20
default via 10.20.20.230 dev ppp0  src 60.241.248.86  metric 20 
default via 144.132.144.1 dev vlan2  src 144.132.145.38  metric 30 


This works fine for me, I have tracked packets with tcpdump on both the server
and the client.

Alex



 
 Thanks!
 
 Jordi Segues
 
 On 22 Jan 2007 09:49:28 +0100, sAwAr [EMAIL PROTECTED] wrote:
 Hi,
 
 my company have just bought new network and I have question about one 
 problem.
 As in topic we must use few completely different dsl's and balance traffic 
 between them.
 2M/0,5M 4Mb/0,5M 8M/0,5M
 M=Mb/s
 I've never done such thing before so I have doubts how it will work. If 
 the links are symmetric 2/2 4/4 8/8 there is no problem because with 
 weights I can compensate  the difference between them and achieve nice 
 results. But what in my situation?
 My questions are: how to set load balancing to get all links equally 
 loaded and avoid situation when the up load will be full and download 
 almost empty? I believe this situation can happen due to fact that load 
 balancing is based on flows and for example p2p or smpt/pop3 will eat 
 whole upload.
 If my problem isn't clear I'll try to explain it better later.
 
 
 Thanks in advance.
 Pozdrawiam
 sawar
 
 --
 Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8
 
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 
 
 
 -- 
 Jordi Segués Daina
 ---
 Andorra GSM: (+376) 35 35 68
 France GSM: (+33) (0)6 81 88 35 55
 [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
 AIM: superjordix
 Skype: callto://superjordix
 ---
 http://www.JordiX.com
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 


signature.asc
Description: Digital signature
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] routing patches seem to break output nat

2007-01-22 Thread Tim Haak

Hi

We have applied the routing patches from 
http://www.ssi.bg/%7Eja/#routes. To 2.6.15 this seems to have broken our 
output natting. Has anyone else experienced this or any advice on how to 
fix. Is this working on the newer kernel i.e. 2.6.19 ? Any help would be 
appreciated.


--
Tim Haak

email: [EMAIL PROTECTED]
cel:   0837787100

First love is only a little foolishness and a lot of curiosity, no really
self-respecting woman would take advantage of it.
-- George Bernard Shaw, John Bull's Other Island

begin:vcard
fn:Tim Haak
n:Haak;Tim
email;internet:[EMAIL PROTECTED]
tel;cell:+28 83 778 7100
version:2.1
end:vcard

___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread Jordi Segues

the above is actually covered in the wiki howto.  Bu tyou need to setup snat on
each interface, then connection tracking takes care of sending each stream out
the right interface, you need to use snat and not MASQ.


Great news :)
And thankyou for the details.
But could you give the link to the wiki howto?
I only found old doc.

Thanks!



Then you need to setup up some ip rule tables for each of the interfaces.


my ip ru looks like this

0:  from all lookup local
200:from 144.132.145.38 lookup cable
201:from 60.241.248.86 lookup adsl
32766:  from all lookup main
32767:  from all lookup default


my ip r sh tab default

default  proto static  metric 5
nexthop via 144.132.144.1  dev vlan2 weight 1
nexthop via 10.20.20.230  dev ppp0 weight 20
default via 10.20.20.230 dev ppp0  src 60.241.248.86  metric 20
default via 144.132.144.1 dev vlan2  src 144.132.145.38  metric 30


This works fine for me, I have tracked packets with tcpdump on both the server
and the client.

Alex




 Thanks!

 Jordi Segues

 On 22 Jan 2007 09:49:28 +0100, sAwAr [EMAIL PROTECTED] wrote:
 Hi,
 
 my company have just bought new network and I have question about one
 problem.
 As in topic we must use few completely different dsl's and balance traffic
 between them.
 2M/0,5M 4Mb/0,5M 8M/0,5M
 M=Mb/s
 I've never done such thing before so I have doubts how it will work. If
 the links are symmetric 2/2 4/4 8/8 there is no problem because with
 weights I can compensate  the difference between them and achieve nice
 results. But what in my situation?
 My questions are: how to set load balancing to get all links equally
 loaded and avoid situation when the up load will be full and download
 almost empty? I believe this situation can happen due to fact that load
 balancing is based on flows and for example p2p or smpt/pop3 will eat
 whole upload.
 If my problem isn't clear I'll try to explain it better later.
 
 
 Thanks in advance.
 Pozdrawiam
 sawar
 
 --
 Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8
 
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 


 --
 Jordi Segués Daina
 ---
 Andorra GSM: (+376) 35 35 68
 France GSM: (+33) (0)6 81 88 35 55
 [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
 AIM: superjordix
 Skype: callto://superjordix
 ---
 http://www.JordiX.com
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc



-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFFtJ1/kZz88chpJ2MRAhGKAJ9xthAZnQ/ovr82sa/x5j4BFJGgWwCgvtWa
dS7qseaia3GnZK/n8szE98Y=
=zLpL
-END PGP SIGNATURE-


___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc






--
Jordi Segués Daina
---
Andorra GSM: (+376) 35 35 68
France GSM: (+33) (0)6 81 88 35 55
[EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
AIM: superjordix
Skype: callto://superjordix
---
http://www.JordiX.com
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread Alex Samad
On Mon, Jan 22, 2007 at 01:21:32PM +0100, Jordi Segues wrote:
 the above is actually covered in the wiki howto.  Bu tyou need to setup 
 snat on
 each interface, then connection tracking takes care of sending each stream 
 out
 the right interface, you need to use snat and not MASQ.
 
 Great news :)
 And thankyou for the details.
 But could you give the link to the wiki howto?
 I only found old doc.
been a while since i had a look, quick google gave me this

http://lartc.org/howto/lartc.rpdb.multiple-links.html

I have this booked market as the wiki
http://linux-net.osdl.org/index.php/Main_Page

But I think the former is what you want

 
 Thanks!
 
 
 Then you need to setup up some ip rule tables for each of the interfaces.
 
 
 my ip ru looks like this
 
 0:  from all lookup local
 200:from 144.132.145.38 lookup cable
 201:from 60.241.248.86 lookup adsl
 32766:  from all lookup main
 32767:  from all lookup default
 
 
 my ip r sh tab default
 
 default  proto static  metric 5
 nexthop via 144.132.144.1  dev vlan2 weight 1
 nexthop via 10.20.20.230  dev ppp0 weight 20
 default via 10.20.20.230 dev ppp0  src 60.241.248.86  metric 20
 default via 144.132.144.1 dev vlan2  src 144.132.145.38  metric 30
 
 
 This works fine for me, I have tracked packets with tcpdump on both the 
 server
 and the client.
 
 Alex
 
 
 
 
  Thanks!
 
  Jordi Segues
 
  On 22 Jan 2007 09:49:28 +0100, sAwAr [EMAIL PROTECTED] wrote:
  Hi,
  
  my company have just bought new network and I have question about one
  problem.
  As in topic we must use few completely different dsl's and balance 
 traffic
  between them.
  2M/0,5M 4Mb/0,5M 8M/0,5M
  M=Mb/s
  I've never done such thing before so I have doubts how it will work. If
  the links are symmetric 2/2 4/4 8/8 there is no problem because with
  weights I can compensate  the difference between them and achieve nice
  results. But what in my situation?
  My questions are: how to set load balancing to get all links equally
  loaded and avoid situation when the up load will be full and download
  almost empty? I believe this situation can happen due to fact that load
  balancing is based on flows and for example p2p or smpt/pop3 will eat
  whole upload.
  If my problem isn't clear I'll try to explain it better later.
  
  
  Thanks in advance.
  Pozdrawiam
  sawar
  
  --
  Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8
  
  ___
  LARTC mailing list
  LARTC@mailman.ds9a.nl
  http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
  
 
 
  --
  Jordi Segués Daina
  ---
  Andorra GSM: (+376) 35 35 68
  France GSM: (+33) (0)6 81 88 35 55
  [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
  AIM: superjordix
  Skype: callto://superjordix
  ---
  http://www.JordiX.com
  ___
  LARTC mailing list
  LARTC@mailman.ds9a.nl
  http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 
 
 
 -BEGIN PGP SIGNATURE-
 Version: GnuPG v1.4.6 (GNU/Linux)
 
 iD8DBQFFtJ1/kZz88chpJ2MRAhGKAJ9xthAZnQ/ovr82sa/x5j4BFJGgWwCgvtWa
 dS7qseaia3GnZK/n8szE98Y=
 =zLpL
 -END PGP SIGNATURE-
 
 
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 
 
 
 
 
 -- 
 Jordi Segués Daina
 ---
 Andorra GSM: (+376) 35 35 68
 France GSM: (+33) (0)6 81 88 35 55
 [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
 AIM: superjordix
 Skype: callto://superjordix
 ---
 http://www.JordiX.com
 ___
 LARTC mailing list
 LARTC@mailman.ds9a.nl
 http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
 


signature.asc
Description: Digital signature
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] traffic shaping question

2007-01-22 Thread Nagy Gabor Peter
Hi list,

I have read the lartc 9th chapter, the bandwidth management part.

I think I understand the principle, but I have a question.

So I have a firewall that has several different interfaces. The most
important for my question is the Internet interface, which is a 2mbps
leased line.
I have an interface into the protected network, I have a DMZ interface,
and I have an interface with direct connection to a client.

Here is what I need:
Internet - DMZ + Internet - LAN + Internet - firewall traffic
together should not exceed 1.5mbps

At the moment I have a tbf, that limits everything that goes to the LAN,
and another that limits everything going to the internet.

I would like to shape the incoming traffic from the internet. OK, I
understand that I cannot influence the senders out there not to try to
send me packets, I can only influence how fast these packets are sent
from me.

But can I somehow treat all incoming traffic together?

Because my knowledge at the moment is only some shaping possibilities on
the LAN interface and on the DMZ interface.

I have only one idea, but I don't know if it is feasible, and if it is,
how to do that.

So I thought that I will create a virtual interface, and route all
traffic from the Internet through this one. So incoming on Internet
interface, outgoing on virtual interface, and from there incoming on the
firewall machine, or outgoing on the LAN or the DMZ interface.

Does it sound good? How can I do that? (I suppose I have to read other
chapters in the lartc guide. Could you point me out where to start? What
to look for?)

Or is there another solution? What would you recommend?

Cheers,
Gabor
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] traffic shaping question

2007-01-22 Thread Marco Berizzi
Nagy Gabor Peter wrote:

 So I thought that I will create a virtual interface, and route all
 traffic from the Internet through this one. So incoming on Internet
 interface, outgoing on virtual interface, and from there incoming on
the
 firewall machine, or outgoing on the LAN or the DMZ interface.

 Does it sound good? How can I do that? (I suppose I have to read other
 chapters in the lartc guide. Could you point me out where to start?
What
 to look for?)

Yes. The virtual interface is called IFB. Look at
the iproute2 package source under doc/actions


___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


RE: [LARTC] HTB? (NEWBIE)

2007-01-22 Thread Flechsenhaar, Jon J
How the guide say I expect to see on router (with command tc -s qdisc ls dev
eth0) that unclassified traffic, like ssh, get rounded to 20: but it doesn't 
happen.

-  Traffic just doesn't get rounded to a class.  If you want traffic going to a 
class you need to specify a filter.  Did you by chance man the default class 
1:30.  Any traffic not classified should end up there.  

-  Packet shaping is only done on the Egress/root side of an interface.  That 
is why you will only see stats for that.  You won't see stats for the ingress 
side.  Does this answer any of your questions?

Jon Flechsenhaar
Boeing WNW Team
Network Services
(714)-762-1231
202-E7

-Original Message-
From: Simone84bo [mailto:[EMAIL PROTECTED] 
Sent: Friday, January 19, 2007 3:05 PM
To: lartc@mailman.ds9a.nl
Subject: [LARTC] HTB? (NEWBIE)

Hi to all
I am studying HTB on LARTC how to. I realize a simple configuration on
router:
tc qdisc add dev eth0 root handle 1: htb default 30 tc class add dev eth0 
parent 1: classid 1:1 htb rate 3mbit burst 15k tc class add dev eth0 parent 1:1 
classid 1:10 htb rate 2mbit burst 15k tc class add dev eth0 parent 1:1 classid 
1:20 htb rate 1mbit burst 15k tc qdisc add dev eth0 parent 1:10 handle 10: sfq 
perturb 10 tc qdisc add dev eth0 parent 1:20 handle 20: sfq perturb 10 tc 
filter add dev eth0 protocol ip parent 1:0 prio 1 u32 match ip dport 80 0x 
flowid 1:10

After this configuration I make a ssh connection between client and server.
How the guide say I expect to see on router (with command tc -s qdisc ls dev
eth0) that unclassified traffic, like ssh, get rounded to 20: but it doesn't 
happen. The count of ssh traffic packet result only on root qdisc. Why?
A second question if i want to limited rate of all my router which 
configuration can i realize?

Thanks
 --
 Email.it, the professional e-mail, gratis per te: http://www.email.it/f
 
 Sponsor:
 Problemi di Liquidità? Con Logos Finanziaria 30.000 € in 24 ore a dipendenti e 
lavoratori autonomi con rimborsi fino a 120 mesi, clicca qui

 Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=2911d=20070120


___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] routing in tunnel mode

2007-01-22 Thread Michael P. Soulier
Hello,

Looking here

http://www.ipsec-howto.org/x299.html

I've set up a vpn in transport mode with two linux boxes. I'm now trying to
set it up in tunnel mode. After using the example keys, trying to ping, it
doesn't work because the route network isn't routable. 

This mention is in the howto

If you tunnel is not working, please check your routing. Your hosts need to
know that they should send the packets for the opposite network to you vpn
gateway. The easiest setup would be using your vpn gateway as default
gateway.

But how does one set up a route like that, since the network is multiple hops
away, the route command isn't going to accept it?

[EMAIL PROTECTED] ~]# route add -net 172.16.113.0 netmask 255.255.255.0
gw 10.33.15.145
SIOCADDRT: Network is unreachable

Some help please. 

Mike
-- 
Michael P. Soulier [EMAIL PROTECTED], 613-592-2122 x2522
Any intelligent fool can make things bigger and more complex... It takes a
touch of genius - and a lot of courage to move in the opposite direction.
--Albert Einstein
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] ipip tunnel docs broken

2007-01-22 Thread Michael P. Soulier
Hello,

Looking here

http://lartc.org/howto/lartc.tunnel.ip-ip.html

It says to load a new_tunnel.o module. There is no such module on 2.6.9, so
where would I find up-to-date documentation on ip tunnels in the Linux kernel?

Thanks,
Mike
-- 
Michael P. Soulier [EMAIL PROTECTED], 613-592-2122 x2522
Any intelligent fool can make things bigger and more complex... It takes a
touch of genius - and a lot of courage to move in the opposite direction.
--Albert Einstein
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] aes unsupported algorithm for ipsec?

2007-01-22 Thread Tim Stoop

Hi all,

I'm trying to create a IPsec tunnel from a Debian Etch machine to a
Cisco PIX. Part of my config is the following:

add x.x.x.x x.x.x.x esp 34501 -m tunnel -E aes-ctr abcdefghijklmnop;

When I try to set this using setkey, it fails with the following message:

line 9: unsupported algorithm at [abcdefghijklmnop]
parse failed, line 9.

Can anyone tell me what I'm doing wrong?

--
Gegroet,
Tim
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] two internet providers

2007-01-22 Thread Danut Chereches

hello

i have slackware installed and i have two internet connections , 
ADSL(2,5mbps) + CableModem(1mbps)

i want to share the connections in a small network
NAT for the ADSL connection, and a proxy server for the cablemodem 
connection
i searched all over the internet (probably not where i was supposed to) 
but i could'n find a solution

if someone could give me a tip i would really appreciate it


sorry for my english
___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] LoadBalancing on many asimetric different dsl's.

2007-01-22 Thread sAwAr
Thanks for all your answers. I ask my question in different way because I still 
don't get answer which will be satysfying for me.
Did anybody set similar configuration in the past? Have someone any suggestions 
how to set it to configure loadbalancing to avoid situation when one link is 
empty or only upload is used and other links are full. Does weights ensure that 
upload and download on all links with different up/down speeds will be equally 
loaded?

Pozdrawiam
sawar
 On Mon, Jan 22, 2007 at 01:21:32PM +0100, Jordi Segues wrote:
  the above is actually covered in the wiki howto.  Bu tyou need to setup
 
  snat on
  each interface, then connection tracking takes care of sending each
 stream 
  out
  the right interface, you need to use snat and not MASQ.
  
  Great news :)
  And thankyou for the details.
  But could you give the link to the wiki howto?
  I only found old doc.
 been a while since i had a look, quick google gave me this
 
 http://lartc.org/howto/lartc.rpdb.multiple-links.html
 
 I have this booked market as the wiki
 http://linux-net.osdl.org/index.php/Main_Page
 
 But I think the former is what you want
 
  
  Thanks!
  
  
  Then you need to setup up some ip rule tables for each of the
 interfaces.
  
  
  my ip ru looks like this
  
  0:  from all lookup local
  200:from 144.132.145.38 lookup cable
  201:from 60.241.248.86 lookup adsl
  32766:  from all lookup main
  32767:  from all lookup default
  
  
  my ip r sh tab default
  
  default  proto static  metric 5
  nexthop via 144.132.144.1  dev vlan2 weight 1
  nexthop via 10.20.20.230  dev ppp0 weight 20
  default via 10.20.20.230 dev ppp0  src 60.241.248.86  metric 20
  default via 144.132.144.1 dev vlan2  src 144.132.145.38  metric 30
  
  
  This works fine for me, I have tracked packets with tcpdump on both the
 
  server
  and the client.
  
  Alex
  
  
  
  
   Thanks!
  
   Jordi Segues
  
   On 22 Jan 2007 09:49:28 +0100, sAwAr [EMAIL PROTECTED] wrote:
   Hi,
   
   my company have just bought new network and I have question about
 one
   problem.
   As in topic we must use few completely different dsl#039;s and balance 
  traffic
   between them.
   2M/0,5M 4Mb/0,5M 8M/0,5M
   M=Mb/s
   I#039;ve never done such thing before so I have doubts how it will 
   work.
 If
   the links are symmetric 2/2 4/4 8/8 there is no problem because
 with
   weights I can compensate  the difference between them and achieve
 nice
   results. But what in my situation?
   My questions are: how to set load balancing to get all links
 equally
   loaded and avoid situation when the up load will be full and
 download
   almost empty? I believe this situation can happen due to fact that
 load
   balancing is based on flows and for example p2p or smpt/pop3 will
 eat
   whole upload.
   If my problem isn#039;t clear I#039;ll try to explain it better later.
   
   
   Thanks in advance.
   Pozdrawiam
   sawar
   
  
 --
   Wolne adresy pocztowe @interia.eu  http://link.interia.pl/f19e8
   
   ___
   LARTC mailing list
   LARTC@mailman.ds9a.nl
   http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
   
  
  
   --
   Jordi Segués Daina
   ---
   Andorra GSM: (+376) 35 35 68
   France GSM: (+33) (0)6 81 88 35 55
   [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
   AIM: superjordix
   Skype: callto://superjordix
   ---
   http://www.JordiX.com
   ___
   LARTC mailing list
   LARTC@mailman.ds9a.nl
   http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
  
  
  
  -BEGIN PGP SIGNATURE-
  Version: GnuPG v1.4.6 (GNU/Linux)
  
  iD8DBQFFtJ1/kZz88chpJ2MRAhGKAJ9xthAZnQ/ovr82sa/x5j4BFJGgWwCgvtWa
  dS7qseaia3GnZK/n8szE98Y=
  =zLpL
  -END PGP SIGNATURE-
  
  
  ___
  LARTC mailing list
  LARTC@mailman.ds9a.nl
  http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
  
  
  
  
  
  -- 
  Jordi Segués Daina
  ---
  Andorra GSM: (+376) 35 35 68
  France GSM: (+33) (0)6 81 88 35 55
  [EMAIL PROTECTED] / MSN: [EMAIL PROTECTED]
  AIM: superjordix
  Skype: callto://superjordix
  ---
  http://www.JordiX.com
  ___
  LARTC mailing list
  LARTC@mailman.ds9a.nl
  http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
  
 
 



--
Co robi Indianin w banku? Zobacz  http://link.interia.pl/f19e4

___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


Re: [LARTC] DGD patch not detecting dead gateway

2007-01-22 Thread Tom Lobato



Hi! Thank you.


Manish Kathuria escreveu:

On 1/19/07, Tom Lobato [EMAIL PROTECTED] wrote:

  Hello all!

  I applied http://www.ssi.bg/~ja/routes-2.6.8-10.diff patch to kernel
2.6.8.1 and it works fine, or almost fine. It does the load balancing
well, but when one link is dropped it continues to try it.
  At the end of http://www.ssi.bg/~ja/nano.txt it is said to ping
gateway 1 and gateway 2, for the kernel to know if that route is
working, but since my linux is connected to the links through 1
dedicated link and one adsl modem, I tryied to:
1) remove ethernet cable from linux nic: the patch worked well,
began to send traffic only to the yet working, link.
2) remove telephone line from adsl modem (or external ethernet
cable from the dedic. link switch): the patch didn't work, continued
trying to send traffic to the dropped link.
  So, I think its happening because linux, since it can ping the
switch (or adsl modem) thinks that link is good.

  Did you have this problem? Some hint?
  Thank you!



My experience has been mixed. The patch worked very well in many cases
but in some it worked only if the first hop gateway was down and not
any of the subsequent hops. So as you mentioned its happening since it
can ping the switch / modem, it thinks the link is good. You can make
a script which will keep on running in the background and check it the
links are up or not and if any of the links is down, it can change the
default route and provide a failover.



Oh yes, in really I already made such scripts, before to know this
patch, using
4.2. Routing for multiple uplinks/providers from Adv-Routing-HOWTO
information.
But facing this problem, I think the best solution is to use it again.

Somebody know if there is working in progress for solve this?
Is there some goal for include this patch to the mainstream kernel? What
is the possibility of it?



Tom Lobato



___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc


[LARTC] determine internet connection duration

2007-01-22 Thread Ming-Ching Tiew

Anyone has idea of what would be the best way to track 
connection time some a particular user to the internet ?

Imagine a wifi network where the users will connect to the
system via DHCP ( there is no PPPOE session involved ).
If there is a need to track internet usage based on connection
time ( to the internet ), what would be the best way to track it ?

Appreciate any input or ideas.

 



___
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc