LARTC - Mar 2007 - DGD of upstream routers

Hi

I have read various info, and mailing list archives, but have not found 
an answer to the following.

I have a few servers with configurations similar to the following:

They each have multiple uplinks to the Internet, and a sample config is 
as follows:

eth1 is 192.168.0.1, connected to 192.168.0.2
eth2 is 192.168.1.1, connected to 192.168.1.2

My default route looks like:
ip route add scope global equalize nexthop via 192.168.0.2 dev eth1 \ 
weight 1 nexthop via 192.168.1.2 dev eth2 weight 1

If one line goes down, I would like the second to be used exclusively 
until the first comes back up

However, the IPs 192.168.0.2 and 192.168.1.2 are always available and 
reachable.   It is the connection past those devices which may drop.

Do any of the available options or patches take this into account, or do 
I have to write custom ping scripts to try reach remote hosts via each 
gateway, and modify the routes if a line appears to be down.

Hi all

I have seen the below question asked a few times, but not seen any
answers.  Is this because

1) it is not possible

2) It is really simple and I shouldn''t even be asking the question  :)


Can somebody please enlighten me.

Thanks

Chris


On Thu, 2007-03-08 at 18:48 +0200, Chris Picton wrote:
> Hi
> 
> I have read various info, and mailing list archives, but have not found 
> an answer to the following.
> 
> I have a few servers with configurations similar to the following:
> 
> They each have multiple uplinks to the Internet, and a sample config is 
> as follows:
> 
> eth1 is 192.168.0.1, connected to 192.168.0.2
> eth2 is 192.168.1.1, connected to 192.168.1.2
> 
> My default route looks like:
> ip route add scope global equalize nexthop via 192.168.0.2 dev eth1 \ 
> weight 1 nexthop via 192.168.1.2 dev eth2 weight 1
> 
> If one line goes down, I would like the second to be used exclusively 
> until the first comes back up
> 
> However, the IPs 192.168.0.2 and 192.168.1.2 are always available and 
> reachable.   It is the connection past those devices which may drop.
> 
> Do any of the available options or patches take this into account, or do 
> I have to write custom ping scripts to try reach remote hosts via each 
> gateway, and modify the routes if a line appears to be down.
> 
> 
> _______________________________________________
> LARTC mailing list
> LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
---------------------+----------------------------------------------------
        Chris Picton | PGP Key ID: 9D28A988 (wwwkeys.pgp.net)
  Technical Director | PGP Key Fingerprint:
     Tangent Systems | 2B46 29EA D530 79EC D9EA 3ED0 229D 6DD6 9D28 A988
        011 447 8096 | "Quid quid latine dictum sit, altum viditar"
chrisp@tangent.co.za | http://www.tangent.co.za/keys/chrisp.asc
---------------------+----------------------------------------------------


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

Chris Picton napisał(a):
> Hi all
> 
> I have seen the below question asked a few times, but not seen any
> answers.  Is this because
> 
> 1) it is not possible
> 
> 2) It is really simple and I shouldn''t even be asking the question
:)
> 
Usually you do such things using some dynamic routing protocols (OSPF, 
BGP, RIP).

But it need cooperation from upstream provider. Quite good solution is 
using some kind of monitoring tool (nagios for example), which 
reachibility of some hosts and using that information you can alter 
routing table.

Or you can use shell script, combined with ping and cron :)

-- 
Michał Margula, alchemyx@uznam.net.pl, http://alchemyx.uznam.net.pl/
"W życiu piękne są tylko chwile" [Ryszard Riedel]