Hi David,
from the HowTo
Instead of choosing one of the two providers as your default route, you now set up
the default route to be a multipath route. In the default kernel this will balance
routes over the two providers. It is done as follows (once more building on the
example in the section on split-access):
It looks like you did BOTH, and have two "default" routes. If you want load
balancing, I believe this is saying use the 2nd method and remove the first.
David Kempe wrote:
Hi Sluggers,
I have a script for dual internet connections that does this:
-------------------------
#!/bin/bash
IF1=eth1
IP1=203.219.190.106
P1=203.219.190.105
P1_NET=203.219.190.104
IF2=eth2
IP2=220.245.224.46
P2=220.245.224.45
P2_NET=220.245.224.44
IF0=eth0
P0_net=192.168.0.0
TABLE1=inet1
TABLE2=inet2
ip route add $P1_NET dev $IF1 src $IP1 table $TABLE1
ip route add default via $P1 table $TABLE1
ip route add $P2_NET dev $IF2 src $IP2 table $TABLE2
ip route add default via $P2 table $TABLE2
ip route add $P1_NET dev $IF1 src $IP1
ip route add $P2_NET dev $IF2 src $IP2
# preference for default route
ip route add default via $P1
# route out a given table choose what table to route with
ip rule add from $IP1 table $TABLE1
ip rule add from $IP2 table $TABLE2
ip route add $P0_NET dev $IF0 table $TABLE1
ip route add $P2_NET dev $IF2 table $TABLE1
ip route add 127.0.0.0/8 dev lo table $TABLE1
ip route add $P0_NET dev $IF0 table $TABLE2
ip route add $P1_NET dev $IF1 table $TABLE2
ip route add 127.0.0.0/8 dev lo table $TABLE2
#route balancing
ip route add default scope global nexthop via $P1 dev $IF1 weight 1
nexthop via $P2 dev $IF2 weight 1
-----------------------
its taken straight from the lartc howto.
however, it doesn't appear to be route balancing at all. I have kernel
2.4.27 with the extra route balancing patches from this page:
http://www.ssi.bg/~ja/#routes
Traffic into the second interface (IF2) works fine, even portforwards
through that interface work fine. You can ping it from out side etc.
However, no traffic from inside the network ever appears to be nated out
from it, nor does traffic originating from the router box ever go out
over that interface.
Is there some weight better than 1 that I need to adjust? What are
useful values to balance the routes evenly over the two interfaces? Or
is there something I am missing.
btw, the network diagram looks exactly like this one:
http://www.lartc.org/howto/lartc.rpdb.multiple-links.html
is there some proc magic I can query to check the kernel has all the
right bits?
thanks
Dave
smime.p7s
Description: S/MIME Cryptographic Signature
--
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
