From: Oliver Geisler <lartc@docawk.org>
To: lartc@vger.kernel.org
Subject: [LARTC] Double gateway - Two interfaces to same subnet
Date: Mon, 03 Feb 2003 16:50:44 +0000 [thread overview]
Message-ID: <marc-lartc-104429104413386@msgid-missing> (raw)
Hi everybody,
Last week I posted a message concerning how to gain control on routing
aliased IP with the goal to establish two gateways to the localnet. It
became a problem 'two interfaces to same subnet'. Thanks to Martin for
your hints:
http://mailman.ds9a.nl/pipermail/lartc/2003q1/006930.html
I've read the docs mentioned. The arp-flux problem should be solved by
switching /proc/sys/net/ipv4/conf/(all,eth0,eth1)/arp_filter and
../rp_filter to 1. But there remained a problem. only one interface is
ansering the arp requests, even for ip of the other interface.
for more understanding this is my setup:
hamburg:~# ip addr
1: lo: <LOOPBACK,UP> mtu 16436 qdisc noqueue
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
2: teql0: <NOARP> mtu 1500 qdisc noop qlen 100
link/void
3: dummy0: <BROADCAST,NOARP> mtu 1500 qdisc noop
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
4: bond0: <BROADCAST,MULTICAST,MASTER> mtu 1500 qdisc noop
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
5: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 00:e0:7d:e2:9e:8d brd ff:ff:ff:ff:ff:ff
inet 192.168.10.9/24 brd 192.168.10.255 scope global eth0
6: eth1: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 00:e0:7d:e2:9e:8f brd ff:ff:ff:ff:ff:ff
inet 192.168.10.8/24 brd 192.168.10.255 scope global eth1
7: eth2: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 00:e0:7d:e2:9e:94 brd ff:ff:ff:ff:ff:ff
inet 62.x.x.90/29 brd 62.x.x.95 scope global eth2
8: eth3: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 00:e0:7d:e2:9e:96 brd ff:ff:ff:ff:ff:ff
inet 62.x.x.90/29 brd 62.x.x.95 scope global eth3
9: eth4: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 100
link/ether 00:e0:7d:dc:95:2f brd ff:ff:ff:ff:ff:ff
inet 192.168.11.8/24 brd 192.168.11.255 scope global eth4
hamburg:~# ip route
62.x.x.89 dev eth2 scope link
62.x.x.88/29 dev eth3 scope link
192.168.11.0/24 dev eth4 proto kernel scope link src 192.168.11.8
192.168.10.0/24 dev eth0 proto kernel scope link src 192.168.10.9
192.168.10.0/24 dev eth1 proto kernel scope link src 192.168.10.8
default via 62.x.x.89 dev eth0
from a host within the localnet an arping to both ips results in:
moskau:~ # arping -c 3 192.168.10.9
ARPING 192.168.10.9 from 192.168.10.1 eth0
Unicast reply from 192.168.10.9 [00:E0:7D:E2:9E:8D] 0.665ms
Unicast reply from 192.168.10.9 [00:E0:7D:E2:9E:8D] 0.619ms
Unicast reply from 192.168.10.9 [00:E0:7D:E2:9E:8D] 0.619ms
Sent 3 probes (1 broadcast(s))
Received 3 response(s)
moskau:~ # arping -c 3 192.168.10.8
ARPING 192.168.10.8 from 192.168.10.1 eth0
Unicast reply from 192.168.10.8 [00:E0:7D:E2:9E:8D] 0.680ms
Unicast reply from 192.168.10.8 [00:E0:7D:E2:9E:8D] 0.635ms
Unicast reply from 192.168.10.8 [00:E0:7D:E2:9E:8D] 0.631ms
Sent 3 probes (1 broadcast(s))
Received 3 response(s)
so it seems for both arp-requests the interface eth0 on machine hamburg
is responding.
with the docs from julians site and others I think the arp response is
based on the reverse routing path. both interfaces eth0 and eth1 set a
path to the localnet 192.168.11.0/24. so eth0 responds for both ips
becaue it's in the first place in the routing table. right?
my next idea was to use julians 'ip arp' extension to iproute2, in order
to gain control to the arp mechanism.
I'm running a debian system with kernel 2.4.18 with lodable module
support switched off and a minimal driver support, but all ipnetworking
modules compiled in.
I've downloaded
ftp://ftp.inr.ac.ru/ip-routing/iproute2-2.4.7-now-ss020116-try.tar.gz
http://www.linuxvirtualserver.org/~julian/iproute2-iparp-3.diff
after unpacking the tar-ball and moving the directory iproute2 to
iproute2-2.4.7-now-ss020116-try, I patched the source with
hamburg:/usr/src# patch -p0 < iproute2-iparp-3.diff
patching file iproute2-2.4.7-now-ss020116-try/ip/Makefile
patching file iproute2-2.4.7-now-ss020116-try/ip/ip.c
patching file iproute2-2.4.7-now-ss020116-try/ip/ip_common.h
patching file iproute2-2.4.7-now-ss020116-try/ip/iparp.c
patching file iproute2-2.4.7-now-ss020116-try/ip/ipmonitor.c
(option -p1 didn't work, however)
setting a link /usr/src/linux to my actual kernel source and running
make in the iproute2 source directory ... somethings going terribly wrong.
(...)
gcc -D_GNU_SOURCE -O2 -Wstrict-prototypes -Wall -g -I../include-glibc
-I/usr/include/db3 -include ../include-glibc/glibc-bugs.h
-I/usr/src/linux/include -I../include -DRESOLVE_HOSTNAMES -c -o
iproute.o iproute.c
gcc -D_GNU_SOURCE -O2 -Wstrict-prototypes -Wall -g -I../include-glibc
-I/usr/include/db3 -include ../include-glibc/glibc-bugs.h
-I/usr/src/linux/include -I../include -DRESOLVE_HOSTNAMES -c -o
iprule.o iprule.c
gcc -D_GNU_SOURCE -O2 -Wstrict-prototypes -Wall -g -I../include-glibc
-I/usr/include/db3 -include ../include-glibc/glibc-bugs.h
-I/usr/src/linux/include -I../include -DRESOLVE_HOSTNAMES -c -o
iparp.o iparp.c
iparp.c:68: `ARPA_TABLE_INPUT' undeclared here (not in a function)
iparp.c:68: initializer element is not constant
iparp.c:68: (near initialization for `arp_tables[0].code')
iparp.c:69: `ARPA_TABLE_OUTPUT' undeclared here (not in a function)
iparp.c:69: initializer element is not constant
iparp.c:69: (near initialization for `arp_tables[1].code')
iparp.c:70: `ARPA_TABLE_FORWARD' undeclared here (not in a function)
iparp.c:70: initializer element is not constant
iparp.c:70: (near initialization for `arp_tables[2].code')
iparp.c:113: `ARPA_TABLE_ALL' undeclared here (not in a function)
iparp.c:113: initializer element is not constant
iparp.c:113: (near initialization for `f.tb')
iparp.c: In function `print_arprule':
iparp.c:174: `RTM_NEWARPRULE' undeclared (first use in this function)
iparp.c:174: (Each undeclared identifier is reported only once
iparp.c:174: for each function it appears in.)
iparp.c:174: `RTM_DELARPRULE' undeclared (first use in this function)
iparp.c:180: dereferencing pointer to incomplete type
iparp.c:184: dereferencing pointer to incomplete type
iparp.c:187: dereferencing pointer to incomplete type
(...)
iparp.c:861: `ARPA_OIF' undeclared (first use in this function)
iparp.c:879: `RTM_GETARPRULE' undeclared (first use in this function)
make[1]: *** [iparp.o] Error 1
make[1]: Leaving directory `/usr/src/iproute2-2.4.7-now-ss020116-try/ip'
make: *** [all] Error 2
can anybody tell me what's missing?
... and am I on the right way to solve my initial problem 'two
interfaces on the same subnet?'
thanks in advance,
oli:-)
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
next reply other threads:[~2003-02-03 16:50 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-02-03 16:50 Oliver Geisler [this message]
2003-02-03 17:43 ` [LARTC] Double gateway - Two interfaces to same subnet Martin A. Brown
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=marc-lartc-104429104413386@msgid-missing \
--to=lartc@docawk.org \
--cc=lartc@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.