[LARTC] Simple route 2nd look please

[LARTC] Simple route 2nd look please

["C. Bergström"]

I want B to route (temporarily) to both the .65 gw and eventually move 
to xxx.xxx.xxx.83 being the default gw, but I can't add that route..

I'm missing some obvious, but if someone would take a 2nd look it would 
be appreciated.  I also have requested to get access to the switch ,but 
that's still waiting.


Server B
ip a s
1: eth0: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 100
    link/ether 00:0b:db:91:84:53 brd ff:ff:ff:ff:ff:ff
    inet xxx.xxx.xxx.87/26 brd xxx.xxx.xxx.127 scope global eth0
2: eth1: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 100
    link/ether 00:0b:db:91:84:54 brd ff:ff:ff:ff:ff:ff
    inet xxx.xxx.xxx.84/32 scope global eth1

arping -I eth1 xxx.xxx.xxx.83
ARPING xxx.xxx.xxx.83 from xxx.xxx.xxx.84 eth1
Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C0]  0.956ms <-- 
Correct interface
Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  1.210ms <-- Incorrect
Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  0.712ms
Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  0.711ms
---------------
ip route add default via xxx.xxx.xxx.83 dev eth1 table T1
RTNETLINK answers: Network is unreachable
eris ~ # route add -net xxx.xxx.xxx.84/31 gw xxx.xxx.xxx.83
SIOCADDRT: Network is unreachable

ip r s
127.0.0.0/8 dev lo  scope link
default via xxx.xxx.xxx.65 dev eth0



Server C
2: eth0: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 1000
    link/ether 00:00:25:c1:cc:c0 brd ff:ff:ff:ff:ff:ff
    inet xxx.xxx.xxx.83/31 scope global eth0
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 1000
    link/ether 00:00:25:c1:cc:c1 brd ff:ff:ff:ff:ff:ff
    inet xxx.xxx.xxx.82/26 scope global eth1

(Temporary)
ip r s
xxx.xxx.xxx.64/26 dev eth1  proto kernel  scope link  src 207.135.120.82
xxx.xxx.xxx.64/26 dev eth0  proto kernel  scope link  src 207.135.120.83
127.0.0.0/8 dev lo  scope link
default via xxx.xxx.xxx.65 dev eth1
_______________________________________________
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc

Re: [LARTC] Simple route 2nd look please

[Martin A. Brown]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Greetings,

 : I want B to route (temporarily) to both the .65 gw and eventually 
 : move to xxx.xxx.xxx.83 being the default gw, but I can't add that 
 : route..
 : 
 : I'm missing some obvious, but if someone would take a 2nd look it 
 : would be appreciated.  I also have requested to get access to the 
 : switch ,but that's still waiting.

This is an L3 problem.  After reading your description, I'm guessing 
that each of your servers has two physical connections to the same 
L2 (broadcast domain) and you have made modifications to the 
routing table (at least on Server B) before you tried to solve this 
problem below.

 : Server B
 : ip a s
 : 1: eth0: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 100
 :    link/ether 00:0b:db:91:84:53 brd ff:ff:ff:ff:ff:ff
 :    inet xxx.xxx.xxx.87/26 brd xxx.xxx.xxx.127 scope global eth0
 : 2: eth1: <BROADCAST,MULTICAST,UP,10000> mtu 1500 qdisc pfifo_fast qlen 100
 :    link/ether 00:0b:db:91:84:54 brd ff:ff:ff:ff:ff:ff
 :    inet xxx.xxx.xxx.84/32 scope global eth1
 :
 : arping -I eth1 xxx.xxx.xxx.83
 : ARPING xxx.xxx.xxx.83 from xxx.xxx.xxx.84 eth1
 : Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C0]  0.956ms <-- Correct interface
 : Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  1.210ms <-- Incorrect
 : Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  0.712ms
 : Unicast reply from xxx.xxx.xxx.83 [00:00:25:C1:CC:C1]  0.711ms

I call the above problem ARP flux [0].  It's an extraordinarily 
common problem when you have multiple connections to the same 
Ethernet.

 : ip r s
 : 127.0.0.0/8 dev lo  scope link
 : default via xxx.xxx.xxx.65 dev eth0
 :

Unless you are running some weirdo networking startup scripts you 
have made changes to the routing table or lost routes on this box 
since you brought up the interface on eth0.

Note!  The "ip address" output for eth0 shows that you have an L3 
address of 207.135.120.87/26.  This means you should have had a 
network route that looked like this:

  207.135.120.64/26 dev eth0 proto kernel scope link src 207.135.120.87

Since this route is missing on Server B, something has removed it.  

 : ip route add default via xxx.xxx.xxx.83 dev eth1 table T1
 : RTNETLINK answers: Network is unreachable
 : eris ~ # route add -net xxx.xxx.xxx.84/31 gw xxx.xxx.xxx.83
 : SIOCADDRT: Network is unreachable

RTNETLINK is telling you that it has no way to reach 207.135.120.83.  
You can do two things:

  * restore the network route, 207.135.120.64/26:  "ip route add 
    207.135.120.64/26 dev eth0 src 207.135.120.87"
  * create a host route to the L3 address you want to use as a next 
    hop:  "ip route add 207.135.120.83 dev eth0"

Good luck!

- -Martin

 [0] http://linux-ip.net/html/ether-arp.html#ether-arp-flux

     (Sorry for the character encoding mismatch.)

- -- 
Martin A. Brown
http://linux-ip.net/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: pgf-0.72 (http://linux-ip.net/sw/pine-gpg-filter/)

iD8DBQFF8hdAHEoZD1iZ+YcRAnMNAJ4y+0/GKY3sUEx85IshFuKrCQ4mXwCfeQLO
YmGSNeQgmGX8LDGqGySG9CA=hYRK
-----END PGP SIGNATURE-----
_______________________________________________
LARTC mailing list
LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc