SEO

Hi all,


I'm runing Solaris 10 on SunFire V100, it has 2 ethernet cards: dmfe0
and dmfe1. Now, I only have 1 cable run to this server. When I plug the

cable to dmfe0, everything is ok, i can connect to server from LAN and
vice versa easily. However, if I move the cable to dmfe1, the
connection is lost although the status of dmfe1 comes to up.


Here is output:


1- I plug cable to dmfe0 and try to ping to my pc at 192.168.1.44. It's

ok.


-bash-3.00# ifconfig -a
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv 4,VIRTUAL> mtu
8232 index 1
inet 127.0.0.1 netmask ff000000


dmfe0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500
index 2
inet 192.168.1.2 netmask ffffff00 broadcast 192.168.1.255
ether 0:3:ba:11:bf:9b
dmfe1: flags=1000803<UP,BROADCAST,MULTICAST,IPv4> mtu 1500 index 3
inet 192.168.1.4 netmask ffffff00 broadcast 192.168.1.255
ether 0:3:ba:11:bf:9c


-bash-3.00# ping 192.168.1.44
192.168.1.44 is alive


2- I move cable to dmfe1 and try to ping 192.168.1.44 again but it fail

-bash-3.00# ifconfig -a
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv 4,VIRTUAL> mtu
8232 index 1
inet 127.0.0.1 netmask ff000000
dmfe0: flags=1000803<UP,BROADCAST,MULTICAST,IPv4> mtu 1500 index 2
inet 192.168.1.2 netmask ffffff00 broadcast 192.168.1.255
ether 0:3:ba:11:bf:9b
dmfe1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500
index 3
inet 192.168.1.4 netmask ffffff00 broadcast 192.168.1.255
ether 0:3:ba:11:bf:9c


-bash-3.00# ping 192.168.1.44
no answer from 192.168.1.44


I wonder why my dmfe1 doesn't work!!!

"SauDoi" (anhtais@gmail.com) writes:
> Hi all,
>
>
> I'm runing Solaris 10 on SunFire V100, it has 2 ethernet cards: dmfe0
> and dmfe1. Now, I only have 1 cable run to this server. When I plug the
>
> cable to dmfe0, everything is ok, i can connect to server from LAN and
> vice versa easily. However, if I move the cable to dmfe1, the
> connection is lost although the status of dmfe1 comes to up.

Hrm.

A couple of ideas...

Try as root to 'ifconfig dmfe1 plumb up' - occasionally kernel needs to be
reminded that the interface is plumbed.

You have two different interfaces with two different MAC addresses on the
same ethernet segment. If you are connecting them to the same ethernet
cable, you might need to wait a few minutes for a switch to notice the MAC
change, and deal with it. If you are in a managed environment, it's
possible that your admin might have hardcoded your MAC to your port to
prevent unauthorized use (plugging in arbitrary hardware, etc).


Lastly, the following was valid under Solaris 2.6, I don't know if it is still
the case. It is based on my observations.

If you are connecting two ethernet interfaces of a sun box to the same
ethernet segment and give two distinct IP addresses to each one of the
interfaces on the same segment, kernel will get confused which interface
to use as the source MAC/IP address. As a result it will round-robin
between the two (Which was a correct behavior accoring to Stevens' book).
In turn this will confuse anything it tries to connect to, as only half TCP
session will be seen by the other system, and other system will reject the
other half of the packets.

So it is possible that you might need to shut down the first interface if
you are connecting to the second one.

'ifconfig dmfe0 down' *should* do it.

HTH. HAND.



--
Trouble rather the tiger in his lair, then the Sysadmin amongst his UNIX boxen.
For to you Programs and their Source Code are things mighty and enduring,
But to him they are but toys of the moment,
To be overturned by flicking of the power switch.... Computer Lessons: SNV '97

Thanks. I will try to shut dmfe0 and test again. Sun box could get
confused as you said when I moved cable to dmfe1 and tried to ping
192.168.1.44, Sun box didn't which interface should be used to forward
packet, it chose dmfe0 and .... my problem appeared

Thx for your answer and explaination.

--
Best Regards.