Two Servers via ethernet crossover cable

Hi,

I'm trying to communicate two servers (HP DLProliant 380 G5) via a crossover cable, but I don't get them to ping each other.

I am working in RHEL 4.

These are the steps I've performed:

1) Plugged the crossover cables in eth2.
(note: The cable is brand new. Also I built one myself and the behaviour is the same.)

2) I have configured the interfaces in both servers as follows:

Server1:

# cat /etc/sysconfig/network-scripts/ifcfg-eth2
DEVICE=eth2
ONBOOT=NO
BOOTPROTO=static
IPADDR=192.168.101.1
NETMASK=255.255.255.0
NOZEROCONF=yes
HWADDR=00:24:81:7D:5E:61
TYPE=Ethernet

Server2:

cat /etc/sysconfig/network-scripts/ifcfg-eth2
DEVICE=eth2
ONBOOT=NO
BOOTPROTO=static
IPADDR=192.168.101.2
NETMASK=255.255.255.0
NOZEROCONF=yes
HWADDR=00:24:81:7D:62:3D
TYPE=Ethernet

3) I restart the interfaces in both servers: ifdown eth2 / ifup eth2

See green LED on both sides (not blinking but fixed)

4) ifconfig output:

Server1:

eth2      Link encap:Ethernet  HWaddr 00:24:81:7D:5E:61
          inet addr:192.168.101.1  Bcast:192.168.101.255  Mask:255.255.255.0
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
          Base address:0x5000 Memory:fdbe0000-fdc00000

Server2:

eth2      Link encap:Ethernet  HWaddr 00:24:81:7D:62:3D
          inet addr:192.168.101.2  Bcast:192.168.101.255   Mask:255.255.255.0
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:0 (0.0 b)  TX bytes:0 (0.0 b)
          Base address:0x5000 Memory:fdbe0000-fdc00000

5)The routes are taken into account:

In both servers:

# netstat -nr
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
192.168.101.0   0.0.0.0         255.255.255.0   U         0 0          0 eth2

6) I try to ping from one to the other but not success at all:

Server1 to Server2:

ping 192.168.101.2
PING 192.168.101.2 (192.168.101.2) 56(84) bytes of data.
From 192.168.101.1 icmp_seq=0 Destination Host Unreachable

Running at same time a tcpdump (in Server1), don't see any traffic over eth2:

# tcpdump -i eth2
tcpdump: listening on eth2

0 packets received by filter
0 packets dropped by kernel

7) Check the arp but the entry is incompleted (not catch of MAC)

# arp
Address                  HWtype  HWaddress           Flags Mask            Iface
192.168.101.2                        (incomplete)                              eth2

8) And here is where the issue must be.
I found that ethtool and mii-tool don't agree in the link status. First one indicates it as dow, while mii indicates up.

And here is the output for ethtool and mii-tool:

# ethtool eth2
Settings for eth2:
        Supported ports: [ TP ]
        Supported link modes:   10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
                                1000baseT/Full
        Supports auto-negotiation: Yes
        Advertised link modes:  10baseT/Half 10baseT/Full
                                100baseT/Half 100baseT/Full
                                1000baseT/Full
        Advertised auto-negotiation: Yes
        Speed: Unknown! (65535)
        Duplex: Unknown! (255)
        Port: Twisted Pair
        PHYAD: 1
        Transceiver: internal
        Auto-negotiation: on
        Supports Wake-on: umbg
        Wake-on: d
        Current message level: 0x00000007 (7)
        Link detected: no


# mii-tool eth2
eth2: negotiated 100baseTx-FD flow-control, link ok

I think I should believe ethtool, but I see the LED lights green, the cables is properlly plugged (tried with two different ones).
Any idea why interefaces are down and can't communicate?

Last:

Here are the drivers:

# ethtool -i eth2
driver: e1000
version: 7.4.24b-NAPI
firmware-version: 5.12-2
bus-info: 0d:00.0

Thanks

This:

UP BROADCAST MULTICAST  MTU:1500  Metric:1

...is odd. On my systems I have

UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

It really doesn't believe the cable's plugged in. The LED's might just be hardwired. mii-tool talks with the transceiver itself, so it's the one I'd trust... Very strange that it autonegotiated a 100baseT link with a 1000baseT peer.

Have you tried a normal cable, not a crossover cable? I've sometimes experienced odd behavior with ports capable of autodetecting the cable type. You might also try forcing the speed and duplex, gigabit links often have problems negotiating -- though if you've discovered a gigabit interface that's incapable of negotiating with itself that's certainly something new! :slight_smile:

Corona688 is correct. Generally no need for a crossover cable for Gigabit NICs.

If you connect these systems into your network, can they communicate with each other?

I see you have other network adapters in your systems. Are they in use?

Are you using the exact same NICs (make, model#, driver) on both systems? NC373i?

I wondered too. I assumed it's a two-port nic and wondered why eth2 and not eth0 or eth1. Besides I looked at the support pages @ hp.com, namely the driver download section and I found the driver the OP has mentioned (e1000), but there is a e1000e too. I don't know what's the difference between them. There are even at least two non-intel drivers (Broadcom, QLogic) :eek: