HP-UX 11.11 cannot ping gatway.

[jmckinzie]

I have a server that has two nics running at 1000FD (autonegotiate) and for some reason, I cannot ping the gateway (located on the same subnet.

I can ping loopback and ping the address of the server itself but, cannot ping the gateway.

Any help would be appreciated.

-Jody

[Tim Nelson]

post.

netstat -in
netstat -rn

Does your router respond to pings ?
Can the gateway/router ping the host ?

[jmckinzie]

Name Mtu Network Address Ipkts Opkts
lan1 1500 159.17.99.0 159.17.99.67 128 0
lan0 1500 159.17.98.0 159.17.98.67 2164 2164
lo0 4136 127.0.0.0 127.0.0.1 1697 1697
root@smtmwq1:/sbin/init.d
# netstat -rn
Routing tables
Destination Gateway Flags Refs Interface Pmtu
127.0.0.1 127.0.0.1 UH 0 lo0 4136
159.17.98.67 159.17.98.67 UH 0 lan0 4136
159.17.99.67 159.17.99.67 UH 0 lan1 4136
159.17.98.0 159.17.98.67 U 2 lan0 1500
159.17.99.0 159.17.99.67 U 2 lan1 1500
127.0.0.0 127.0.0.1 U 0 lo0 0
default 159.17.99.1 UG 0 lan1 0
default 159.17.98.1 UG 0 lan0 0


The gateway does respond to pings and i am getting the switch guys to verify they can ping the address.

[Rita C Workman]

If you have confirmed your nic is set properly - the next step I always do is call the networking person (if it's not you) and confirm that the port on the switch is set the same and that they locked it down.

Just to rule that out.

Rgrds,
Rita

[Tim Nelson]

You cannot have two default routes...

[jmckinzie]

They are on different networks.

[Tim Nelson]

As a test remove the default using lan0 and try to ping some addresses on lan1..
(route delete 159.17.98.1 1 )

if you want traffic to go over the other interface you will have to provide address ranges.

e.g.
route all 10.10.1.0 over lan1
route add 10.10.1.0 159.17.99.1 1

all other traffic out lan0
route add default 159.17.98.1 1

or vice-versa

[TTr]

I don't think the multiple default gateways is an issue. It may not be clean but it does not block lan traffic entirely.

Some routers do not reply to ping requests. Maybe this is the case here.

Have you tried traceroute? Have you tried pinging other servers?

Assuming you have checked your wiring and the link status is good, check the soft config of each LAN interface with "lanscan" and "lanadmin".
In lanadmin select "lan" and "display" for each interface and check the settings.

Then ask your network person to check on their end. Ping the servers from the gateway and check the gateway ports for being enabled and for traffic coming in.

[Bill Hassell · Accepted solution]

It is quite common today for network admins to purposely disable ping responses for gateways and routers. It is a security measure to make network topology more difficult to discover. This also has an effect on HP-UX where (depending on version and patches) will cause a LAN interface to be downed after a few minutes. This is called the dead gateway detection feature, but now it must be disabled. Check if it is enabled:

ndd -get /dev/ip ip_ire_gw_probe

If it is not 0, then it should be turned off:

ndd -set /dev/ip ip_ire_gw_probe 0

And to make permanent, add an entry in /etc/rc.config.d/nddconf:

TRANSPRT_NAME[0]=ip
NDD_NAME[0]=ip_ire_gw_probe
NDD_VALUE[0]=0

[jmckinzie]

This was due to the network cables in the correct ports.