ping -i in HPUX11

[Petteri Kettunen]

I'd like to do simple check if some I/F is up or not. I've configured lan0&lan1 to one network with IPs 10.1.1.2 10.1.1.3. Then I try the simplest command to check if connection of one of these I/Fs is ok by pinging the switch whose IP is 10.1.1.1:

ping -i 10.1.1.2 10.1.1.1

No matter if lan0 cable is connected or not, ping always finds 10.1.1.1. Are there any tricks to make this work or am I completely lost?

[A. Clay Stephenson]

Hi:

You're not lost but it appears that you configured 2 ports on the same subnet and that is causing ambigiuos results. Typically packets get interleaved between the two ports on the same subnet.


Clay

[Magdi KAMAL]

Hi Petteri,

Tryusing the following :

1. lanscan on the local server and fetch out id the interfaces are UP.

2. ifconfig on local server for both interfaces and fetch out if they are UP.

3. traceroute from another subnet to the local server on both IPs to see how the IPS are solved.

Magdi

[linuxfan]

Hi Petteri,

Try linkloop (man linkloop for more information). you would need to know the PPA number and the mac address (which can be determined using lanscan). You would need the mac address of the remote node(host/switch).
(arp -a)

This way you know you are actually using the interface you intend to use.

-HTH
I am RU

[rick jones]

my read of the HP-UX 11 manpage for ping leads me to believe that -i only affects pings to multicast addresses, not unicast. if that is correct, when you have two interfaces configured into the same subnet you have no great control over how things are routed.

however.... you might try a tusc trace of ping to see if it does anything differently with -i specified when the destination is a unicast address. if it does, then you might consider setting "ip_strong_es_model" to a value of one. that will cause routing lookups to include the source IP address, and that may get you what you want. no guarantees however.

if you are configuring both NICs into the same subnet for failover, you might consider using Auto Port Aggregation. it will load balance across the NICs and do failover and such. it presents a single virtual interface to the transport and you could assign both IPs to that interface.

[MANOJ SRIVASTAVA]

Hi Petteri Kettunen

Try this ,

just disconnect physical cable connection to 10.1.1.1 and try pinging that address , incase you are able to ping it then u have another device configured for that IP and hence this error.

Manoj Srivastava