No reply using ping even when network card is ok

[Clarence J]

Hi,

I'm currently running a HP-9000 D370 server. I do not seem to be able to ping the default gateway and all other nodes. I used to be able to do this until recently. Some settings may have been unknowingly been changed recently.

I checked the network card and looks ok. However, when I do:

arp -a

I do not get any information. Can someone help me to diagnoise how to overcome the problem?

Thanks!!

[Steven Sim Kok Leong]

Hi,

Check the following:
1) Perform a linkloop from the server to another system on the same LAN/subnet. Check whether the linkloop is successful.
2) Perform another continuous ping and check the traffic stats of the switch port that this server is connected, ie. whether the inbound traffic increases.

One possibility is that the DUPLEX mode on the switch port does not match the DUPLEX mode on the server. If you are using auto-negotiation, it might have failed. Try to hardcode the DUPLEX mode on both the switch end and the server end.

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Printaporn_1]

Hi,

make sure that you start from basic
1.#lanscan
make sure that all hardware state is UP.
2.#ifconfig lan0 (or other lan interface you're using)
they should say up.. you can see IP address , netmask.
3.ping your own IP
4.ping IP in same subnet
5.ping gateway IP

[Clarence J]

Hi Steven,

Thank you for your reply. I'm not able to ping to anywhere. I'm still using the same hub port as before the problem.

Any other things I can check?

[Philip Chan_1]

Clarence,

Is your machine configured with an IP address? try ping'ing to itself and do "ifconfig lan0", check if the IP address was present.

Rgds,
Philip

[Steven Sim Kok Leong]

Hi,

When you said your network interface is okay, I had assumed that you have performed the necessary checks on the IP address, netmask etc. I had also assumed that you would be using a switch. Note that using a hub for a server is highly insecure as traffic can be sniffed easily.

Have you also checked through the steps that Printaporn has posted? In addition, was the link status on your network card okay? What about the link status for the hub port that your server is connected to? linkloop will test the network connectivity at the link/MAC level.

Another check is your /var/adm/syslog/syslog.log for the possibility of an IP address conflict.

We will need the responses to the tests we have posted in order to proceed with the next step of troubleshooting or resolution where applicable.

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Stefan Schulz]

Think we need more information here. Do you ping using the IP Adress or the hostname?

If you are using the IP adress, do you get a timeout or any other erromessage?

If you are using hostname what happens if you use the IP adress of the desired system? Could be a hostname resolution problem.

Are you using DNS? What does a nslookup tell you? What does a nslookup tell you?

Could you provide us with the output of
lanscan
ifconfig lan0
ping your own IP
ping IP in same subnet
ping gateway IP

Also /etc/nsswitch.conf and /etc/resolv.conf could be interesting.

Regards Stefan

[Vincenzo Restuccia]

Edit /etc/rc.config.d/netconf verify the GATEWAY,/sbin/init.d/net stop|start, verify the new configuration with netstat -rn.

[Clarence J]

Hi All,

Thanks for your replies.

The following are the findings based on the info. you have requested:

- ifconfig1 lan1

lan1: flags=843
inet 192.168.166.220 netmask ffffff00 broadcast 192.168.166.255

- ping own IP
There IS A REPLY

- ping IP in same subnet
NO REPLY

- ping gateway IP
NO REPLY

The light is lit at the network card and hub.

- No errors in /var/adm/syslog/syslog.log

- Ping using both IP-Address or hostname
There IS REPLY

lanscan
Hardware Path: 8/16/6
Station Address: 0x080009AF8B59
Crd In#: 1
Hdw State: UP
Net-Interface NamePPA : lan1 snap1
NM ID: 1
MAC Type: ETHER
HP-DLPI Support: Yes
DLPI Mjr#: 119

- /etc/nsswitch.conf
hosts: files
services: files
network: files
aliases: files
protocol: files

/etc/resolv.conf
domain exchange
nameserver 192.168.166.1 # default GW


Note: my default gateway is 192.168.166.1

Thanks to all of you for all the responses. I hope you can help me resolve this issue.

Have a nice week ahead!

[Printaporn_1]

Hi,

Try with
1.netfmt -t 50 -f /var/adm/nettl.LOG00 > /tmp/nettl

check /tmp/nettl , any ERROR ?
2.change lan cable ? if possible change with machine that you sure that it work.
3.change to another hup port ?
4.use lanadmin select nmid that corresponding to youe lan , and reset , any error ?

[Steven Sim Kok Leong]

Hi,

What is the response to the linkloop on the MAC address of the adjacent system on the same hub/LAN? The linkloop is needed to identify any problems at the Ethernet connectivity to the other server. In addition, perform another linkloop from the adjacent system to your current system. Is there any response?

An example of how to perform a linkloop:

A successful linkloop would flag "OK" ie:

# linkloop 0x001083FDB6C6
Link connectivity to LAN station: 0x001083FDB6C6
-- OK

An unsuccessful linkloop would flag the following:

# linkloop 0x001083FDB6C7
Link connectivity to LAN station: 0x001083FDB6C7
error: get_msg2 getmsg failed, errno = 4
-- FAILED
frames sent : 1
frames received correctly : 0
reads that timed out : 1

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Philip Chan_1]

Check the LAN cable, and the hub (or switch) port at the other end of the cable. Make sure they're both all right.

Do the linkloop test as others suggested as well.

Can you post the output of "netstat -r" too?

~Philip

[Clarence J]

Hi again,

I have run the following:

netfmt -t 50 -f /var/adm/nettl.LOG00 > /tmp/nettl

******* STREAM/UX *******
Subsystem : STREAMS
Log Class: Error

10:59: 36 009469 1 T..5306 5 ar_rput_dlpi:DL_UDERROR_IND, dl_dest_addr_leng4

When I run a linkloop

error: expected primitive 0x30, got DL_ERROR_ACK
dl_error_primitive = 0x2d
dl_errno = 57
error- did not receive data part of message

Thanks alot people!!

[Clarence J]

netstat -r

Routing tables
Dest/Netmask Gateway Flags Refs Use Inter Pmtu
localhost localhost UH 0 261 la0 4136
hpd370 hpd370 UH 0 12 lan1 4136
192.168.166.0 hpd370 U 2 0 lan1 1500
127.0.0.0 localhost U 0 0 la0 4136
default gateway UG 0 0 lan1 1500

Hope this helps!

[Printaporn_1]

Hi,

Can you sucessful reset lan without error using lanadmin ?
if you already change cable / Hup port and not help , it can be HW problem.

[Steven Sim Kok Leong]

Hi,

Based on your linkloop error messages, can you verify the that the port speed is compatible on both the hub and your server? If your NIC card is a 100BT (ie. that does not support 10BT) and your hub is a 10BT hub, you will get similar linkloop error messages) due to incompatibility of network speed.

If that is the case, you will either need to change your hub to one that supports 100BT or change your NIC card to one that supports 10BT.

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Steven Sim Kok Leong]

Hi,

If your NIC card is a 10/100 BT card, please check your /etc/rc.config.d/hpbase100conf that you did not manually set it to the wrong speed. Look out for the HP_BASE100_SPEED[subscript] settings:

# HP_BASE100_SPEED Manually set the driver's speed. Can be one of: 10HD, 10FD, 100HD, 100FD, auto_on

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Clarence J]

Hi,

I've tried to use lanadmin to reset the LAN Interface but have the following error:

Link connectivity to LAN station:
error: expected primitive 0x30, got DL_ERROR_ACK
dl_error_primitive = 0xe8
dl_errno = 0x04
dl_unix_errno=57 (16)-unable to reset LAN Interface
error-did not receive data part of message.

When I ran the display on the lanadmin, the following was seen:

Operation Status(value)= down(2)


Does it mean that I have problem with the network card?

Thanks everyone for your help so far!!!

[Philip Chan_1]

Clarence,

Besides the link speed, watch out for duplex mode too. You should be able to get both information (link speed and duplex mode) from your network administrator. And, you could probably referencing other machines in the same segment for that.

Rgds,
Philip

[Clarence J]

Hi,

I've obtained the following from lanadmin, display:

SPEED= 10000000

From /etc/rc.config.d/hpbase100conf:

HP_BASE100_INTERFACE_NAME[0]=
HP_BASE100_STATION_ADDRESS[0]=
HP_BASE100_SPEED[0]=

Do I have to make any setting here? How do I go about it?

Thanks!

[Steven Sim Kok Leong · Accepted solution]

Hi,

What is the speed that your hub supports? Is it 10BT or 100BT? There are hubs that only supports 100BT? What is the speed that your NIC card can support? Check at the back of the NIC card itself. Once you have identified these unknowns, we will be able to tell whether it is necessary to change your configuration settings or whether you need to change your NIC card or hub.

Since you are using a hub and not a switch, the DUPLEX mode is definitely at HALF-DUPLEX.

Hope this helps. Regards.

Steven Sim Kok Leong
Brainbench MVP for Unix Admin
http://www.brainbench.com

[Clarence J]

Hey Everyone!!!

Thanks very much!!! You have helped me narrow down the problem and have it solved!!!

The UTP cable was actually plugged into the wrong port on the server box!!! We did not see the 10-BaseT port below the server. Thanks to your good deduction, I went to take a look at the server again and found that the cable was pulgged to the wrong port.

BTW, the card into which the cable was plugged in is a J3515A. Any idea what is this? 100-BaseT card?

Thanks to everyone!! You guys r REALLY FANTASTIC!

[Philip Chan_1]

J3515A is a 100BT network card (so you got 2 NICs in your server).

Glad to hear you have your problem solved.

Cheers,
Philip