After repair -- NIC interface not found

[David Greenberg_1]

HP-UX 11.0 server just went though a reboot into maintenance mode (I blew away etc/fstab)
When system ws repaired (with HP rep's help), network seemed OK (ping worked). Now I find I cannot telnet to machine.
rc.log gives error message starting with S112kminit:

interface_state[01]:not found

This error repeats many times in rc.log
Network config files have not changed. Not sure ehat this means or where to go with this.
(BTW: system's built in NIC is not used - an added 100MB E'net card is used instead)

Many thanks for your help...

David

[PIYUSH D. PATEL]

Hi,

Do #ioscan -fnC lan
and check whether the device files of your lan card exists and it is configured in kernel.

Run commands lanscan and landiag. If the hardware status is down in the lanscan command then run

#ifconfig lan0 up

Or go to landiag and Reset the lancard.

Piyush

[S.K. Chan]

Do you by any chance make a copy of startup configuration files and leave those behind in /etc/rc.config.d or /sbin/init.d ? If you find any files that are not suppose to be in those 2 directories, move 'em elsewhere then reboot your machine again. If the answer is NO you can ignore my suggestion.

[harry d brown jr]

David,

Check your /dev/rc.config.d/netconf file.

Look for INTERFACE_STATE. a shrt of mine is :

# cat /etc/rc.config.d/netconf|grep -v "^#"|grep -v "^$"
HOSTNAME="vpart2"
OPERATING_SYSTEM=HP-UX
LOOPBACK_ADDRESS=127.0.0.1
INTERFACE_NAME[0]="lan0"
IP_ADDRESS[0]="141.123.226.41"
SUBNET_MASK[0]="255.255.240.0"
BROADCAST_ADDRESS[0]=""
INTERFACE_STATE[0]=""
DHCP_ENABLE[0]=0
ROUTE_DESTINATION[0]="default"
ROUTE_MASK[0]=""
ROUTE_GATEWAY[0]="141.123.224.1"
ROUTE_COUNT[0]="1"
ROUTE_ARGS[0]=""
GATED=0
GATED_ARGS=""
RDPD=0
RARP=0
INTERFACE_NAME[1]="lan5"
IP_ADDRESS[1]="141.123.226.42"
SUBNET_MASK[1]="255.255.240.0"


live free or die
harry

[MANOJ SRIVASTAVA]

Hi David

In all probabilities the /etc/rc.config.d/netconf is corrupted or pointing a wron interface.

1. lanscan will give the lan interface the machine has sensed . check for the 100 mbps interface instance . Incase the card is not sensed then you need to get the interafce up .

2.Incase you can see the 2nd instance then you need to edit the /etc/rc.config.d.netconf file to look somehting like the file attached.

3. Stop the start the network daemon
/sbin/init.d/net stop and start .



Manoj Srivastava

[Wilfred Chau_1]

Something wrong with your netconf.
First, change this interface_state[01] to
interface_state[1] or interface_state[0]

Or else try the rest if the above doesn't solve
your problem.

the builtin lan uaually is lan0.

do a lanscan and look for the hardware path
of the new nic.

replace lan0 with lan?(new nic)

/sbin/init.d/net start

Does it work now?

[sven verhaegen]

Hi

having a look at the issue apparently it does point to an issue with the parameter described , one would expect indeed this format [01] to be somewhat strange , the value between those brackets is used in a 'while do' sequence so shouldn't be much of a problem unless the script does check on it's correctness which is what I fear ,so cahnge it to [1] ... there is however a small but very usefull way of finding out what is wrong at the startup in the scripts , that way is called


set -x


this command put in front of the relevant startup script (being S112 and the S008 net and lancard S300-400 script ranges pecific script) will show the execution of them during startup you can then see at what exact command it fails and what specific parameter is read at that moment , a good way of knowing if it is script related is to try after a failure to do manual configuration dong

ifconfig x.x.x.x netmask x.x.x.x up

if this works fine it most surely is a script error..

[David Greenberg_1]

I'm working again!
Turns out my netconf file was screwy (had interface 0 and 01 defined -- both pointing to same card!), and my etc/servies file got wiped out.

Thanks for the help!