[Jisco]

Thanks Steve

The following are the steps I took to build the NIM server:

1) Installed NIM from CD and the following are the listings:

root@nim_server:/var/log/csm$ lslpp -L all | grep bos.sysmgt.nim*
bos.sysmgt.nim.client 5.3.0.20 C F Network Install Manager -
bos.sysmgt.nim.master 5.3.0.20 C F Network Install Manager -
bos.sysmgt.nim.spot 5.3.0.0 C F Network Install Manager – SPOT

2) root@nim_server:/var/log/csm$ lsnim -c machines
master machines master
nim_client machines standalone

3) The following is the listing of the machines:
root@nim_server:/var/log/csm$ lsnim -l master
master:
class = machines
type = master
max_nimesis_threads = 20
if_defined = chrp.up.ent
comments = machine which controls the NIM environment
platform = chrp
netboot_kernel = mp
if1 = en3 nim_master 0
cable_type1 = N/A
Cstate = ready for a NIM operation
prev_state = ready for a NIM operation
Mstate = currently running
serves = boot
serves = bosinst_data
serves = inst_script
serves = lpp_source1
serves = nim_script
serves = spot3
master_port = 1058
registration_port = 1059
reserved = yes

root@nim_server:/var/log/csm$ lsnim -l nim_client
nim_client:
class = machines
type = standalone
comments = Created on Thu 5 Oct 17 11 58 2006
connect = shell
platform = chrp
netboot_kernel = up
if1 = en3 nim_client 0
cable_type1 = N/A
Cstate = ready for a NIM operation
prev_state = ready for a NIM operation
Mstate = not running

Something that I have noticed is that when I view the machine properties from WSM, the cable type defaults to vnc (instead of N/A, though this is Ok from command line) and the “network adapter logical device name” defaults to ent0 which I do not want (because that interface is configured with an address that is not patched in so cannot be contacted remotely)

Having backed up the NIM db (using your script nim_backup_db.sh), I again used wsm to reconfigure the NIM environment. (the menu path is: overview and tasks à configure the network installation management environment à configure this host as NIM server. On the next screen I specify all options such as primary network interface for the NIM master, input device for installation images, lpp_source name and directory). This creates successfully and the resources are listed from the CLI. For example,

root@nim_master:/var/log/csm$ lsnim -l spot3
spot3:
class = resources
type = spot
comments = Created on Thu 5 Oct 17 17 11 2006
plat_defined = chrp
arch = power
bos_license = yes
Rstate = ready for use
prev_state = verification is being performed
location = /oradb_backup/nim_lpp_spot_res/spot_3/spot3/usr
version = 5
release = 3
mod = 0
oslevel_r = 5300-00
alloc_count = 0
server = master
Rstate_result = success
mk_netboot = yes
mk_netboot = yes

I then use HMC to activate the client in SMS mode and then select an Ethernet adapter to use. I then set the properties (ip address, server ip address and subnet mask) of the NIM client to ensure that the server and client are in the same subnet. I test ping and it returns “ping success”. I then select the configured adapter as my boot device and then boot as normal. From here it is obvious that it is communicating with the server because returns the name of the bootfile as configured on the master. After a few seconds the errors are thrown.

When I specify the gateway address of the bootfile name, I get a tftpd error. If I remove this entry completely or specify gateway address instead of bootfile name, I get a bootp error.