wiki:Internal/Nodes/NodeTesting

Version 4 (modified by ssugrim, 12 years ago) ( diff )

A brief Guide to Node Testing

The procedure is different for different generations of nodes. We'll break it up into two classes:

Old rabbit Based CM

  1. First locate the node4-11 Node ID box, it looks like
    Node ID box
  2. Wire the control, CM, monitor, and keyboard to their respective ports
    Wired up node
  3. Connect Power, the led on the top of the box should start flashing
  4. Connect via ssh to console.outdoor.orbit-lab.org
  5. Run
    ssugrim@console.outdoor:~$ omf-5.2 stat
     INFO NodeHandler: init OMF Experiment Controller 5.2.408
    -----------------------------------------------
     INFO Topology: Loading topology 'system:topo:all'.
     Testbed : outdoor.orbit-lab.org
     .
     .
     .
     Node n_4_10 - State: NODE NOT AVAILABLE
     Node n_4_11 - State: POWEROFF
     Node n_4_12 - State: NODE NOT AVAILABLE
    -----------------------------------------------
    
    Note the status of node 4-11.
  6. In a seprate session telnet to cons4-11 3025
    ssugrim@console.outdoor:~$ telnet cons4-11 3025
    Trying 10.1.224.59...
    Connected to cons4-11.outdoor.orbit-lab.org.
    Escape character is '^]'.
    
  7. Turn on the node and then watch the console session for the pxe message and quickly collect the mac address. It should look like:
    Yukon PXE v3.06  (20041109)
    (C)Copyright 2003-2004 Marvell(R). All rights reserved.
    
    Pre-boot eXecution Environment (PXE) v2.1
    (C)Copyright 1997-2000 Intel Corporation.
    
    CLIENT MAC ADDR: 00 0F EA 02 87 F4
    GUID: 30303046-4541-3032-3837-4634FFFFFFFF
    DHCP.-
    
  8. Once you have the mac, turn off the node with omf-5.2 tell offh [4,11].
    ssugrim@console.outdoor:~$ omf-5.2 tell offh [4,11]
    
     INFO NodeHandler: init OMF Experiment Controller 5.2.408
    ---------------------------------------------------
     Testbed : outdoor.orbit-lab.org - Command: offHard
     Node n_4_11 - Ok
    ---------------------------------------------------
    
  9. ssh to dhcp1 and edit the /etc/dhcpd3/dhcpd.conf file. (You may need to bounce of internal4 to do this)
  10. Find the line for node 4-11, in the outdoor group, and replace the mac address with the one you copied from the console. The original line may look like
        host node4-1 {hardware ethernet 00:03:1d:08:19:23; fixed-address 10.40.4.1;}
    
        # Dummy node for TESTING. This node should be wired to the outdoor control vlan.
        host node4-11 {hardware ethernet 00:03:1d:0a:9f:23; fixed-address 10.40.4.11;}
    
        host node1-201 {hardware ethernet 00:0F:EA:7C:C3:D1; fixed-address node1-201.outdoor.orbit-lab.org;}
    
  11. Restart the dhcp server:
    root@dhcp1:/etc/dhcp3# /etc/init.d/dhcp3-server restart
     * Stopping DHCP server dhcpd3                                                                                                                                                                          [ OK ]
     * Starting DHCP server dhcpd3
    
  12. Now try to load an image with omf-5.2 load [4,11] ImageName and observe the console. If you see faulty behavior you want to capture, log the telnet session.
    ssugrim@console.outdoor:~$ omf-5.2 load [4,11] inventory.ndz
    Imaging nodes: '[4,11]' with image 'inventory.ndz'
    (Domain:  default from hostname)
    (Timeout:  800 sec.)
     INFO NodeHandler: init OMF Experiment Controller 5.2.408
    
    You may also see error messages on the monitor attached to the node.

New Lantronix Based CM

Attachments (2)

Note: See TracWiki for help on using the wiki.