Jump to content United States-English
HP.com Home Products and Services Support and Drivers Solutions How to Buy
» Contact HP
More options
HP.com home
HP XC System Software : Installation Guide > Chapter 4 Configuring and Imaging the System

Task 4: Run the discover Command to Discover System Components

» 

Technical documentation

Complete book in PDF
» Feedback
Content starts here

 » Table of Contents

 » Glossary

 » Index

The second step in the configuration process is the discovery of all system components. In addition to configuring the system topology by discovering all hardware components, the discover command also configures the administration network and interconnect, starts the MySQL service, and initializes the system configuration database.

  1. Begin this procedure as the root user on the head node.

  2. Enter the following command to capture command output into a file; this step is optional:

    # script your_filename 

  3. Start the automatic discovery of all network components in your system:

    # ./discover --system [--super] [--oldmp] [--outfile file_name]  \
       [--ic=AdminNet] [--verbose]
    Notes About the discover Command Line:
    • The --super option is required for large-scale systems (systems with up to 1024 compute nodes) with a hardware configuration that consists of a Super Root Switch into which multiple Root Administration Switches are connected.

    • The --oldmp option is required for Itanium-based systems because their management processors (MPs) have statically set IP addresses and are not configured to use DHCP.

      Do not include the --oldmp option on the command line for Opteron- or Xeon-based systems because those systems use iLO and LO-100i as their console port management devices, and as a result, should be using DHCP to obtain their IP addresses.

    • Use --outfile file_name to save a copy of the HP XC system information that will be written to the system management database. You must supply an existing file name on the head node in which to save the information.

      Specifying [--outfile file_name] enables you to use the --infile file_name option should it become necessary to restore the HP XC system to the point in the installation just after the discover command was run but before the cluster_config utility is run.

      However, if any MAC addresses have changed in the system since the time the first discover command was run, it may be better to perform a new full discovery process. You can modify the changed MAC addresses in your saved file, but do so carefully.

      The --verbose option is the only other valid option on the discover command line when the [--infile file_name] option is used.

    • Use the --ic=AdminNet option only if you are configuring the interconnect on the administration network. When the interconnect is configured on the Administration Network, a single LAN is used; the administration network and the interconnect share the same ports and switches. All other interconnect types (Myrinet, InfiniBand, and Quadrics®) are discovered automatically based on the interconnect found on the head node; specific command options are not necessary for those interconnect types.

    • HP recommends that you include the --verbose option because it provides useful feedback and enables you to follow the discovery process.

    • Table 4-1 and discover(8) contain information about additional keywords you can add to the command line to bypass some of the questions that will be asked during the discovery process. Use of these keywords is strictly optional.

    If you encounter problems during the discovery process, refer to “Troubleshooting the Discovery Process” for troubleshooting guidelines.

    The discover command does not properly discover HP ProLiant DL140 and DL145 nodes until the password is set. Running the discover command now sets up the IP address of the console ports so you can then telnet into them and set the password. This process is described in step 6.

  4. Follow along with the command output. The following example output was created using the --verbose option on a small, six-node Opteron-based system. The output you see will be different, especially if you have a large-scale system. The information you supply depends on whether or not you entered additional keywords on the command line.

    Table 4-1 describes each prompt and provides information to help you with your answers.

    Enter the actual total number of nodes in this cluster: 6
    Enter the number of nodes that are workstations: 0 
    
    Discovering Root switches...
    Enter the MAC address of the admin root switch
    in the format xx:xx:xx:xx:xx:xx : your_MAC_address    1
    Please enter the ProCurve switch Administrator  password: your_password  2
    Please re-enter password: your_password
    Restarting dhcpd
    
    Enter the common user name for all console port management 
    devices: your_user_name   3
    Please enter the password for your_user_name: your_password
    Please re-enter password: your_password
    
     waiting for network component at 172.20.65.1 to become available.
    ............
    Discovering 172.20.65.1 port 24 ... done
    Restarting dhcpd
     waiting for network component at 172.20.65.2 to become available.
    ...
    Discovering 172.20.65.2 port 25 ... Retrying switch discover in 5 seconds 
    No Switch Found
    Restarting dhcpd
    
    Discovering Branch switches...
    Discovering 172.20.65.1 port 1 ... Retrying switch discover in 5 seconds 
    No Switch Found
    Discovering 172.20.65.2 port 1 ... Retrying switch discover in 5 seconds 
    No Switch Found
    Restarting dhcpd
    
    Discovering Console Ports...
    running port_discover on 172.20.65.2 
    running port_discover on 172.20.65.2 
    Number of CPs found 6, Number expected 6
    Discovering 172.20.65.2 port 22 ... 
    Console Port Not Found. Disabled Console Port Entry Created.
    Discovering 172.20.65.2 port 21 ... Console Port OK
    Discovering 172.20.65.2 port 20 ... Console Port OK
    Discovering 172.20.65.2 port 19 ... Console Port OK
    Discovering 172.20.65.2 port 18 ... Console Port OK
    Discovering 172.20.65.2 port 17 ... Console Port OK
    Restarting dhcpd
    discovered_cps is 6
    Checking if all console ports are reachable ... 
    number of cps to check, 5
    pinging 172.21.0.13  no response from 172.21.0.13
    
    pinging 172.21.0.15  no response from 172.21.0.15
    
    pinging 172.21.0.12  
    pinging 172.21.0.14  
    pinging 172.21.0.11  
    Following console ports did not respond:
     172.21.0.13 172.21.0.15
    
    number of cps to check, 2
    pinging 172.21.0.13  no response from 172.21.0.13
    
    pinging 172.21.0.15  
    Following console ports did not respond:
     172.21.0.13
    
    number of cps to check, 1
    pinging 172.21.0.13  .done
    Opening /etc/hosts 
    Opening /etc/hosts.new.XC
    Opening /etc/powerd.conf
    Building /etc/powerd.conf ... Querying cp-n11
    Querying cp-n12
    Querying cp-n13
    Querying cp-n14
    Querying cp-n15[1]
    done
    
    Attempting to power on nodes with nodestring n[11-15]
    Powering on all known nodes ... done
    
    Discovering Nodes...
    running port_discover on 172.20.65.1 
    nodes Found = 1 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 1 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 1 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 1 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 1 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 2 nodes Expected = 6
    
    running port_discover on 172.20.65.1 
    nodes Found = 6 nodes Expected = 6
    
    All nodes initialized.
    Powering off all nodes but head node ... done
    Switch 172.20.65.1 port 22 ... Node Found
    Switch 172.20.65.1 port 21 ... Node Found
    Switch 172.20.65.1 port 20 ... Node Found
    Switch 172.20.65.1 port 19 ... Node Found
    Switch 172.20.65.1 port 18 ... Node Found
    Switch 172.20.65.1 port 17 ... Node Found
    Switch 172.20.65.1 port 16 ... NO Node Found 4
    Restarting dhcpd
    node n11 set to model ProLiant DL140 G2 
    node n12 set to model ProLiant DL140 G2 
    node n16 set to model ProLiant DL380 G4 
    node n14 set to model ProLiant DL360 G4
    node n13 set to model ProLiant DL360 G4
    node n15 set to model ProLiant DL380 G4
    Opening /etc/hosts 
    Opening /etc/hosts.new.XC
    Head Node CP device type set to iLO
    Waiting for power daemon to restart... done
    1

    Enter the MAC address of the switch that is connected to the administration ports. Do not enter the MAC address of the switch connected to the console ports.

    2

    Enter the password for the Root Administration Switch that you previously defined when you prepared the hardware. If you did not preset a password, press the Enter key.

    3

    Enter the common user name and password that you previously defined for the console port management devices (the iLO, LO-100i, or MP devices) when you prepared the hardware. If you did not predefine these items, press the Enter key.

    4

    This is expected behavior. Because this is a cluster discovery process, the command output No Switch Found or No Node Found is expected. This message does not indicate a failure unless a network component is plugged into that port on the switch.

    If necessary, see Appendix K: Troubleshooting for information about troubleshooting problems you may encounter during the discovery process.
    Note:

    The following prompts are displayed for large-scale systems:

    The discover process has detected 2 regions. Is this correct? [y/n] y
    switchName nems0-1-0 switchIP 172.20.65.1 type 2848 
    Enter the actual total number of nodes in region 1: 526
    Enter the number of nodes that are workstations: 0
    Enter the actual total number of nodes in region 2: 504
    Enter the number of nodes that are workstations: 0
    discoverSystem called with switch nems1-1-1 and Region Number 1 
  5. If you captured the output of the discover command in a file (described in step 2), press Ctrl+d to end the script now.

    This file is for your information only; store this file in a directory where you keep administrative data.

  6. Do the following on HP XC systems with HP ProLiant DL140 G2 and DL145 G2 nodes; otherwise, skip this step for all other server models:

    1. Use the telnet command to log in to each node's LO-100i console management device and change the password.

      To determine console port names, view the /etc/dhcpd.conf file and look for the characters cp- in the host name. Use the factory default user name admin and the default password admin to log in.

      # telnet cp-node_name
      login: admin
      password: admin

    2. Press Escape and Shift+9 to enter the command-line mode.

    3. Use the C[hange Password] option to change the console port management device password. The factory default password is admin; change it to the password of your choice. This password must be the same on every node in your system.

      Lights-Out> C 
      Type the current password> admin 
      Type the new password (max 16 characters)> your_password 
      Retype the new password (max 16 characters)> your_password 
      New password confirmed.Lights-Out exit

    4. Change the password for every HP ProLiant DL140 G2 and DL145 G2 in your system.

    5. Go back to step 3 and rerun the discover command with the same options.

Proceed to “Task 5: Define and Set Up Your System Environment” .



[1] See Appendix K if the discover process hangs while trying to discover console ports.

Printable version
Privacy statement Using this site means you accept its terms Feedback to webmaster
© 2003 Hewlett-Packard Development Company, L.P.