LAM/MPI logo

LAM/MPI General User's Mailing List Archives

  |   Home   |   Download   |   Documentation   |   FAQ   |   all just in this list

From: nadzhatul sheema (naskatui_83_at_[hidden])
Date: 2005-04-05 23:02:15


hello...i have problem with my LAM configuration in my cluster. the master node has 2 nic card and all the slaves have one nic card. i have configure the connection to use ssh instead of rsh. my lam_bhost.def is look like this
master
node01
node02
node03
i have a problem with lamboot and when i run the recon -d
 
[aizura_at_master lam-7.0.6]$ recon -d
n-1<26486> ssi:boot: Opening
n-1<26486> ssi:boot: opening module globus
n-1<26486> ssi:boot: initializing module globus
n-1<26486> ssi:boot:globus: globus-job-run not found, globus boot will not run
n-1<26486> ssi:boot: module not available: globus
n-1<26486> ssi:boot: opening module rsh
n-1<26486> ssi:boot: initializing module rsh
n-1<26486> ssi:boot:rsh: module initializing
n-1<26486> ssi:boot:rsh:agent: /bin/ssh -x
n-1<26486> ssi:boot:rsh:username: <same>
n-1<26486> ssi:boot:rsh:verbose: 1000
n-1<26486> ssi:boot:rsh:algorithm: linear
n-1<26486> ssi:boot:rsh:priority: 10
n-1<26486> ssi:boot: module available: rsh, priority: 10
n-1<26486> ssi:boot: finalizing module globus
n-1<26486> ssi:boot:globus: finalizing
n-1<26486> ssi:boot: closing module globus
n-1<26486> ssi:boot: Selected boot module rsh
n-1<26486> ssi:boot:base: looking for boot schema in following directories:
n-1<26486> ssi:boot:base: <current directory>
n-1<26486> ssi:boot:base: $TROLLIUSHOME/etc
n-1<26486> ssi:boot:base: $LAMHOME/etc
n-1<26486> ssi:boot:base: /usr/etc
n-1<26486> ssi:boot:base: looking for boot schema file:
n-1<26486> ssi:boot:base: lam-bhost.def
n-1<26486> ssi:boot:base: found boot schema: /usr/etc/lam-bhost.def
n-1<26486> ssi:boot:rsh: found the following hosts:
n-1<26486> ssi:boot:rsh: n0 localhost (cpu=1)
n-1<26486> ssi:boot:rsh: n1 master (cpu=1)
n-1<26486> ssi:boot:rsh: n2 node01 (cpu=1)
n-1<26486> ssi:boot:rsh: n3 node02 (cpu=1)
n-1<26486> ssi:boot:rsh: n4 node03 (cpu=1)
n-1<26486> ssi:boot:rsh: resolved hosts:
n-1<26486> ssi:boot:rsh: n0 localhost --> 127.0.0.0
n-1<26486> ssi:boot:rsh: n1 master --> 10.0.0.1 (origin)
n-1<26486> ssi:boot:rsh: n2 node01 --> 10.0.0.2
n-1<26486> ssi:boot:rsh: n3 node02 --> 10.0.0.3
n-1<26486> ssi:boot:rsh: n4 node03 --> 10.0.0.4
n-1<26486> ssi:boot:rsh: starting RTE procs
n-1<26486> ssi:boot:base:linear: starting
n-1<26486> ssi:boot:base:linear: booting n0 (localhost)
n-1<26486> ssi:boot:rsh: starting recon on (localhost)
n-1<26486> ssi:boot:rsh: starting on n0 (localhost): tkill -N -d
n-1<26486> ssi:boot:rsh: launching remotely
n-1<26486> ssi:boot:rsh: attempting to execute "/bin/ssh -x localhost -n echo $SHELL"
-----------------------------------------------------------------------------
LAM failed to execute a process on the remote node "localhost".
LAM was not trying to invoke any LAM-specific commands yet -- we were
simply trying to determine what shell was being used on the remote
host.
 
LAM tried to use the remote agent command "/bin/ssh"
to invoke "echo $SHELL" on the remote node.
 
This usually indicates an authentication problem with the remote
agent, or some other configuration type of error in your .cshrc or
.profile file. The following is a list of items that you may wish to
check on the remote node:
 
        - You have an account and can login to the remote machine
        - Incorrect permissions on your home directory (should
          probably be 0755)
        - Incorrect permissions on your $HOME/.rhosts file (if you are
          using rsh -- they should probably be 0644)
        - You have an entry in the remote $HOME/.rhosts file (if you
          are using rsh) for the machine and username that you are
          running from
        - Your .cshrc/.profile must not print anything out to the
          standard error
        - Your .cshrc/.profile should set a correct TERM type
        - Your .cshrc/.profile should set the SHELL environment
          variable to your default shell
 
Try invoking the following command at the unix command line:
 
        /bin/ssh -x localhost -n echo $SHELL
 
You will need to configure your local setup such that you will *not*
be prompted for a password to invoke this command on the remote node.
No output should be printed from the remote node before the output of
the command is displayed.
 
When you can get this command to execute successfully by hand, LAM
will probably be able to function properly.
-----------------------------------------------------------------------------
n-1<26486> ssi:boot:base:linear: Failed to boot n0 (localhost)
n-1<26486> ssi:boot:base:linear: aborted!
-----------------------------------------------------------------------------
recon was not able to complete successfully. There can be any number
of problems that did not allow recon to work properly. You should use
the "-d" option to recon to get more information about each step that
recon attempts.
 
Any error message above may present a more detailed description of the
actual problem.
 
Here is general a list of prerequisites that *must* be fulfilled
before recon can work:
 
        - Each machine in the hostfile must be reachable and operational.
        - You must have an account on each machine.
        - You must be able to rsh(1) to the machine (permissions
          are typically set in the user's $HOME/.rhosts file).
 
        *** Sidenote: If you compiled LAM to use a remote shell program
            other than rsh (with the --with-rsh option to ./configure;
            e.g., ssh), or if you set the LAMRSH environment variable
            to an alternate remote shell program, you need to ensure
            that you can execute programs on remote nodes with no
            password. For example:
 
        unix% ssh -x pinky uptime
        3:09am up 211 day(s), 23:49, 2 users, load average: 0.01, 0.08, 0.10
 
        - The LAM executables must be locatable on each machine, using
          the shell's search path and possibly the LAMHOME environment
          variable.
        - The shell's start-up script must not print anything on standard
          error. You can take advantage of the fact that rsh(1) will
          start the shell non-interactively. The start-up script (such
          as .profile or .cshrc) can exit early in this case, before
          executing many commands relevant only to interactive sessions
          and likely to generate output.
-----------------------------------------------------------------------------
n-1<26486> ssi:boot:rsh: finalizing
n-1<26486> ssi:boot: Closing

                
---------------------------------
Yahoo! Messenger
 Show us what our next emoticon should look like. Join the fun.