Re: [Wien] parallel ssh error

Gavin Abo Mon, 30 Sep 2019 04:42:32 -0700

An additional comment, /home/username/WIEN2k (or ~/WIEN2k) is where Ihave WIEN2k installed. Whereas, you have installed WIEN2k at/servernode1 [1]. In the examples of my previous posts (e.g. [2]) youmight find some typographical errors were I forget to replace my/home/username/WIEN2k with your /servernode1.

It is best to have WIEN2k setup at a common path location on all nodes[3,4] (i.e., your vlsi1-vsli4). Therefore, I recommend to not haveWIEN2k at other locations among your system nodes like the/home/username/WIEN2K as you mentioned below.

[1]https://www.mail-archive.com/[email protected]/msg19036.html[2]https://www.mail-archive.com/[email protected]/msg19061.html[3]https://www.mail-archive.com/[email protected]/msg17988.html[4]https://www.mail-archive.com/[email protected]/msg09229.html


On 9/30/2019 12:59 AM, Peter Blaha wrote:

So there is progress as now the environment seems to be accepted inthe remote shell.

lapw1para (called by x_lapw, which is called by run_lapw -p) createsthe splitted klists-files (case.klist_1,...) and def fileslapw1_1.def,...


It uses the $cwd variable and executes basically:

ssh vlsi1 "cd $cwd; lapw1c lapw1_1.def "

Does this work on your computers ?



On 9/29/19 7:16 PM, Indranil mal wrote:

Now echo $WIENROOT is giving the $WIENROOT location.

echo $WIENROOT/lapw*

/home/username/WIEN2K/lapw0 /home/username/WIEN2K/lapw0_mpi/home/username/WIEN2K/lapw0para /home/username/WIEN2K/lapw0para_lapw/home/username/WIEN2K/lapw1 /home/username/WIEN2K/lapw1c/home/username/WIEN2K/lapw1c_mpi /home/username/WIEN2K/lapw1cpara/home/username/WIEN2K/lapw1_mpi /home/username/WIEN2K/lapw1para/home/username/WIEN2K/lapw1para_lapw /home/username/WIEN2K/lapw2/home/username/WIEN2K/lapw2c /home/username/WIEN2K/lapw2c_mpi/home/username/WIEN2K/lapw2cpara /home/username/WIEN2K/lapw2_mpi/home/username/WIEN2K/lapw2para /home/username/WIEN2K/lapw2para_lapw/home/username/WIEN2K/lapw3 /home/username/WIEN2K/lapw3c/home/username/WIEN2K/lapw5 /home/username/WIEN2K/lapw5c/home/username/WIEN2K/lapw7 /home/username/WIEN2K/lapw7c/home/username/WIEN2K/lapwdm /home/username/WIEN2K/lapwdmc/home/username/WIEN2K/lapwdmcpara /home/username/WIEN2K/lapwdmpara/home/username/WIEN2K/lapwdmpara_lapw /home/username/WIEN2K/lapwso/home/username/WIEN2K/lapwsocpara /home/username/WIEN2K/lapwso_mpi/home/username/WIEN2K/lapwsopara /home/username/WIEN2K/lapwsopara_lapw


ssh vlsi1 'echo $WIENROOT/lapw*'



However getting the same error

  stop error


grep: *scf1*: No such file or directory
cp: cannot stat '.in.tmp': No such file or directory
FERMI - Error
grep: *scf1*: No such file or directory
Parallel.scf1_1: No such file or directory.
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
bash: fixerror_lapw: command not found
bash: lapw1c: command not found
  LAPW0 END
hup: Command not found.


and lapw2 error file

  'LAPW2' - can't open unit: 30
  'LAPW2' -        filename: Parallel.energy_1
**  testerror: Error in Parallel LAPW2

On Sat, Sep 28, 2019 at 11:58 PM Gavin Abo <[email protected]<mailto:[email protected]>> wrote:


    The "sudo service sshd restart" step, which I forgot to copy and
    paste, that is missing is corrected below.

    On 9/28/2019 12:18 PM, Gavin Abo wrote:


    After you set both "SendEnv *" and "AcceptEnv *", did you restart
    the sshd service [1]?  The following illustrates steps that might
    help you verify that WIENROOT appears on a remote vlsi node:

    username@computername:~$ echo $WIENROOT

    username@computername:~$ export WIENROOT=/servernode1
    username@computername:~$ echo $WIENROOT
    /servernode1
    username@computername:~$ ssh vlsi
    Welcome to Ubuntu 18.04.3 LTS (GNU/Linux 4.15.0-64-generic x86_64)
    ...
    Last login: Sat Sep 28 12:04:07 2019 from xxx.x.x.x
    username@computername:~$ echo $WIENROOT

    username@computername:~$ exit
    logout
    Connection to vlsi closed.
    username@computername:~$ sudo gedit /etc/ssh/ssh_config
    [sudo] password for username:

    username@computername:~$ sudo gedit /etc/ssh/sshd_config

    username@computername:~$ grep SendEnv /etc/ssh/ssh_config
        SendEnv LANG LC_* WIENROOT
    username@computername:~$ grep AcceptEnv /etc/ssh/sshd_config
    AcceptEnv LANG LC_* WIENROOT

        username@computername:~$ sudo service sshd restart


    username@computername:~$ ssh vlsi
    ...
    username@computername:~$ echo $WIENROOT
    /servernode1
    username@computername:~$ exit

    [1]
https://askubuntu.com/questions/462968/take-changes-in-file-sshd-config-file-without-server-reboot

    On 9/28/2019 11:22 AM, Indranil mal wrote:

    Sir I have tried with " SetEnv * " Still nothing is coming with
    echo  commad and user name by mistake I posted wrong Otherwise no
    issue with user name and I have set the parallel options file
    taksset "no" and remote options are 1 1 in server and client
    machines.


    On Sat, 28 Sep 2019 11:36 Gavin Abo, <[email protected]
    <mailto:[email protected]>> wrote:

        Respected Sir, In my linux(Ubuntu 18.04 LTS) in ssh_config,
        and in sshd_config there are two line already "SendEnv LANG
        LC_*" "AcceptEnv LANG LC_*" respectively.


        The "LANG LC_*" probably only puts just the local language
        variables in the remote environment.  Did you follow the
        previous advice [1] of trying to use "*" to put all variables
        from the local environment?

        [1]
https://www.mail-archive.com/[email protected]/msg19049.html

        However, ssh vsli1 'echo $WIENROOT' gives nothing (blank).


        That seems to be the main cause of the problem as it should
        not return (blank) but needs to return "/servernode1" as you
        previously mentioned [2].

        [2]
https://www.mail-archive.com/[email protected]/msg19036.html

        Perhaps the message below is a clue.  It you had set the
        WIENROOT variable in .bashrc of your /home/vlsi accounts on
        each system, you likely have to login and use that same
        /home/vlsi account on the head node as the output below seems

to indicate login to a different /home/niel account. Alternatively, setting the WIENROOT variable in .bashrc of

        all /home/niel accounts on each node might work too.

           The command ssh vsli1 'pwd $WIENROOT' print "/home/vlsi"
        the common home directory and
        ssh vlsi1 "env"
        ...
        USER=niel
        PWD=/home/niel
        HOME=/home/niel
        ...
        this is similar as server, and other nodes.

        Sir After changing the parallel option file in $WIENROOT in
        server to

        setenv TASKSET *"yes" from "no"*
        if ( ! $?USE_REMOTE ) setenv USE_REMOTE 1
        if ( ! $?MPI_REMOTE ) setenv MPI_REMOTE 1
        setenv WIEN_GRANULARITY 1
        setenv DELAY 0.1
        setenv SLEEPY 1

setenv WIEN_MPIRUN "mpirun -np _NP_ -machinefile _HOSTS__EXEC_"

        setenv CORES_PER_NODE 1

        the error is not coming but the program is not increasing
        steps after lapw0 it stuck in lapw1


        Since it seemed to be throwing an appropriate error message
        with TASKSET previously unlike when set to "yes", probably
        you should change it back to "no".

_______________________________________________
Wien mailing list
[email protected]
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:  
http://www.mail-archive.com/[email protected]/index.html

Re: [Wien] parallel ssh error

Reply via email to