[Wien] lapw0 problem after supercell

[=?iso-2022-jp?B?GyRCMiZaXxsoQg==?=]

Dear Stefaan:

  I *have checked* the archives, that was why I also put the limit result on
email. I was not sure then this was the problem of stack size and could not
change the setup of limit by myself. Please do not doubt my reliability. I
will not intentionally waste anyone’s time. 

  Warm regards

 

 

Dear Peter:

  Thank you very much for your patience. The stack size had been changed to
2097152 kbytes. It is ok now for lapw0. But after this, scf stop
automatically. Following is the output:

Lapw1.error:

**  Error in Parallel LAPW1

**  LAPW1 STOPPED at Tue May 9 19:04:58 JST 2006

**  check ERROR FILES!

Error in LAPW1

Error in LAPW1

Error in LAPW1

Error in LAPW1

Error in LAPW1

 

Dayfile is:

on reed with PID 16555

 

    start       (Tue May  9 19:03:41 JST 2006) with lapw0 (40/20 to go)

 

    cycle 1     (Tue May  9 19:03:41 JST 2006)  (40/20 to go)

 

>   lapw0 -p    (19:03:41) starting parallel lapw0 at Tue May  9 19:03:41
JST 2006

--------

running lapw0 in single mode

52.642u 15.747s 1:08.39 99.9%   0+0k 0+0io 1498pf+0w

>   lapw1  -c -p        (19:04:49) starting parallel lapw1 at Tue May  9 19:
04:49 JST 2006

->  starting parallel LAPW1 jobs at Tue May  9 19:04:49 JST 2006

running LAPW1 in parallel mode (using .machines)

4 number_of_parallel_jobs

**  LAPW1 crashed!

0.125u 0.200s 0:09.04 3.5%      0+0k 0+0io 12003pf+0w

 

*          stop error

 

 

and lapw1_*.error have contents:

Error in LAPW1

 

Also gaas.scf1_* and gaas.output1_* exist.

I think there is nothing to do with NFS and mpi ( not in use) because there
is no problem for other materials. Here I increased NUME to 3000. 

 

What is wrong?

 

Thank you very much

 

Best wishes

 

Hao

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20060509/e75f71d7/attachment.html