[Wien] Fwd: Wien2k in parallel mode_error.

[Gavin Abo]

It looks like the initialization finished okay in single mode as I see 
"init_lapw finished ok" in the image.

The "PBS_NODEFILE: Undefined variable" indicates a problem with your 
parallel setup or an Linux environment setup issue.

There might be a problem with how you setup your jobs file to create the 
.machines file on the fly for mpi parallel.

If you are using w2web, there may be a problem with the script file that 
you have used.

On some systems, there may be a statement that you have to add to your 
job script so that the environment variables are propagated [ 
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15985.html 
].

Clusters for mpi calculations are specialized (i.e., have unique 
hardware and software).  So it is best to talk to your local Information 
Technology expert (administer or Help Desk) for you cluster to have them 
help you setup your job script or w2web execution script.  Your local IT 
expert and you may want to refer to the WIEN2k usersguide as well as the 
following two links:

http://susi.theochem.tuwien.ac.at/reg_user/faq/pbs.html
https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg01658.html

On 4/21/2018 12:54 PM, Ashwani Kumar wrote:
> Hi,
> Created Hafnia 2*2*2 P-type supercell. Called super struct file to 
> StructGen in the interface. Split one atom of Hf from equivalency, 
> changed its Z, Symbol (Ta), saved struct file (wanted to keep dope 
> level ≤1%). Initialization executed, hangs at NN program (for 2 days), 
> stop with the error.  Redone supercell formation. .machine file was 
> prepared for parallel execution. Initialization successful but dstart 
> ran in single mode.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20180421/276fa698/attachment.html>