[Nek5000-users] How to read NEK data generated on another machine (with different endianness)

[nek5000-users at lists.mcs.anl.gov]

Dear Ambrish,


Thank you for the files.


I find the logfile (your .out file) a bit confusing.

It looks like it starts time stepping prior to loading

the restart file ??


(Just some notes.  I see you're running on P=1024 ranks.

You have nsteps=1.  What happens if you have nsteps=5, say?)


When I grep uvw from the logfile I see everything is zero.


Why does the code timestep prior to loading?  Are you doing something

in the user file?


Finally, I urge you to update to the latest version of the code

from the webpage (the v17.0 release).    The team has put a lot

of effort into having this as the stable and supportable release.

Although I have no reason to suspect issues with the version

you are using, we won't be able to debug it---we are all working

only with the current version at this point.


Best,

Paul


________________________________
From: Nek5000-users <nek5000-users-bounces at lists.mcs.anl.gov> on behalf of nek5000-users at lists.mcs.anl.gov <nek5000-users at lists.mcs.anl.gov>
Sent: Friday, January 5, 2018 9:25:13 AM
To: nek5000-users at lists.mcs.anl.gov
Subject: Re: [Nek5000-users] How to read NEK data generated on another machine (with different endianness)


Dear Paul,


Thanks a lot for your prompt response.


I am attaching the error and output files for the job, which did not work. Could you please have a look into it, and suggest something?


When we restart on Juqueen, everything works fine.


Moreover, we are using version 1.0rc1 / SVN r1115 on SuperMUC, whereas 1.0rc1 / SVN r1088 on Juqueen. I don't know, but might this be causing some problem?


Thanks and best regards,

Ambrish



________________________________

Dear Ambrish,


Nek5000 supports either endian conditions and one

can in principle go back and forth.


What does the error look like?


I'm assuming that the issue does not arise when you

restart on juqueen?


Paul



With best regards,
Ambrish

------------------------------------------------------
Ambrish Pandey
Post Doctoral Researcher
Institut für Thermo- und Fluiddynamik
Technische Universität Ilmenau
Ilmenau 98693
Germany
Emails: ambrish.pandey at tu-ilmenau.de
ambrishiitk at gmail.com
------------------------------------------------------
________________________________
From: Pandey Ambrish TU Ilmenau
Sent: Thursday, January 4, 2018 1:13:31 PM
To: nek5000-users at lists.mcs.anl.gov
Subject: How to read NEK data generated on another machine (with different endianness)


Dear All,


We have generated some data on JUQUEEN (Julich Supercomputing Centre), and want to analyze them on SuperMUC (Leibniz Supercomputing Centre). But NEK (1.0rc1 / SVN  r1115) on SuperMUC can't process the data. It is giving some error after reading the file.


Original data is in big endian format, and SuperMUC generates data with little endian format. Could this be a problem? If so, how can we tackle this issue?


Thanks,

With best regards,
Ambrish

------------------------------------------------------
Ambrish Pandey
Post Doctoral Researcher
Institut für Thermo- und Fluiddynamik
Technische Universität Ilmenau
Ilmenau 98693
Germany
Emails: ambrish.pandey at tu-ilmenau.de
ambrishiitk at gmail.com
------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20180105/33758591/attachment.html>