Run a test case on blue waters

[anshchaube]

This issue can be closed when a Nek job is successfully executed on Blue Waters, and the data from the run has been transferred and visualized.

[dshaver-ANL]

This doesn’t look bad actually, just not fully developed. Can you run it longer? You would at least need to reach time = L/U.

-Dillon

From: Ansh Chaube [mailto:notifications@github.com] Sent: Friday, June 21, 2019 11:21 AM To: arfc/sneks4me Cc: Subscribed Subject: Re: [arfc/sneks4me] Run a test case on blue waters (#3)

Same test case ran on desktop and yielded good results. But results from Blue Waters look unphysical:

[visit0000]https://user-images.githubusercontent.com/31393012/59936823-8f8bb580-9416-11e9-968d-f5f936bdce69.png

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/arfc/sneks4me/issues/3?email_source=notifications&email_token=AE3653AGQG4XXE7ESLYUFDLP3T5VJA5CNFSM4HZ42N2KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODYI5KVA#issuecomment-504485204, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AE3653GI6656GE4G323NRVDP3T5VJANCNFSM4HZ42N2A.

[anshchaube]

I recently switched to Ubuntu 18, and I'm having trouble with visit. When I export the results, they look fine(which I realized after posting on Github). But when I'm viewing them in visit, I see big blocks in the middle of the channel where the velocity is zero. Currently trying to find an appropriate version of visit.

[anshchaube]

Data looks reasonable (although the boundary layer velocity seems to be zero). The problem was with VisIt. I think I can move on to #5 and start submitting bigger jobs that run for longer and close this issue @dshaver-ANL