Nek5000 / nekRS

our next generation fast and scalable CFD code
https://nek5000.mcs.anl.gov/
Other
290 stars 76 forks source link

Update parReader.cpp #515

Closed jhgoebbert closed 1 year ago

jhgoebbert commented 1 year ago

In the documentation "elapsedTime" is mentioned with a capital T https://nekrsdoc.readthedocs.io/en/latest/input_files.html#general-section

stgeke commented 1 year ago

Did you run into an issue? The keys will be converted to lower case anyway (see makeStringsLowerCase).

jhgoebbert commented 1 year ago

I have this settings

[GENERAL]
polynomialOrder = 9
#startFrom = "restart.fld"
stopAt = elapsedTime
endTime = 300.0
elapsedTime = 5.0

but nekRS stopped with

step= 1863  t= 1.76746035e+02  dt=1.0e-01  C= 4.17  UVW: 17  P: 1  S: 7  elapsedStep= 2.75e-01s  elapsedStepSum= 3.78288e+02s

>>> runtime statistics (step= 1863  elapsed= 600.409s):

I was expecting it to stop much earlier - so had a look at the code and come across this line.

jhgoebbert commented 1 year ago

I changed to

[GENERAL]
polynomialOrder = 9
#startFrom = "restart.fld"
stopAt = elapsedtime
endTime = 300.0
elapsedtime = 5.0

and now it works:

step= 447  t= 4.47000000e+01  dt=1.0e-01  C= 0.00  UVW: 1  P: 1  S: 1  elapsedStep= 1.87e-01s  elapsedStepSum= 8.41537e+01s

>>> runtime statistics (step= 447  elapsed= 300.321s):

=> elapsed= 300.321s = 5.0 * 60

stgeke commented 1 year ago

Hmm if I try this on my end I'll run into a different issue. Are you using vanilla v23?

jhgoebbert commented 1 year ago

No, sorry. I am still on a version from December 2022.

jhgoebbert commented 1 year ago

Hi Stefan, don't you think this fix should be merged?

stgeke commented 1 year ago

I don't think this PR is needed as it seems to work when using v23. If you think there is still an issue, feel free to re-open this PR.