Luca Bertagna
Luca Bertagna
Ah, sorry about that. I thought the dashboard subfolder was the only needed one. Thanks for catching that.
> I don't know how 1.14.3 headers are getting picked up. Maybe from the system? I do see two hdf5 installations in /opt/cray/pe/lib64, which is linked to by the netcdff...
> Luca: if it were trying to use newer hdf from the .so's, then the warning would be flipped -- ie the issue is that it's complaining that headers are...
Downloading the version of this file from the ANL server, and running ncdump -k shows "classic". I think "classic" is ok (gauging from @dqwu message), so maybe we just need...
@ekdejong The v1 tests fail the RUN phase, so something may be amiss with the logic you implemented. ``` BACKTRACE: /home/runner/_work/E3SM/E3SM/components/eamxx/src/share/io/eamxx_output_manager.cpp:432 Error! Output file already exists: - filename: ERS.ne4pg2_ne4pg2.F2010-SCREAMv1.ghci-snl-cpu_gnu.eamxx-prod.C.20251125_214248_yvz7p8.scream.1hourlyAVG_native.h.AVERAGE.nhours_x1.1994-10-07-00000.nc -...
Tbc, I do think we should put checks in buildnml, but a more verbose runtime error explaining what went wrong is already an improvement. After all, even if the yaml...
@ekdejong one last request: rather than have 12 commits, 10 of which are just pairs of "blah" and "Revert blah", can you rewrite the git history and remove those commits,...
So the error happens with a CICE file...but does it only happen with EAMxx compsets? That'd be weird...
> I can happen with non-EAMxx cases, but we rarely (if ever) run with CICE in other E3SM activities 👀 Ok, so if it happens with non-eamxx cases as well,...