error in ERI.f19_g16.B1850

[jlu@coaps_fsu_edu]

The error information is:


forrtl: severe (36): attempt to access non-existent record, unit 92, file /panfs/storage.local/common/inputdata/ocn/pop/gx1v6/ic/ts_PHC2_jan_ic_gx1v6_20090205.ieeer8

Thanks.

[The run was with new CESM1_0_3 version. The file ts_PHC2_jan_ic_gx1v6_20090205.ieeer8 exists in the target directory]

 

[jlu@coaps_fsu_edu]

This error solved by adding
-assume byterecl
in the fortran compiler flags.

 

[yh3019@columbia_edu]

Hi, I encountered the same question here.May I ask how to add this flag "-assume byterecl" in the fortran compiler flags?Thanks a lot!

 

[jedwards]

You would add it to the Macros.* file in your case directory, add it to the line that begins with FFLAGS := note that this is a compiler specific flag and if you are not using the same compiler you will need to find the flag appropriate for the compiler you are using.

 

[yh3019@columbia_edu]

Hi Jedwards,I just checked my files and found that "-assume byterecl" is already in my Macros. But i still have the error:"forrtl: severe (36): attempt to access non-existent record, unit 95, file XXX/inputdata/ocn/pop/gx1v6/ic/ts_PHC2_jan_ic_gx1v6_20090205.ieeer8". Can you tell me what I should do to solve this problem?Thank you!

 

[jedwards]

Two possibilities:

1. The file can't be found from the compute node - is it on a filesystem which is only available on the login nodes?   The fie path provided earlier makes me suspect that that's the case - /panfs/storage.local/ -
2. The file is corrupted - the expected md5 sum is: 44c595e9b1d1340fbc6c680c9200b3f8 If the command md5sum ts_PHC2_jan_ic_gx1v6_20090205.ieeer8does not give this result the file is corrupted and should be replaced - remove the current file and the model should replace it.

 

[yh3019@columbia_edu]

Hi,I checked the two possibilities as you suggested but I think the problem is not related to them. I can find the file in the correct path /work/05383/huangy/stampede2/inputdata/ocn/pop/gx1v6/ic/ts_PHC2_jan_ic_gx1v6_20090205.ieeer8 and the md5 sum is the same as you posted.I deleted the file and that did not help. Do you have any other idea about this?Thank you a lot for your help! Yu

 

[jedwards]

I'm wondering if the "-assume byterecl" is inappropriate for your hardware.   Have you tried compiling without it?

 

[yh3019@columbia_edu]

Hi Jedwards,I finally found that there was actually a different error in my configuration but the model somehow reported this irrelevant problem of .ieeer8 file.Thank you a lot for your help!Yu