Hi teachers,
I submitted job to run smk_edgar_HEMI108k.csh but it stuck in TEMPORAL program and no error returned in log files. It just run without stop for hours and not into SMKMERGE program. I checked invtable, gsref and gspro files and didnt find any issues.
temporal.edgar.HEMI_108k.log:
It could be an out-of-space issue, either related to lack of RAM or disk space. If the you delete the intermediate files and rerun temporal, does it get hung up at the same exact timestep?
Thank you! I changed serial ./smk_edgar.csh to mpi run it to get more RAM and it didnt stuck any more. But but now I want to run multiple SMOKE programs simultaneously (with each program processing a different source) to improve efficiency. I copied the original SMOKE folder and then started running several SMOKE programs separately. However, when running them at the same time, some of the programs always terminate unexpectedly, and in the end, only one program remains running. What could be causing this issue?
Hi alison,
Two weeks ago, I solved the issue of getting stuck at the ‘temporal’ step by increasing the number of nodes to get more RAM, and it has been running smoothly since then. However, last night, I suddenly encountered this issue again. This time, I found that increasing the number of nodes no longer helps to make it run successfully. I’m wondering if the failure is due to the fact that I changed the location of the intermediate files atsup and atmp. However, the intermediate files are too large for me to generate them in the default path. I would really appreciate your advice.
What sector are you trying to run and how large is the inventory you are trying to process?
Is this your own inventory or one of the EPA inventories?
The inventory I used is the latest version of EDGAR and nine pollutants were put for inputs as below where each one is around 500 MB. However, I changed atsup file output path just now and it worked again this time. Something like the problem from path of intermediate file set? I felt so confused.