You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Expected behavior
User should not have to clean up $ROTDIR before running getic again after it fails
Current behavior
After I ran the getic task and it failed (see issue #1155), the job was resubmitted. The job gave an error when running the following statement.
Job getic task successfully completed on first attempt. It is recommended only rerun the failed job as needed.
The job getic.sh should be updated with exception handlining in place to fail the job if the target directory is not empty.
Will provide a PR for the fix.
If getic is run with an existing `$ROTDIR` (possibly because it failed the first time), it would fail because the directory already exists. Now any existing directory is deleted before attempting to move the ICs.
Fixes#1156
Expected behavior
User should not have to clean up $ROTDIR before running getic again after it fails
Current behavior
After I ran the getic task and it failed (see issue #1155), the job was resubmitted. The job gave an error when running the following statement.
mv ${EXTRACT_DIR}/${CDUMP}.${yy}${mm}${dd}/${hh}/* ${ROTDIR}/${CDUMP}.${yy}${mm}${dd}/${hh}
mv: cannot move <EXTRACT_DIR>/gfs.20221115/00/* to <ROT_DIR>/gfs.20221115/00/atmos': Directory not empty
Machines affected
All
To Reproduce
run getic.sh before fixing it for issue #1155 and then resubmit job
The text was updated successfully, but these errors were encountered: