No master input files. input dataset is empty


I am trying to submit a job to the grid which runs over a user-produced input container, and my job goes into “broken” state with the message “Logged status: no master input files. input dataset is empty”.

Here is the panda job:

The container is user.jburzyns.313415.PowhegPythia8EvtGen_ZH_H125_a16a16_4b_ctau100.recon.AOD.e8323_e5984_s3227_r12627_r12347_StreamAOD

If I run rucio list-files on this container, it returns the full list of files in the container as expected. I can download the files and everything looks normal. But for some reason when trying to process these datasets on the grid it cannot find them. What is strange is that the job clearly recognizes that the container exists, as it shows the correct datasets in the input list (see screenshot [1]). If I select one of these datasets though, it shows that there are no files, and the creation date is the time of submission of my attempted job (see screenshot [2]). If I look at the output of the job that produced this input, I can clearly see that the datasets are nonempty (see screenshot [3]). The only difference here is that the dataset has the additional scope clarifier user.jburzyns at the beginning, but I did not explicitly specify that in the submission. If anyone has any ideas as to what might be happening here I would greatly appreciate it!


1 Like

Hi Jackson,

I had a similar issue a few weeks ago (link), but I’m not sure if it is exactly the same. How did you submit your job? Can you share the submission command/script? Did you try adding the user.jburzyns: scope to the input dataset when you submitted? Did you somehow filter out the input files in the input dataset as I did in my case?