This problem is shared by everybody I would say.
Globally, it seems the jobs are now run without issue (explaining the green status), but somehow it is not possible to retreive the data obtained. For example, you may have information about the number of lines and columns of a tabular file obtained as output (meaning the job ran and finished with your results ok), but no display of the content through the "eye" button, and an empty file when trying to download it. And you also get red if you try to use these datasets as input of other jobs, as it will find an empty file also. So at some point the content of the datasets just got lost after the analysis ended, even though it existed at some point (since the metadata about lines/columns do exist).
There are other tickets with the same issue (like this, this and this). Part of these tickets are currently taken in charge by someone from @team.galaxy so we will wait for now.
Best regards,
Mélanie