The rsync Data Manager has been problematic. It will very likely be replaced near term (still in progress).
For now, I suggest fetching fasta datasets directly from the external source and creating your own indexes from there. Fetch the genome, index for Sam (allow to complete), then Picard (allow to complete), index for 2bit (can be run concurrent and no downstream indexes are dependent), and then continue with any other indexes associated with the tools that will be in use.
Thanks, Jen, Galaxy team