We had a dxdmchk process running to fetch and clean files on an extended drive...  DX policies were set so that writes were only going to the Isilon and would move and purge immediately...  As files were fetched then they would in turn be moved and stubbed on the Isilon... Doing this would fetch files back from both media groups, but as the majority were located on the Centera at the time the process was started, this was not a big deal.

Due to event outside of DX the VM running DX was rebooted and had only managed to get about half way through the migration.  The process was running for about 4 months before it was killed...  I am now left with the dilemma as to how to limit the fetches to only include the Centera and finish the migration without having to fetch files back from the Isilon and extend my migration window...

Switches allow for us to chose the media based on ID, but there are 1551 pieces of media on the Centera (I know...) So now I am trying to figure out how to best resume fetching and cleaning files back through DX, but limit to the Centera as a source without having to do one media at a time...