Hsin-Fang Chiang Your plan sounds good. Thanks for tracing through all of this. This re-organization will very much help make the file system tree provide more information to future users.

Great!

With Greg Daues's help, /project/wmwv/nidever/ now has everything as originally transferred from /lsst8/decam/data on NFS. Michael Wood-Vasey has the full ownership of the /project/wmwv/ folder.

I'll go ahead and reorganize /datasets/decam as planned. Will do DM-11381 at the same time.

Files have been moved to /datasets/decam/_internal as proposed (my previous comment on Jul 26 and a slight change to satisfy RFC-308). Unneeded files such as rerun attempts and non-fits files have been removed as proposed.

When I tried to remove the duplicates in Dey's COSMOS folders (20130209, 20130210, 20130211, 20130212), I found all 136 visits that have 'COSMOS' in its OBJECT header key are duplicates and were already added by RFC-263. Other files in those folders seem to be raw calibration files, test exposures, or other non-COSMOS objects. Therefore, I propose to simply remove all of these folders (20130209, 20130210, 20130211, and 20130212) from /datasets/decam/_internal/raw/ if there are no objections.

As I see no objections, I'm removing the Dey's COSMOS folders (named 20130209, 20130210, 20130211, 20130212) from {[/datasets/decam/}}. The COSMOS data from RFC-263 stay in /datasets/decam/_internal/raw/cosmos/February2013/. Raw data from this ticket are /datasets/decam/_internal/raw/allen_neo, /datasets/decam/_internal/raw/nidever_smash, and /datasets/decam/_internal/raw/saha_bulge. They are ingested into /datasets/decam butler repo using the stack w_2017_33.

Data have been reorganized as proposed; raw are ingested.