Missing restarts in C90C.input case, C48f case is dead

[mathomp4]

Questions for @rtodling

C90C and C90C_replay Restart issue

Both @weiyuan-jiang and myself are having issues running the C90C.input test case.

The first issue is that the restart bundle:

FVICS? [/archive/u/jstassi/restarts/GEOSadas-5_24_0]
> /discover/nobackup/projects/gmao/dadev/rtodling/archive/Restarts/x0044_C90/rs/Y2020/M12/x0044_C90.rst.20201215_21z.tar

Doesn't have a gwd_internal_rst file. Now, I can solve that and as a test I made:

/discover/nobackup/mathomp4/x0044_C90.rst.20201215_21z.tar

which is just that tarfile with a c90 gwd_internal_rst

I think C90C_replay.input would have the same issue.

C48 and other dead tests

There are some tests that just can't be run at the moment because FVICS points to dead /archive space:

C48f.input
84:FVICS? [/archive/u/jstassi/restarts/GEOSadas-5_24_0]
85-> /archive/u/jstassi/restarts/GEOSadas-5_24_0

C90C_ens.input
83:FVICS? [/archive/u/jstassi/restarts/GEOSadas-5_24_0]
84-> /archive/u/jstassi/restarts/GEOSadas-5_25_0/C90CS_x0039_p6.rst.20190729_21z.tar

Not sure how to fix these. Maybe NCCS can find them on the tape and try to pull off? I mean for C48f, I could just remap some restarts. But not sure how to handle C90C_ens