Enable rt.sh dependency relations

[MinsukJi-NOAA]

Issue #103
*Used to create baselines in one step without manual copying of files
*Used for restart rt to get initial data from work directory
*Merge cpld_fv3_mom6_cice_atm_flux_restart_run.IN into cpld_fv3_mom6_cice_atm_flux_run.IN

[DeniseWorthen]

So looking through the tests, it seems like for now we're not implementing the restart test to use the intermediate restarts. Is the plan is first to implement the dependency function and then later move to using the intermediate restarts? At that point, the 3d run tests could be deleted.

[MinsukJi-NOAA]

That's right. The restart run in this PR stays mostly the same as before. The only change is that it is grabbing the restart data from a current run (not the baseline data) whose name is specified as a dependency in rt.conf.

[DeniseWorthen]

I understand the changes in test/fv3_conf, tests/tests and rt.conf and they look right.

[DeniseWorthen]

Is there any reason why the dependency would not work with the 35d optional tests that Bin added?

[MinsukJi-NOAA]

I did not look into the 35d tests. Looks like we just need to add the dependencies in the 6th column of rt_35d_*.conf files.

[DeniseWorthen]

I ran into a problem trying to run the 35d test w/ the dependency on orion. I got
cp: cannot stat '/work/noaa/stmp/jminsuk/RT/FV3-MOM6-CICE5/develop-20200530/WW3_input_data/ww3_multi_35d.inp': No such file or directory

@binli2337 Is there additional input required in the baseline for the 35d test

[MinsukJi-NOAA]

Which test case was that?

[DeniseWorthen]

I tried the rt_35d_cmeps.conf after you added the dependency columns. I merged your rt-dependency branch to my develop branch.

[MinsukJi-NOAA]

I am running rt_35d_cmeps.conf on Orion right now.

[MinsukJi-NOAA]

Dependency implementations in this PR actually does not affect the 35 day runs, because we are neither generating the baselines nor testing the restart capability. Nevertheless, the 35d cmeps cold run were completed successfully. Warm runs did not complete because of the issue here: #111 (comment)

[MinsukJi-NOAA]

This PR is ready for merge. Other 35-day run issues should be addressed separately.