CIME error in wave watch archiving

[rljacob]

CIME error in wave watch archiving

[rljacob]

@jgfouca your name is on those lines.

[jgfouca]

@rljacob , strange code that I don't remember writing. Git blame does show me as the last person who touched these lines. The associated commit is:

Update CIME to ESMCI cime5.7.0 (PR #2414)
    
    Squash merge of jgfouca/branch-for-to-acme-2018-06-28
    
    Features:
    
    * Allow users to go up through the setup phase on non-local machines
    * Documentation updates
    
    Bug fixes:
    
    * fix issue with e3sm mali file archiving
    * Use ConfigParser rather than SafeConfigParser
    * Fix for passing run arguments in LSF
    
    Fixes #2393
    Fixes #2390
    
    [BFB]

So this commit was just a bulk of changes being synced, not stuff really authored by me.

It's possible @jedwards4b knows the reason for this.

[rljacob]

Oh I forgot we can't trust gitblame during the subtree era.

[jgfouca]

Yeah, the subtree era mucked-up the git histories of both E3SM and CIME.

[jgfouca]

Thinking about it some more, it seems like this bullet must be the reason:

* fix issue with e3sm mali file archiving

[github-actions]

This issue is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 5 days.

[jgfouca]

@jedwards4b , any recollection of the "ww" stuff?

[jedwards4b]

We have ww3 and wwdev and this is intended to match either. I don't think that there is an archiving issue with ww3 in cesm.

[jgfouca]

Thanks, @jedwards4b . So this is likely an issue with an E3SM XML file. I've assigned @jasonb5

[jasonb5]

1173    for comp in list(components):
1174        compname = self.get_value("COMP_{}".format(comp))
...
1186         comp_expected = archive.get(comp_archive_spec, "compname")
1187         if comp_expected == "ww3":
1188             comp_expected = "ww"
...
1201            expect(
1202                comp_expected in comps_in_case,
1203               "env_archive defines component {} not defined in case".format(

After some investigating I found a CESM test (PET.f19_g17_rx1.2000_DATM%NYF_SLND_CICE5_POP2_DROF%NYF_SGLC_WW3) that I could trace to understand what's happening. When this code runs for CESM the case's compname is ww and the archive compname is ww3 so the code in question handles that difference. For E3SM we don't need this between the compnames match.

I'll add a comment for the future and bypass this for E3SM.

[jgfouca]

Thanks, @jasonb5 !

[rljacob]

When this code runs for CESM the case's compname is ww and the archive compname is ww3
@jedwards4b can that difference be removed so we don't need this code at all?