[Ncep.list.nems.announce] [Nems.tickets] NEMS trunk FREEZE and Re: NEMS Ticket #161: Unifying merge of branches/NUOPC/development and NEMS trunk
Gerhard Theurich
theurich at sourcespring.net
Thu Apr 14 15:55:34 UTC 2016
Hi Ratko, I totally see your point, it is always best to have most
cahnges in before starting. My only concern is time. I want to make sure
we have enough time to work out the issues. There isn't only the svn
based merge (with plenty of manual intervention as I foresee it), but
then there is also the whole unification on the actual code (removal of
#ifdefs). At this stage we will need to get the current "trunk mode"
working under the unified NUOPC-based EARTH component. I was hoping to
get to working on that maybe even by end of this week.
On the other hand, I am still cleaning up compsets as per a call we had
on Tuesday. It is taking me longer than I thought, so even the
NUOPC/development branch isn't quite ready yet.
If you don't mind, could you still just make copies of the
REGRESSION_TEST dirs into REGRESSION_TEST_r74057? Just in case we do
decide to start a first attempt of the merge work with that trunk
revision? Thanks, and I appreciate your assistance during the process
very much!
-Gerhard
On 04/13/2016 06:50 PM, Ratko Vasic wrote:
> Gerhard,
> I'd wait until weekend (before freeze).
> There is one commit in the line (GSM stuff), and as agreed on today's
> meeting with NMMB group, we'll update trunk with changes (at least 90%
> of changes) going into next production.
> GSM commit is adding one more regression test, and NMMB commit will for
> sure change NMMB results.
> I would like to tag that version as a reference point.
> I'll test everything on both machines (theia and wcoss).
>
> Ratko
>
>
>
> On 4/13/2016 5:39 PM, Gerhard Theurich wrote:
>> Hi Jun,
>>
>> Thank you very much! That is very helpful!
>>
>> Unless I hear otherwise let's declare NEMS -r74057 / GSM -r73903 to be
>> the trunkRev1.
>>
>> Ratko, could you please make copies of your current
>> .../Ratko.Vasic/REGRESSION_TEST dirs on wcoss and Theia to
>> .../Ratko.Vasic/REGRESSION_TEST_r74057. This will allow us to work
>> with trunkRev1 for the next few days without loosing the ability to
>> validate the trunk regression tests - just in case the trunk keeps
>> moving and requires changes to the baselines.
>>
>> Patrick and/or Dusan, will you be available to run NMMB trunk
>> regression tests on wcoss and Theia when we get to that point?
>>
>> -Gerhard
>>
>> On 04/13/2016 02:24 PM, Jun Wang - NOAA Affiliate wrote:
>>> Gerhard,
>>>
>>> For gsm side, I think the current trunk version (nems -r74057, gsm
>>> -r73903) is a stable version. Maybe someone from NMMB can confirm that
>>> for nmmb. If both gsm and nbb confirm the version, I think you can
>>> create nems branch trunkRev1. Dusan just committed the code, so
>>> currently all the regression tests for both GSM and NMMB for this
>>> version passed on both wcoss and theia. Please see the log files at
>>> trunk/tests/RegressionTests_wcoss.log and
>>> trunk/tests/RegressionTests_theia.log. The baseline from the trunk
>>> version above are saved in Ratko's directory on both theia and wcoss:
>>>
>>> wcoss:
>>> /meso/noscrub/Ratko.Vasic/REGRESSION_TEST
>>> Theia:
>>> /scratch4/NCEPDEV/meso/noscrub/Ratko.Vasic/REGRESSION_TEST
>>>
>>> I can help test gsm regression test once you have a stable merged
>>> version. Thanks.
>>>
>>> Jun
>>>
>>>
>>>
>>> On Wed, Apr 13, 2016 at 4:52 PM, Gerhard Theurich
>>> <theurich at sourcespring.net <mailto:theurich at sourcespring.net>> wrote:
>>>
>>> Hi NEMS developers,
>>>
>>> I just opened a new NEMS ticket (#161) to serve as a reference
>>> for the
>>> upcoming merge task. This more than a simple merge in that we are
>>> trying
>>> to unify the coupled system with the default "trunk mode" of
>>> NEMS. The
>>> result will be much cleaner superstructure code (removal of
>>> #ifdefs in
>>> the EARTH component). Also the emerging system will exclusively
>>> use the
>>> latest ESMF release (v7.0.0).
>>>
>>> As previously discussed, there is an upcoming FREEZE of the NEMS
>>> trunk
>>> for the entire next week (i.e. week of 04/18/16). We will try to
>>> have an
>>> early test candidate of the new NEMS trunk available some time
>>> middle of
>>> next week.
>>>
>>> There are some critical pieces that I need help with from EMC:
>>>
>>> 1) Prior to the trunk freeze next week, I would need someone to
>>> identify
>>> a recent and stable NEMS trunk revision. Let's call this revision
>>> trunkRev1
>>>
>>> 2) I will need someone (or a few folks) that are well versed at
>>> running
>>> the NEMS trunk regression tests. I am thinking it would be good
>>> to have
>>> GSM and NMMB experts available for this. They would run the
>>> regression
>>> tests on trunkRev1 and make a backup copy of the results, so that
>>> we can
>>> retest against the same baseline, even if the default baseline on
>>> Theia
>>> and WCOSS may still change (because trunk is not yet frozen).
>>>
>>> 3) During the unification process I am expecting to have versions of
>>> NEMS available at different times. It would be great of the same
>>> people
>>> from #2 were available to check out these revisions and test them
>>> again
>>> against the saved baseline from #2.
>>>
>>> Thanks,
>>> -Gerhard
>>>
>>>
>>>
>>> On 04/13/2016 01:26 PM, NEMS Trac Ticket wrote:
>>> > #161: Unifying merge of branches/NUOPC/development and NEMS trunk
>>> > -------------------------+-----------------
>>> > Reporter: theurich@… | Owner:
>>> > Type: task | Status: new
>>> > Priority: major | Milestone:
>>> > Component: NEMS | Version: 3.1
>>> > Resolution: | Keywords:
>>> > -------------------------+-----------------
>>> > Changes (by theurich@…):
>>> >
>>> > * type: defect => task
>>> >
>>> >
>>> > --
>>> > Ticket URL:
>>> <https://svnemc.ncep.noaa.gov/trac/nems/ticket/161#comment:1>
>>> > NEMS <https://svnemc.ncep.noaa.gov/trac/nems>
>>> > My example project
>>> >
>>> _______________________________________________
>>> Ncep.list.emc.nems.tickets mailing list
>>> Ncep.list.emc.nems.tickets at lstsrv.ncep.noaa.gov
>>> <mailto:Ncep.list.emc.nems.tickets at lstsrv.ncep.noaa.gov>
>>> https://www.lstsrv.ncep.noaa.gov/mailman/listinfo/ncep.list.emc.nems.tickets
>>>
>>>
>>>
>
More information about the Ncep.list.nems.announce
mailing list