[Ncep.list.nems.announce] NEMS Upcoming commits: NEMS Ticket #174 & GSM #102: to the NEMS, GSM, NMM, phys, NEMSmodulefiles and chem trunks
Eugene Mirvis - NOAA Affiliate
eugene.mirvis at noaa.gov
Fri Aug 12 13:53:25 UTC 2016
Hello,
Thanks Tom's updates on nmmb trunk.
Now NEMSLegacy trunk has been changed accordingly and all externals now
pointing to the components trunks:
NEMS
https://svnemc.ncep.noaa.gov/projects/nems/trunk
NEMS/src/atmos/gsm -r 80392
https://svnemc.ncep.noaa.gov/projects/gsm/trunk
NEMS/src/atmos/nmm -r 80429
https://svnemc.ncep.noaa.gov/projects/nmmb/trunk
NEMS/src/chem -r 80344
https://svnemc.ncep.noaa.gov/projects/nceplibs/chem/trunk
That sould conclude the commits of NEMS ticket #174 and GSM #102
Thanks to all,
-Eugene
On Thu, Aug 11, 2016 at 1:42 PM, Eugene Mirvis - NOAA Affiliate <
eugene.mirvis at noaa.gov> wrote:
> Hello NEMS developers,
>
> Committed on the NEMS trunk (see NEMS ticket #174 and GSM ticket # 102)
> All test have been passed on WCOSS and Theia.
>
> Please synch all nems branches accordingly. NMMB team cooperation is
> needed:
>
> The branch:
> https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nmmbtrunk_MD
>
> needed to be reintegrated into NMMB trunk.
> Dusan and Tom kindly agreed on that during preliminary conversations -
> it's only 1 little change in nmmb makefile - pointing on PHYS_INC location,
> while build within NEMS.
> Please contact me right away if you'll have any questions.
> Without that little changes nmmb build in NEMS wouldn't work. Thanks!
>
> I will update NEMSLegacy trunk after NMMB will inform on fixing their
> trunk!
>
> Thank you all for your cooperation.
>
> Cheers,
>
> -Eugene
>
>
>
> -------- README ------------------------------
> -----------------------------------------------------------------------
> by E.Mirvis and M.Iredell
>
> Motivation
> There are competing forces for pulling physics out of the models into its
> own library and for pushing physics into each model directory. For a
> separate physics library are the forces to share physics across models, to
> run physics within a single column model, multi- physics ensembles, to
> develop a new Common Community Physics Package (CCPP), to use a shared
> physics driver, and to enable the outside community to readily contribute
> physics schemes. For model-specific physics are the forces to more easily
> develop code in one environment, to remove requirements for testing physics
> for many models, and to tune physics for one model. The following is a
> proposed compromise solution that may meet the forces halfway and satisfy
> all parties.
>
> Outline of solution
>
> Use the phys subproject under nceplibs project in subversion.
> However, phys would have a less flat structure than other nceplibs
> subprojects.
> It would still have trunk, tags, and branches, but it would also have
> a Components subdirectory containing physics from individual models
> (components), plus a shared_phys directory used by more than 1 model.
> On the next page is a template for how it might look like.
> Note it has a separate subdirectory for gsm_phys under the Components.
> Under the gsm trunk would be externals to the gsm_phys trunk and the
> shared_phys trunk.
> Releases of gsm are tags which point to tags in gsm_phys and
> shared_phys. These tags may have the same name and script can be easily
> automated.
> The idea is that the gsm code manager would manage the gsm_phys
> subproject too.
> Commits to gsm and gsm_phys trunks will be done at the same time under
> the same review, because of the external properties.
> The phys subproject manager would manage the non-component phys
> routhines as well as to coordinate with component code managers commits at
> the shared_phys trunks.
> New physics in gsm_phys trunk, being already tested, would
> automatically be externals in phys trunk.
> As usual, the development of the components’ phys will be done on the
> branches, tagged separately, sync with local trunks and integrated phys
> trunk will be updated based on that.
> If any development by phys group of the components is required, phys
> code manager should go through the Components code manager approval.
> The shared_phys directory may contain the shared community physics
> driver.
> Releases of the phys library are some planned tags and built as
> libraries under nceplibs production and research locations.
> Unit tests might be established for individual physics entry points.
> Consider also building gsm_phys+shared_phys as a released library.
>
>
> `--phys
> | `--trunk
> | | `--gsmphys (external to phys/Components/gsmphys/trunk)
> | | `--nmmphys (external to phys/Components/nmmphys/trunk)
> | `--tags
> | `--branches
> | `--Components
> | | `--gsmphys (owned by gsm group)
> | | | `--branches
> | | | `--tags
> | | | `--trunk
> | | `--additional-model-unique-phys
> | | | `--branches
> | | | `--tags
> | | | `--trunk
> …
> …
> | | `--shared_phys
> | | | `--branches
> | | | `--tags
> | | | `--trunk
>
> `--gsm
> | `--trunk
> | |… `--gsmphys (external to phys/Components/gsmhys/trunk)
>
> `--nmm
> | `--trunk
> | |… `--nmmphys (external to phys/Components/nmmphys/trunk)
> So the developer does not need to do multiple tickets or multiple sets of
> tests or multiple submits for review. The added work for working on
> multiple projects is just the re-pointing of the external. The code manager
> also does not have much added work except the re-pointing back of the
> external.
> During the code manager’s procedure, at no time should the trunk not work.
>
> Actual code is located now under gsm/gsmphys and nmm/nmmphys project and
> nceplibs/phys -
> unified has the external to both physiscs and nems/src/conf for
> consistency of the environment
> comment:11 Changed 2 hours ago by eugene.mirvis@…
>
> New NEMS features: a standalone integrated pre-build Phys and Chem
> libraries. ==
>
> The current commit provides a standalone capabilities for NEMS
> isolated physics and chemistry separately.
> The phys and chem library sources and build system will be located
> now as the sub-projects of nceplibs in EMC repository, inheriting all NEMS
> development history as a regular nceplibs sub-projects, contain branches,
> tags and the trunks, tagged currently as v10.0.10, accordingly of the NEMS
> restructuring history.
> Both: phys and chem are using the same configuration NEMS
> procedure, therefore, each has NEMS conf directory as the external links.
> That secures consistency of configurations between phys, chem, and NEMS
> framework.
> A dual built capability is provided for the both phys and chem
> libraries:1) as the stand-alone installations 2) the builts within a
> traditional NEMS structure for phys and chem development process.
> That selection is provided via the two external parameters:
> $PHYS_MODE and $CHEM_MODE.
>
> If these parameters are set as “compile ”the built system assumes full
> compilation if such library. In this case a directory chem source should
> have an external link to chem at the “src” while the phys dir an external
> link at src/atmos level, defined at the subversion properties at NEMSLegacy
> externals. By default, both will be set to “compile” by now, as currently
> agreed with component code manager and will be assigned and are located in
> configuration files in conf directory for Theia and WCOSS.
>
> If options to use of the pre-built libraries on operational and R&D
> EMC machines are needed to be exercised, $ PHYS_MODE and/or $CHEM_MODE the
> value of these parameters must be changed to any other than “compile”. In
> this case, NEMS’s makefile will be acting accordingly to use pre-built
> libraries on each machine and the source code within NEMS structure is not
> necessary to be checked out.
> The pre-built PHYS and CHEM libraries locations
>
> The pre-built libraries are available on Theia & WCOSS at the following
> locations:
> On Theia:
> /scratch3/NSEPDEV/nceplibs/dev/
>
> phys, chem, phys_full and chem_full
>
> The builds of phys and chem contain the pre-builds with an option of
> GOCART_MODE=stub while phys_full and chem_full contain pre-builds with the
> GOCART_MODE=full.
>
> On WCOSS, the locations of the similar builts are at:
> /usrx/local/mceplibs/dev/
>
> phys, chem, phys_full and chem_full consecutively.
>
> --------------------------
> Development procedure for GSM developer who needs to change both gsm and
> gsm_phys
> (assume all this is done during a week when no other trunk commits are
> done)
>
> Developer creates ticket
> Developer copies gsm_phys trunk to gsm_phys branch
> Developer copies gsm trunk to gsm branch
> Developer points gsm_phys external in gsm branch to gsm_phys branch
> Developer checks out gsm branch
> Developer makes changes, tests, and passes branches to GSM group for
> review
> GSM group does review
> Code manager merge-reintegrates gsm branch to gsm trunk
> Code manager merge-reintegrates gsm_phys branch to gsm_phys trunk
> Code manager points gsm_phys external in gsm trunk to gsm_phys trunk
>
> If options to use of the pre-built libraries on operational and R&D
> EMC machines are needed to be exercised, $ PHYS_MODE and/or $CHEM_MODE the
> value of these parameters must be changed to any other than “compile”. In
> this case, NEMS’s makefile will be acting accordingly to use pre-built
> libraries on each machine and the source code within NEMS structure is not
> necessary to be checked out.
> The pre-built PHYS and CHEM libraries locations
>
> The pre-built libraries are available on Theia & WCOSS at the following
> locations:
> On Theia:
> /scratch3/NSEPDEV/nceplibs/dev/
>
> phys, chem, phys_full and chem_full
>
> The builds of phys and chem contain the pre-builds with an option of
> GOCART_MODE=stub while phys_full and chem_full contain pre-builds with the
> GOCART_MODE=full.
>
> On WCOSS, the locations of the similar builts are at:
> /usrx/local/mceplibs/dev/
>
> phys, chem, phys_full and chem_full consecutively.
>
> You can link against each of these libraries directly but you have to
> change the values of PHYS_MODE and CHEM_MODE from the “compile” in the NEMS
> conf/configure.nems.Theia* and conf/configure.nems.WCOSS* files. That will
> automatically assume in the NEMS makefile the pre-built location paths are
> activated on each site.
>
> The chem and phys modulefiles
>
> In order to conveniently manipulate with different options of use, for
> installation procedure and for regression testing; and also to support the
> NCEPLIBS computing environment flexibility and version standards, phys and
> chem modulefiles have been created. These modulefiles are also residing
> under EMC repositiory as the NEMSmodulefiles sub-project of the nceplibs
> project and are also available on Theia and WCOSS currently under:
> /scratch3/NCEPDEV/nceplibs/dev/NEMS modulefiles,
> and
> /usrx/local/nceplibs/dev/NEMS modulefiles, consecutively, until an
> operational implementation of the phys and chem @ nwprod will occure.
>
> Phys and Chem installation and version control procedures
>
> The installation of the phys and chem are simplified to :
> a) Checking out need version, dev branch, tag or the phys trunk of the
> development
> b) gmake clean
> c) running ./Install
>
> Actual changes on the NEMS trunk:
>
> Theia:/scratch4/NCEPDEV/nems/noscrub/Eugene.Mirvis/TRUNK_BEFORE_COMMIT_N174>svn
> commit -FStandalone_Readme
> Sending .
> Sending src
> Sending src/atmos
> Sending src/atmos/makefile
> Deleting src/atmos/phys
> Deleting src/atmos/share/idea_composition.f
> Sending src/atmos/share/makefile
> Deleting src/atmos/share/mersenne_twister.f
> Deleting src/chem
> Sending src/conf/configure.nems.Theia.intel
> Sending src/conf/configure.nems.Theia.intel_gsm
> Sending src/conf/configure.nems.Theia.intel_nmmb
> Sending src/conf/configure.nems.Wcoss.intel_gsm
> Sending src/conf/configure.nems.Wcoss.intel_nmmb
> Adding src/conf/modules.nems.Theia_ESMF_700_gsm
> Adding src/conf/modules.nems.Theia_ESMF_700_nmmb
> Deleting src/conf/modules.nems.Theia_ESMF_7_gsm
> Deleting src/conf/modules.nems.Theia_ESMF_7_nmmb
> Sending src/conf/modules.nems.Wcoss_ESMF_700_gsm
> Sending src/conf/modules.nems.Wcoss_ESMF_700_nmmb
> Sending src/configure
> Adding src/detect_machine.sh
> Adding src/detect_machine1.sh
> Sending src/makefile
> Sending tests
> Sending tests/Compile_theia.log
> Sending tests/Compile_wcoss.log
> Sending tests/RegressionTests_theia.log
> Sending tests/RegressionTests_wcoss.log
> Sending tests/tests/gfs_slg_stochy
> Sending tests/tests/nmm_2way_nests
> Sending tests/tests/nmm_2way_nests_debug
> Sending tests/tests/nmm_cntrl
> Sending tests/tests/nmm_mvg_nests_gens
> Sending tests/tests/nmm_nests
> Sending tests/tests/nmm_reg_filt
> Sending tests/tests/nmm_reg_filt_zombie
> Sending tests/tests/nmm_reg_hur
> Sending tests/tests/nmm_reg_rst_nio_debug
> Sending tests/tests/nmm_reg_timesr
> Transmitting file data ..........................done
>
> On Tue, Aug 9, 2016 at 7:30 PM, Eugene Mirvis - NOAA Affiliate <
> eugene.mirvis at noaa.gov> wrote:
>
>> Greetings,
>>
>> At this moment, NEMS components' gatekeepers have completed the order of
>> previous planned commits on NEMS components trunks and after some machines
>> got fixed now too, I'd like to finalize a commit of the isolated atmos
>> physics and chem (gocart), described in previous e-mail (see below), NEMS's
>> ticket #174 and GSM ticket #102.
>> :
>> After sycned of my branches with latest commits, and all good RTs on 2
>> machines, the reintegration process with NEMS trunk seems to have no
>> problem at this step (see a dry-run below):
>>
>> [Eugene.Mirvis at g10a2 nemstrunk]$ svn merge --dry-run --reintegrate
>> https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nem
>> strunk_NEMSLegacyMD_ph2
>> --- Merging differences between repository URLs into '.':
>> U tests/Compile_theia.log
>> U tests/Compile_wcoss.log
>> U tests/tests/nmm_mvg_nests_gens
>> U tests/tests/nmm_cntrl
>> U tests/tests/nmm_reg_timesr
>> U tests/tests/gfs_slg_stochy
>> U tests/tests/nmm_reg_rst_nio_debug
>> U tests/tests/nmm_reg_hur
>> U tests/tests/nmm_reg_filt
>> U tests/tests/nmm_nests
>> U tests/tests/nmm_2way_nests
>> U tests/tests/nmm_reg_filt_zombie
>> U tests/tests/nmm_2way_nests_debug
>> U tests/RegressionTests_theia.log
>> U tests/RegressionTests_wcoss.log
>> U tests
>> U src/configure
>> U src/atmos/share/makefile
>> D src/chem
>> D src/atmos/share/idea_composition.f
>> D src/atmos/phys
>> D src/atmos/share/mersenne_twister.f
>> U src/atmos/makefile
>> U src/atmos
>> U src/conf/modules.nems.Wcoss_ESMF_700_nmmb
>> U src/conf/configure.nems.Theia.intel_gsm
>> U src/conf/configure.nems.Theia.intel_nmmb
>> U src/conf/configure.nems.Wcoss.intel_gsm
>> U src/conf/configure.nems.Wcoss.intel_nmmb
>> A src/conf/modules.nems.Theia_ESMF_700_gsm
>> U src/conf/modules.nems.Wcoss_ESMF_700_gsm
>> U src/conf/configure.nems.Theia.intel
>> A src/conf/modules.nems.Theia_ESMF_700_nmmb
>> D src/conf/modules.nems.Theia_ESMF_7_gsm
>> D src/conf/modules.nems.Theia_ESMF_7_nmmb
>> A src/detect_machine1.sh
>> A src/detect_machine.sh
>> U src/makefile
>> U src
>> U .
>>
>> tests/tests/ changes are unplanned, result of the WCT linmit changes
>> only and occured becouse of the extrime slowness of Theia this weekend.
>> These changes have no impact on the performance or testing, but increase
>> wallclocktime limits only. Are Ok on testing for both machines.
>>
>> All regression tests passed OK on Theia and WCOSS now , the successful
>> logs are located @
>>
>> https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nem
>> strunk_NEMSLegacyMD_ph2/tests/
>>
>> I'd like to thank Raghu and RDHPCS sys admins who cooperated to speedup
>> RT during the latest Theia H/W troubles.
>>
>> The main difference between last description (below) and current commit
>> is -- the rest of the atmos/phys is located now under gsm/gsmphys project
>> and "integared atmos phys" external linked to that.
>>
>> This restructuring passed all tests as well, including 35 NMMB tests.
>>
>> The NMMB code managers! (Ratko/Tom/Dusan),
>>
>> Please consider the merge of my little changes of the nmmb makefile (when
>> compiled within NEMS) to be adapted to nmmb trunk.
>> I spoke to Tom and Dusan on that already, and it sounds like there are ok
>> with that - it's only 1 minor parameter to be introduced.
>>
>> Because I have no access now to nmmb project, I've drafted this change @
>> my nems branch:
>> "https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nm
>> mb_NEMSLegacyMD_ph2"
>>
>> There is very a very little change in nmmb makefile, but I hope you'll
>> find it acceptable and can reintegrate for me simultaneously with my NEMS
>> trunk commit:
>>
>> [Eugene.Mirvis at g10a2 nmm]$ svn merge --dry-run --reintegrate
>> https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nmm
>> b_NEMSLegacyMD_ph2
>> --- Merging differences between repository URLs into '.':
>> C makefile
>> U .
>> Summary of conflicts:
>> Text conflicts: 1
>>
>> That small change would allow to use unified pre-build phys on both
>> machines, if you will decide to choose so, by changing on parameter in
>> configuration file
>> to PHYS_MODE=prebld.
>>
>> There are following advantages to do so as we can foresee:
>>
>> - —to share physics across all models within and outside NEMS
>> framework.
>> - —to be able run physics within a single column model and work with
>> R2O,
>> - —To utilize PHYS releases in multi- physics & multi- model ensembles,
>>
>> - —to develop a standalone new Common Community Physics Package
>> (CCPP),
>> - —to use a shared physics driver for multiple or unified models, and
>> to
>> - —enable the outside community to readily contribute physics schemes.
>>
>>
>> Please contact me if any questions.
>>
>> Thanks,
>>
>> -Eugene
>>
>>
>>
>>
>> On Tue, Jul 12, 2016 at 11:55 AM, Eugene Mirvis - NOAA Affiliate <
>> eugene.mirvis at noaa.gov> wrote:
>>
>>>
>>>
>>> On Tue, Jul 12, 2016 at 11:48 AM, Eugene Mirvis - NOAA Affiliate <
>>> eugene.mirvis at noaa.gov> wrote:
>>>
>>>> ==
>>>> NEMS TRAC Ticket #174:New NEMS features: the standalone integrated
>>>> pre-builds of Phys and Chem libraries. ==
>>>>
>>>> ----
>>>>
>>>> 1. The current set of the upcoming commits for the NEMS trac ticket
>>>> #174 provides a standalone capabilities for NEMS isolated physics and
>>>> chemistry separately.
>>>>
>>>> 2. The phys and chem library sources and build system will be located
>>>> now as the sub-projects of nceplibs in EMC repository, inheriting all NEMS
>>>> development history as a regular nceplibs sub-projects, contain branches,
>>>> tags and the trunks, tagged currently as v10.0.10, accordingly of the NEMS
>>>> restructuring history.
>>>>
>>>> 3. Both: phys and chem are using the same configuration NEMS procedure,
>>>> therefore, each has NEMS conf directory as the external links to the nems
>>>> conf disrectory for now. That secures consistency of configurations between
>>>> phys, chem, and NEMS framework.
>>>>
>>>> 4. A dual built capability is provided for the both phys and chem
>>>> libraries:1) as the stand-alone installations 2) the builts within a
>>>> traditional NEMS structure for phys and chem development process.
>>>>
>>>> 5. That selection is provided via the two external parameters:
>>>> $PHYS_MODE and $CHEM_MODE.
>>>> If these parameters are set as “compile ”the built system assumes full
>>>> compilation if such library. In this case a directory chem source should
>>>> have an external link to chem at the “src” while the phys dir an external
>>>> link at src/atmos level, defined at the subversion properties at NEMSLegacy
>>>> externals. By default, both will be set to “compile” by now, as currently
>>>> agreed with component code manager and will be assigned and are located in
>>>> configuration files in conf directory for Theia and WCOSS.
>>>>
>>>> 6. If these options are set to use of the pre-built libraries on
>>>> operational and R&D EMC machines are needed to be exercised, $ PHYS_MODE
>>>> and/or $CHEM_MODE the value of these parameters must be changed to any
>>>> other than “compile”. In this case, NEMS’s makefile will be acting
>>>> accordingly to use pre-built libraries on each machine and the source code
>>>> within NEMS structure is not necessary to be checked out.
>>>>
>>>> 7. The pre-built PHYS and CHEM libraries locations :
>>>>
>>>>
>>>> The pre-built libraries are available on Theia & WCOSS at the
>>>> following locations:
>>>> On Theia:
>>>> /scratch3/NSEPDEV/nceplibs/dev/
>>>> phys, chem, phys_full and chem_full
>>>> The builds of phys and chem contain the pre-builds with an option of
>>>> GOCART_MODE=stub while phys_full and chem_full contain pre-builds with the
>>>> GOCART_MODE=full.
>>>>
>>>> On WCOSS, the locations of the similar builts are at:
>>>> /usrx/local/mceplibs/dev/
>>>> phys, chem, phys_full and chem_full consecutively.
>>>>
>>>> You can link against each of these libraries directly but you have to
>>>> change the values of PHYS_MODE and CHEM_MODE from the “compile” to, perhaps
>>>> "prebuild" in the NEMS conf/configure.nems.Theia* and
>>>> conf/configure.nems.WCOSS* files. That will automatically assume in the
>>>> NEMS makefile the pre-built location paths are activated on each site
>>>> .
>>>> 8. The chem and phys modulefiles______________________
>>>> In order to conveniently manipulate with different options of use, for
>>>> installation procedure and for regression testing; and also to support the
>>>> NCEPLIBS computing environment flexibility and version standards, phys and
>>>> chem modulefiles have been created. These modulefiles are also residing
>>>> under EMC repositiory as the NEMSmodulefiles sub-project of the nceplibs
>>>> project and are also available on Theia and WCOSS currently under:
>>>> /scratch3/NCEPDEV/nceplibs/dev/NEMS modulefiles,
>>>> and
>>>> /usrx/local/nceplibs/dev/NEMS modulefiles, consecutively, until an
>>>> operational implementation of the phys and chem @ nwprod will occure
>>>> Also at the repository :
>>>> https://svnemc.ncep.noaa.gov/projects/nceplibs/Modules/NEMSmodulefiles/
>>>>
>>>> 9. Phys and Chem installation and version control
>>>> procedures________________________
>>>> The installation of the phys and chem are simplified to :
>>>> a) Checking out need version, dev branch, tag or the phys trunk of the
>>>> development
>>>> b) gmake clean
>>>> c) running ./Install script
>>>> Commit procedure to the phys and chem trunks should be address to the
>>>> phys and chem code manager.
>>>>
>>>> 10. Expected file changes in reintegration:
>>>> 10.1. CHEM:
>>>> [Eugene.Mirvis at t14a1 chem_trunk]$ svn merge --dry-run --reintegrate
>>>> https://svnemc.ncep.noaa.gov/projects/nceplibs/chem/branches
>>>> /Euge/nemstrunk_NEMSLegacyMD
>>>> --- Merging differences between repository URLs into '.':
>>>> A ESMFVersionDefine_ESMF_NUOPC.h
>>>> A configure
>>>> A Compile_theia.log
>>>> A ESMFConvenienceMacros.h
>>>> A InstallM
>>>> A Modulefiles_chem
>>>> A Modulefiles_chem/Modulefiles_phys
>>>> A Modulefiles_chem/chem
>>>> A Modulefiles_chem/chem/10.0.7
>>>> A Modulefiles_chem/chem/10.0.1
>>>> A Modulefiles_chem/chem/10.0.10
>>>> A Modulefiles_chem/chem/build10.0.10
>>>> A Modulefiles_chem/chem/10.0.2
>>>> A Compile_wcoss.log
>>>> A Install
>>>> A detect_machine1.sh
>>>> A clean.h
>>>> A makefile_std1
>>>> A makefile_std2
>>>> U makefile
>>>> U .
>>>> No textual conflicts are expected
>>>>
>>>> 10.2 PHYS reintegration_____________________________:
>>>>
>>>> Phys directory contains 2 externals for now (that reflects what was
>>>> insisted by the MESO group and accepted by M.Iredell for now)
>>>> [Eugene.Mirvis at t14a1 phys]$ svn propget svn:externals
>>>> conf https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nem
>>>> strunk_NEMSLegacyMD/src/conf
>>>> nmmbphys https://svnemc.ncep.noaa.gov/projects/nmmb/trunk/nmmbphys
>>>>
>>>> Reintegration:
>>>> [Eugene.Mirvis at t14a1 phys_trunk]$ svn merge --dry-run --reintegrate
>>>> https://svnemc.ncep.noaa.gov/projects/nceplibs/phys/branches
>>>> /Euge/ncepphys_trunk_NEMSLegacy_1physMD
>>>> --- Merging differences between repository URLs into '.':
>>>> A nmmbphys_Cln
>>>> A share
>>>> A share/mkDepends.pl
>>>> A Install
>>>> A ESMFVersionDefine_ESMF_NUOPC.h
>>>> A Compile_theia.log
>>>> A include
>>>> A include/ESMFVersionDefine_ESMF_NUOPC.h
>>>> A include/ESMFVersionDefine.h
>>>> A include/configure.nems_univ
>>>> A include/ESMFVersionLogic.h
>>>> A include/NCEP_base.mk
>>>> A include/ESMFVersionDefine_ESMF_6.3.0rp1API.h
>>>> A include/NUOPC.inc
>>>> A include/ESMFVersionDefine_ESMF_3.h
>>>> A include/kind.inc
>>>> A Compile_wcoss.log
>>>> A Install0
>>>> A nmmbphys_Get
>>>> U makefile
>>>> A configure
>>>> A detect_machine1.sh
>>>> A makefile_TODO
>>>> A mersenne_twister.f
>>>> A makefile_std1
>>>> A idea_composition.f
>>>> A Install_gsm
>>>> A Install_nmm
>>>> U .
>>>>
>>>> No textual conflicts are expected
>>>>
>>>> 10.3 NEMS_reintegration:
>>>> [Eugene.Mirvis at t14a1 nems_trunk]$ svn merge --dry-run --reintegrate
>>>> https://svnemc.ncep.noaa.gov/projects/nems/branches/euge/nem
>>>> strunk_NEMSLegacyMD
>>>> --- Merging differences between repository URLs into '.':
>>>> U tests/Compile_theia.log
>>>> U tests/rt.conf
>>>> U tests/RegressionTests_theia.log
>>>> U src/configure
>>>> U src/atmos/share/makefile
>>>> D src/chem
>>>> D src/atmos/share/idea_composition.f
>>>> D src/atmos/phys
>>>> D src/atmos/share/mersenne_twister.f
>>>> U src/atmos/makefile
>>>> U src/conf/modules.nems.Wcoss_ESMF_700_nmmb
>>>> U src/conf/configure.nems.Theia.intel
>>>> U src/conf/configure.nems.Theia.intel_gsm
>>>> U src/conf/configure.nems.Theia.intel_nmmb
>>>> U src/conf/configure.nems.Wcoss.intel_gsm
>>>> U src/conf/configure.nems.Wcoss.intel_nmmb
>>>> A src/conf/modules.nems.Theia_ESMF_700_gsm
>>>> A src/conf/modules.nems.Theia_ESMF_700_nmmb
>>>> U src/conf/modules.nems.Wcoss_ESMF_700_gsm
>>>> D src/conf/modules.nems.Theia_ESMF_7_gsm
>>>> D src/conf/modules.nems.Theia_ESMF_7_nmmb
>>>> A src/detect_machine1.sh
>>>> A src/detect_machine.sh
>>>> U src/makefile
>>>> U src
>>>> U .
>>>>
>>>> No textual or tree conflicts
>>>> 10.4 GSM_reintegration
>>>> [Eugene.Mirvis at t14a1 gsm_trunk]$ svn merge --dry-run --reintegrate
>>>> https://svnemc.ncep.noaa.gov/projects/gsm/branches/Euge/gfs_
>>>> NEMSLegacy1physMD
>>>> --- Merging differences between repository URLs into '.':
>>>> U module_GFS_GRID_COMP_stub.F90
>>>> U module_GFS_INTEGRATE.F90
>>>> U phys/makefile
>>>> U phys
>>>> U libutil/makefile
>>>> U libutil
>>>> U module_GFS_GRID_COMP.F90
>>>> U module_GOCART_ROUTINES_stub.F90
>>>> U module_GFS_CORE_SETUP.F90
>>>> U module_GOCART_ROUTINES.F90
>>>> U dyn/gfs_dynamics_initialize_slg_mod.f
>>>> U dyn/makefile
>>>> U dyn
>>>> UU makefile
>>>> U module_GFS_INTERNAL_STATE.F90
>>>> U .No text conflicts are expected or result change
>>>>
>>>> 10. 5
>>>> [Eugene.Mirvis at t14a1 nmmb_trunk_reint]$ svn merge --dry-run
>>>> --reintegrate https://svnemc.ncep.noaa.gov/p
>>>> rojects/nems/branches/euge/nmmbtrunk_MD
>>>> --- Merging differences between repository URLs into '.':
>>>> U nmmbphys/makefile
>>>> U makefile
>>>> No text conflicts expected
>>>>
>>>> After that, src/atmos/phys and src/chem. Are going to be deleted from
>>>> NEMS and introduced on NEMSLegacy level as the subversion external
>>>> properties as the following:
>>>>
>>>>
>>>> [Eugene.Mirvis at t14a1 NEMSLegacy_1physT_MDg]$ svn propget svn:externals
>>>> NEMS https://svnemc.ncep.noaa.gov/p
>>>> rojects/nems/branches/euge/nemstrunk_NEMSLegacyMD
>>>> NEMS/src/atmos/gsm https://svnemc.ncep.noaa.gov/p
>>>> rojects/gsm/branches/Euge/gfs_NEMSLegacy1physMD
>>>> NEMS/src/atmos/nmm https://svnemc.ncep.noaa.gov/p
>>>> rojects/nems/branches/euge/nmmbtrunk_MD
>>>> NEMS/src/atmos/phys https://svnemc.ncep.noaa.gov/p
>>>> rojects/nceplibs/phys/branches/Euge/ncepphys_trunk_NEMSLegacy_1physMD
>>>> NEMS/src/chem https://svnemc.ncep.noaa.gov/p
>>>> rojects/nceplibs/chem/branches/Euge/nemstrunk_NEMSLegacyMD
>>>>
>>>> For now
>>>> Or after the commit:
>>>> NEMS https://svnemc.ncep.noaa.gov/projects/nems/trunk
>>>> NEMS/src/atmos/gsm https://svnemc.ncep.noaa.gov/projects/gsm
>>>> NEMS/src/atmos/nmm https://svnemc.ncep.noaa.gov/projects/nems/trunk
>>>> NEMS/src/atmos/phys https://svnemc.ncep.noaa.gov/p
>>>> rojects/nceplibs/phys/trunk
>>>> NEMS/src/chem https://svnemc.ncep.noaa.gov/p
>>>> rojects/nceplibs/chem/
>>>>
>>>> With associated revisions on the head of the trunks
>>>>
>>>> 11. Working (tested) directories
>>>> All 4 full sets of the 45 regression tests have been PASSED now.
>>>> There are 2 working directories have been tested on each machines,
>>>> tested with the options a) to compile phys and chem within traditional
>>>> structure $PHYS_MODE=compile, $CHEM_MODE=compile ( are default for now) and
>>>> b) tested with the pre-build phys, chem., phys_full and chem._full
>>>> libraries :
>>>> With the phys and chem. prebuild options PHYS_MODE=compile
>>>> CHEM_MODE=compile:
>>>> On Theia:
>>>> PHYS_MODE=compile
>>>> CHEM_MODE=compile
>>>>
>>>> /scratch4/NCEPDEV/nems/noscrub/Eugene.Mirvis/NEMSLegacy_1physT_MDg
>>>>
>>>> On WCOSS __________:
>>>> PHYS_MODE=compile
>>>> CHEM_MODE=compile
>>>> /gpfs/td1/emc/global/noscrub/Eugene.Mirvis/NEMS/NEMSLegacy_1physT_MDg
>>>>
>>>> With the phys and chem. prebuild options:
>>>> PHYS_MODE=prebuild
>>>> CHEM_MODE=prebuild:
>>>> On Theia:
>>>> /scratch4/NCEPDEV/nems/noscrub/Eugene.Mirvis/NEMSLegacy_1phy
>>>> sT_MD_prebld
>>>> On WCOSS :
>>>> /gpfs/td1/emc/global/noscrub/Eugene.Mirvis/NEMS/NEMSLegacy_1
>>>> physT_MDg_prebld
>>>> All full regression tests have been passed in all 4 directories. No
>>>> results changes.
>>>>
>>>> GSM , NMMB code manages and NEMS team please observe upcoming changes
>>>> in 24 hrs and contact me if you have any questions
>>>>
>>>> Thank you,
>>>>
>>>> -Eugene
>>>>
>>>> ---
>>>> Eugene Mirvis, *Tech Lead,*
>>>> Senior Computational Scientist, IMSG @ EMC
>>>> National Centers for Environmental Prediction,
>>>> National Oceanic and Atmospheric Administration
>>>> 5830 University Research Ct., NCWCP rm 2183
>>>> College Park, MD 20740
>>>> Ph. 301.683.3809
>>>>
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.lstsrv.ncep.noaa.gov/pipermail/ncep.list.nems.announce/attachments/20160812/54acc82b/attachment-0001.html
More information about the Ncep.list.nems.announce
mailing list