Version 16 (modified by flavoni, 3 years ago) (diff)

2017 Work on trunk validation

Last edited :

lime

Date Rev number SETTE results Other specific results Comments and propositions
Test of 50 years of trunk (october 2017) 7991 50 years climatological ORCA2_LIM3_PISCES
Active cpp keys: key_trabbl key_lim3 key_zdftke key_zdfddm key_zdftmx_new key_iomput key_mpp_mpi key_top key_nosignedzero
CORE II forcing
Namelist and xml files and plots in pdf
BAD sst!?

After 2016 Merge Party 7614 50 years climatological ORCA2_LIM3_PISCES
Active cpp keys: key_trabbl key_lim3 key_zdftke key_zdfddm key_zdftmx key_iomput key_mpp_mpi key_top key_nosignedzero
CORE II forcing
Namelist and xml files and plots in pdf Convection seems too strong in Antartica (to be confirmed)

7715 ADA CNRS, Simona Flavoni

GYRE_PISCES,ORCA2_LIM3_PISCES,ORCA2_OFF_PISCES,AMM12, ISOMIP andORCA2_LIM3_OBS restart and repro OK,SASrun but not tested restarts, WORCA2AGUL_2_2 run short_NO zoom ok in revision 7746
NOTA BENE :ORCA2_LIM3_OBS is created from reference configurationORCA2_LIM3_PISCES. maybe discuss on how to remove pisces in sette tests (just withoutkey_topcode do not compile: error at trcnam_my_trc routine) WORCA2AGUL_2_2 run short_NO zoom ok in revision 7746: but I saw that I had these cpps: bld::tool::fppkeyskey_trabblkey_lim2key_zdftkekey_zdfddmkey_zdftmxkey_mpp_mpikey_iomputkey_nosignedzerokey_agrif
7731 and 7746 EKMAN IBM, INGV, Damiano Delrosso


GYRE_PISCES,ORCA2_LIM3_PISCES,ORCA2_OFF_PISCES,AMM12, ISOMIP andORCA2_LIM3_OBS restart and repro OK,SASOK
SETTE tests have been carried out with and without openMP AGRIF configs do not compile. trc.f90(154): error [ #6838]: An object must not have both the POINTER and the TARGET attribute. [TRCDTA_BDY]
7723 ATHENA, CMCC, Francesca Mele


GYRE_PISCES,ORCA2_LIM3_PISCES,ORCA2_OFF_PISCES,AMM12, ISOMIP andORCA2_LIM3_OBS restart and repro OK,SASOK
SETTE tests have been carried out with and without openMP
recent SETTE Tests performed on the trunk with openMP before 2nd March
2 March 2017 It is decided that tests should be performed on the trunk (as soon as Silvia has reverted it removing openMP modifications) to check if everything is ok before starting 2017 developments There are still problems with AGRIF when removing key-top.
Agrif was working with zoom after merge: It was not restartable but it was running.
It is not clear if the problem came out when including openMP modifications since there are not tests just before including the hybridization.
Clement is about to end compatibility between AGRIF and LIM3. The best is to have bug corrections in the trunk.
Silvia: Agrif worked just before and after including OpenMP changes, I tested it at rev 7753.
15 March 2017 7797 dell cluster, intel , Clement B GYRE_PISCES, ORCA2_LIM3_PISCES, ORCA2_OFF_PISCES, AMM12: ok
SAS not rst
ISOMIP not exist
21 March 2017 7816 EKMAN IBM, INGV, Damiano Delrosso

GYRE_PISCES,ORCA2_LIM3_PISCES,ORCA2_OFF_PISCES,AMM12, ORCA2_LIM3_OBS,WORCA2AGUL_2_2,WORCA2AGUL_2_2_NAG OK
SAS not rst
Compiling with -check all option the following error shows up in runtime for GYRE_PISCES configs (see stderr):
forrtl: severe (408): fort: (7): Attempt to use pointer PDATA_2D when it is not associated with a target

Image PC Routine Line Source
[…]
opa 0000000001BE8CF7 histcom_mp_histw_ 1863 histcom.f90
opa 0000000001BDDF41 histcom_mp_histwr 1676 histcom.f90
opa 000000000128785D diawri_mp_dia_wri 529 diawri.f90
opa 0000000000514192 step_mp_stp_ 225 step.f90
opa 000000000040637D nemogcm_mp_nemo_g 158 nemogcm.f90
opa 00000000004060E4 MAIN 18 nemo.f90
[…]
This does not happen when compiling without -check all option
24 April 2017 7952 ADA ​CNRS, Simona Flavoni
GYRE_PISCES,ORCA2_LIM3_PISCES,ORCA2_OFF_PISCES,AMM12, ISOMIP andORCA2_LIM3_OBS restart and repro OK,
SAS run but not tested restarts,
ORCA2_AGUL KO restart, ORCA2AGUL ok reproducibility
error for AGULASLONG: File ./1_restart_ice.nc* not found, read ice restart file at time step : 0.000000000000000E+000, in any case we force it to nit000 - 1 : 150
29 June 2017 8218 ADA_DEBUG ​CNRS, Simona Flavoni
GYRE_PISCES, ORCA2_OFFLINE, AMM12_LONG, ORCA2_LIM3_OBS
ORCA2_LIM3_PISCES ok repro but KO restart,
SAS run but not tested restarts,
ORCA2_AGUL KO restart
2 errors:
first Due to key_tmx_new there are some missing files for agrif: 1_mixing_power_bot.nc, 1_mixing_power_bot.nc, 1_mixing_power_cri.nc, 1_decay_scale_bot.nc. Need to be created via nesting tools.
second error: nn_fsbc(child) must be set to 1 otherwise agrif and lim3 do not work properly
5 July 2017 8284 ADA ​CNRS, Clement Rousset
* GYRE_PISCES Ok
* ORCA2_OFFLINE Ok
* AMM12 Ok
* ORCA2_LIM3_OBS Ok
* ORCA2_LIM3_PISCES Ok
* SAS repro not tested and restart not Ok although SAS_BIPER is restartable
* ORCA2_NAG repro Ok
* ORCA2AGUL repro Ok but restart not Ok
To make AGRIF to work (ORCA2AGUL) one has to remove key_zdftmx_new and add key_zdftmx instead otherwise some files are missing
sette_rpt.sh does not always give the right message (in my case, sette_rpt.sh says amm12 is not repro while it in fact is)

Attachments (1)