Skip to content

Changes

Page history
1st commit for gal tapioca authored by guillaume.morvan_ird.fr's avatar guillaume.morvan_ird.fr
Show whitespace changes
Inline Side-by-side
Configurations/TAPIOCA36.md 0 → 100755
View page @ 20bbd084
* [TAPIOCA36](#tapioca36)
* [Description](#tapioca36-description)
* [Context](#tapioca36-description-context)
* [Experience](#tapioca36-description-experience)
* [Physique](#tapioca36-description-physique)
* [Examples of output fields](#tapioca36-description-examples-of-output-fields)
* [Download Instructions](#tapioca36-download-instructions)
* [Prerequisites](#tapioca36-download-instructions-prerequisites)
* [Download Config Repos](#tapioca36-download-instructions-download)
* [Modules to load](#tapioca36-modules-to-load)
* [Preprocessing](#tapioca36-preprocessing)
* [Setup NEMO](#tapioca36-preprocessing-setup-nemo)
* [Prepare inputs](#tapioca36-preprocessing-prepare-inputs)
* [fetch data already prepared](#tapioca36-preprocessing-prepare-inputs-fetch-data-already-prepared)
* [Specific Information](#tapioca36-specific-information)
* [Horizontal Decomposition](#tapioca36-specific-information-horizontal-decomposition)
* [Vertical Decomposition](#tapioca36-specific-information-vertical-decomposition)
* [Tips](#tapioca36-tips)
* [Post it](#tapioca36-tips-post-it)
<a name="tapioca36-description"></a>
===============================
## Description
<a name="tapioca36-description-context"></a>
========================
#### Context
....
<a name="tapioca36-description-experience"></a>
========================
#### Experience
| Configuration | Name | Model | Res. | Lat<br>Lon</br> | ix<br>iy</br><br>iz</br> | Type | Duration Tests | Cluster | Cpus<br>/model</br> | Time step | Time elapse<br>/yr</br> |
|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
| TAPIOCA36 | Tapioca-sud 36e | NEMO4.0.2 (rev. 12389) <br>XIOS2.5 (rev. 1862) </br> | 3km (1/36e) | [XX;XXN]<br>[XX;XXW]</br> | XX<br>XX</br><br>75</br> | Forced | 2007 <br>to 2017</br> | Jean Zay (Idris) | XX<br>XX (I/O)</br> | XX | XX
###### Miscellaneous
....
* Bathymetry [GEBCO 2019](https://www.gebco.net/data_and_products/gridded_bathymetry_data/gebco_2019/gebco_2019_info.html) (15 arc-second grid) : median interpolation on ORCA12 Bathymetry
* Tidal atlas FES2014b : tidal stream atlas is used to predict the direction and speed of tidal currents (15 harmonics constants files)
Use scalar approximation for load potential
* Forcing datasets :
* MERCATOR GLORYS12V1 (On Northeast Brazil) (1/12°) [see](https://git.outils-is.ird.fr/legos_atnum/repos/nemo_configs/wikis/Introduction#introduction-contents-listing-of-global-inputs-stored)
* ERA5 (CDS) : [-101W;31E/-41S;40.9N] (~0.25 / 0.3°)
* Runoffs (ISBA-CTRIP land surface model (daily runoff data from 1993 to 2017)) (B.Decharme)
* Validation datasets :
* [WOA2018](https://www.nodc.noaa.gov/OC5/woa18/)
* Inputs storage (only for configuration use, not common inputs) in user space : 35Go
* Outputs storage (2005 to 2006): 700Go
* Hour Estimation for production (2 years) = near 7000h (Number of procs * time elapse simu / year * years). We assume here that simulation use all the nodes on Jeanzay (40proc by nodes)
* Project name on cluster Jean-zay : lah
* Location : Jeanzay cluster (Contact G.Morvan)
<a name="tapioca36-description-physique"></a>
========================
#### Physics
| | Data | Parameters | Miscellaneous
|:---:|:---:|:---:|:---:|
| Initial Condition | Mercator GLORYS12V1 | T,S | - |
| Boundary Condition | Mercator GLORYS12V1 | T,S, UV2d, UV_3d, SSH daily interanual | 4 layers relaxation scheme
| Atmo | Reanalysis ERA5 | UV10(1h) / (TQ)2(1h) / rad(sw/lw) (1h) / precip/snow (1h) / slp (6h) interanual | - |
| Runoff | ISBA-CTRIP land surface model | Runoff daily interanual | - |
| Solar flux | Orbview-2 SeaWiFS, R2018.0, 0.1°, Global, 1997-2010 (Monthly Composite) | CHLA clim monthly | - |
<a name="tapioca36-description-examples-of-output-fields"></a>
========================
####
....
----
<a name="tapioca36-download-instructions"></a>
===============================
## Download Instructions
<a name="tapioca36-download-instructions-prerequisites"></a>
========================
#### Prerequisites
- Define your path where all directories will remain (as user want), hereafter, we assume your current directory, this variable will be the top level of your environment (keep in mind) :
```console
export WRKDIR=$(pwd)
```
- The git clone command needs your __username you are using to log into gitlab__ :
```console
USERNAME=???
```
- Protocole for cloning repos , by default https :
```console
PROTOCOLE=https
```
- Your tools name directory (actually the tag from the tools repo, to define as :
```console
TOOLSDIR=dev_r12389_config_nemo_legos
```
<a name="tapioca36-download-instructions-download"></a>
========================
#### Download Config Repos
Instructions to do for the first time, it will ask you your password on these steps :
_Download tools repository_:
```console
git clone ${PROTOCOLE}://${USERNAME}@git.outils-is.ird.fr/legos_atnum/repos/nemo_tools.git $WRKDIR/NEMO_TOOLS/$TOOLSDIR
```
_Choose right tag for tools_:
```console
cd $WRKDIR/NEMO_TOOLS/$TOOLSDIR ; git checkout $TOOLSDIR ; cd $WRKDIR
```
_Setup configuration repositories_:
```console
$WRKDIR/NEMO_TOOLS/$TOOLSDIR/preprocessing/setup_environment/getupsetup_repos.sh -c TAPIOCA36 -r 12389 -p $PROTOCOLE -u $USERNAME
```
----
<a name="tapioca36-modules-to-load"></a>
===============================
## Modules to load (for each Jeanzay cluster session)
First of all, please ensure you did well the prerequisites section, see [here](https://git.outils-is.ird.fr/legos_atnum/repos/nemo_configs/wikis/Configurations/TAPIOCA36#tapioca36-download-instructions-prerequisites) , mainly to declare your environment variable _WRKDIR_ (the top level directory of your experiment).
When _Download instructions_ already completed without problem, for each linux terminal opened on Jeanzay, reload all library needed and aliases as written below ...
```console
[ "$WRKDIR" == "" ] && export WRKDIR=$(pwd)
module purge
module use -a $WRKDIR/modulefiles
module load dev_nemo_r12389_TAPIOCA36
```
> **Note** : It is better to purge (_module purge_) to prevent incompatibility issues
> **Note** : We encourage people who wants to process this configuration to append at the end of his __~/.bash_profile__ the appropriate WRKDIR variable and copy/paste this block section above to prevent errors.
<a name="tapioca36-preprocessing"></a>
===============================
<a name="tapioca36-preprocessing-setup-nemo"></a>
========================
## Setup NEMO
---
#### Execution script
Now, You have a script called **build_nemo_main.ksh** from Nemo tools repos, It will use and read the configuration file called **includefile.ini** from the Nemo configuration repository.
So, If the user decides to execute the script **build_nemo_main.ksh**, he must look at the values written in **includefile.ini**, for instance he can activate some flags to do such as :
* PROJECT= # to fill by user, it s the ID project on cluster
* XIOS_COMPIL=FALSE
* NEMO_COMPIL=FALSE
* PREP_RUNDIR=TRUE
* PREP_TOOLS=TRUE
* DOIT_NESTING=(TRUE/FALSE)
* DOIT_DOMAINCFG=(TRUE/FALSE)
* DOIT_MESHMASK=(TRUE/FALSE)
* DOIT_WEIGHTS=(TRUE/FALSE)
* DOIT_REBUILD_NEMO=(TRUE/FALSE)
Wherever you are on your environment , changes can be made on **includefile.ini** file, then source it (reminder):
```console
set -a;. ${WRKDIR}/NEMO_CONFIGS/setup/TAPIOCA36/conf/includefile.ini;set +a
```
or with an alias , it will open an editor here **vim** and change it directly, for this type:
```console
config_edit
```
> **Note** : If you want to change editor tools (other than vim) executed with the alias __config_edit__, change the config modulefile, and reload it...
Finally you can launch it as is :
```console
build_nemo_main.ksh
```
> **Note** : script build_nemo_main.ksh is located here : ${WRKDIR}/NEMO_TOOLS/dev_r12389_config_nemo_legos/preprocessing/build_NEMO/
<a name="tapioca36-preprocessing-prepare-inputs"></a>
========================
## Prepare inputs
<a name="tapioca36-preprocessing-prepare-inputs-fetch-data-already-prepared"></a>
=============
==> **Fetch data already prepared**
...
<a name="tapioca36-specific-information"></a>
===============================
## Specific Information
<a name="tapioca36-specific-information-horizontal-decomposition"></a>
========================
#### Horizontal decomposition
| XP | Nprocs / jpni x jpnj / jpi x jpj | ratio sup (computed / global)
|:---:|:---:|:---:|
_Boundary mesh _ :
see here : ...
<a name="tapioca36-specific-information-vertical-decomposition"></a>
========================
#### Vertical decomposition
_Depth for TAPIOCA (ORCA12)_ :
<a name="tapioca36-specific-information-workflow-chart"></a>
========================
#### Workflow chart
soon
<a name="tapioca36-specific-information-code-decomposition"></a>
========================
#### Code decomposition
soon testing.
<a name="tapioca36-specific-information-Checking-reproductibility"></a>
========================
#### Checking reproductibility (Singularity egg processed on Irene)
soon testing
<a name="tapioca36-tips"></a>
===============================
<a name="tapioca36-tips-post-it"></a>
========================
## Post It (command line)
* Compilation :
```
cd $P_CTL_DIR/..
rm -rf BLD
cd $NEMO_BASE_DIR
./makenemo -m X64_JEANZAY -r ORCA2_ICE_PISCES -n TAPIOCA36-G00 -j8 -v3 > log_build_TAPIOCA36-G00.txt 2>&1
```
* Job command line :
```
squeue | grep $USER
scancel jobid
...
```
----
## Articles,references
empty