Changeset 10819
- Timestamp:
- 2019-04-01T15:10:17+02:00 (6 years ago)
- Location:
- NEMO/branches/2019/fix_sette_ticket2239
- Files:
-
- 2 edited
Legend:
- Unmodified
- Added
- Removed
-
NEMO/branches/2019/fix_sette_ticket2239/README
r10769 r10819 1 1 usage of SETTE package: 2 2 3 INSTALLATION: 4 The sette directory have to be located in the main directory (at the same level as src/ or cfg/). 5 3 6 PRINCIPAL SCRIPT: sette.sh 4 * sette.sh is a simple wrapper running test on the reference configuration, test cases and generate the report at the end. 5 * sette.sh is able to take a list of configuration in argument (sette.sh -t "CFG1 CFG2 CFG3"). The list of available configurations is available in running ./sette_list_avail_cfg.sh. 6 The default for sette is to run all of the reference and test configurations. 7 * User need to set up the param.cfg file correctly and download input file before running the script. 7 * sette.sh is a simple wrapper that runs tests on the reference configurations, test cases and generates the report at the end. 8 * sette.sh is able to take a list of configurations in argument (sette.sh -t "CFG1 CFG2 CFG3"). 9 - The list of available configurations is available in running ./sette_list_avail_cfg.sh. 10 - The default for sette is to run all of the reference and test configurations. 11 * Users need to set up the param.cfg file correctly and download input files before running the script. 8 12 9 13 SECONDARY SCRIPT 10 14 * sette_reference-configuration.sh: run sette tests on the reference configurations 11 15 * sette_test-cases.sh : run sette tests on test cases 12 * sette_rpt.sh : generate the sette report on reproducibility, restartability and compare result with a reference (if there is one specified in param.cfg). 13 See below in USER MODIFICATIONS the usage of the script 14 * sette_list_avail_rev.sh : generate a listing on all the revision available in the validation directory for each configuration. 16 * sette_list_avail_rev.sh : generate a listing on all the revisions available in the validation directory for each configuration. 15 17 * sette_list_avail_cfg.sh : generate a listing of all the available configurations. 16 18 17 VERY IMPORTANT : for a verbose debug do: 18 ./sette.sh 2>&1 | tee out.sette 19 this create out.sette file in ${SETTE_DIR} 20 21 USER MODIFICATIONS 22 * in sette_reference-configuration.sh and sette_test-cases.sh script : 23 OPTIONAL ARGUMENT : compiler among those available in NEMOGCM/ARCH (if nothing will use the one specified in paramd.cfg) 24 BATCH_COMMAND_PAR : command for job submission in batch mode parallel (MMIÂ) specified in param.cfg 25 BATCH_COMMAND_SEQ : command for job submission in batch mode sequential (NB_PROC = 1) 26 default value is the BATCH_COMMAND_PAR value 27 INTERACT_FLAG : "yes" if you want to run in interactive mode 28 "no" if you want to run in batch mode 29 MPIRUN_FLAG : "yes" if you want to run in parallel (MPI) 30 "no" if you want to run in sequential mode (NB_PROC = 1) 31 32 makenemo -m : machine (compilator); see ./makenemo -m help (in NEMOGCM/CONFIG directory) 33 -n : config name (in examples config_name_nproci_nprocj), see ./makenemo -n help 34 -r : reference configuration (if you don't give it you have to choise directories to install) 35 add_key : to add list of keys 36 del_key : to del list of keys 37 and for more details and options of makenemo see ./makenemo -h 38 39 set_namelist : to set namelist values 40 (example : to set 75 time steps for one test : namelist nn_itend 75) 41 ( : to change cn_icerst_in name : namelist_ice cn_icerst_in \"test_00101231_restart_icemod\") 42 ( : to activate control print : namelist ln_ctl false ) 43 prepare_exe_dir.sh creates execution directory for each test 44 : NOTE for EACH JOB EXECUTION it is needed to set TEST_NAME variable 45 TEST_NAME variable gives name to execution directory (created in prepare_exe_dir.sh) 46 IT is mandatory to have different TEST_NAME for each job run if not existing files are re-written (for example namelist) 47 48 prepare_job.sh generates the script to be executed 49 input: INPUT_FILE_CONFIG_NAME NUMBER_PROC TEST_NAME MPIRUN_FLAG JOB_FILE 50 51 fcm_job.sh needs input: TOTAL_NUMBER_OF_PROCS_used JOB_FILE_NAME INTERACT_FLAG MPIRUN_FLAG 52 (for example if you run with 4 procs JOB_FILE batch parallel 53 and only run short test ===> ./fcm_job 4 SHORT no yes ) 54 19 USER SETUP (recommended) 55 20 * in param.cfg : 21 - variables (COMPILER, NEMO_VALIDATION_REF, NEMO_REV_REF) have to be specified in param.cfg 22 - variables (BATCH_CMD, BATCH_STAT, BATCH_NAME, FORCING_DIR, SVN_CMD, ADD_NOSIGNEDZERO) 23 can be exported from your shell startup files. It is advise to setup the default value in your startup file if it doesn't fit your need. 24 If you prefer not doing so, you need to update the default initialisation 25 - variable NEMO_VALIDATION_DIR is sette by default in MAIN_DIR/NEMO_VALIDATION (this is to avoid overwritting results when running sette on different branches). 26 - description of variables listed in param.cfg: 56 27 COMPILER : compiler among those available in NEMOGCM/ARCH 57 FORCING_DIR : directory in which will be found input.tar file (same name in input_CONFIG_NAME.cfg) 58 INPUT_DIR : directory in which store input files (tar file) 59 reference configuration input tar file could be found here : https://zenodo.org/record/1472245 60 TEST_CONFIGS : the list of all the configuration to test in sette.sh (without any arguments) 61 TMPDIR : temporary directory NEEDED ONLY FOR IBM machines (put EXP00 directory) 62 NEMO_VALIDATION_DIR : directory in which create NEMO_VALIDATION tree, and store restart, run.stat, tracer.stat and ocean.output files in 63 tree NEMO_VALIDATION_DIR/WCONFIG_NAME/WCOMPILER_NAME/TEST_NAME/REVISION_NUMBER(or DATE) 64 BATCH_CMD : command for job submission in batch mode 65 SVN_CMD : svn command use to do svn info (default svn). Could be useful if you are using git svn 28 # 29 # reference version for bit reproducibility results 66 30 NEMO_VALIDATION_REF : reference directory 67 31 NEMO_REV_REF : reference revision 68 reference directory for result comparison will be NEMO_VALIDATION_REF/COMPILER/NEMO_REV_REF 32 # 33 # batch command needed 34 BATCH_CMD : command for job submission in batch mode 35 BATCH_STAT : command to check job status 36 BATCH_NAME : generic sette job name (as it appears with $BATCH_STAT command) 37 # 38 # file storing 39 FORCING_DIR : directory in which will be found input.tar file (same name in input_CONFIG_NAME.cfg) 40 reference configuration input tar file could be found here : https://zenodo.org/record/1472245 41 NEMO_VALIDATION_DIR : directory in which will be stored restart, run.stat, tracer.stat and ocean.output files for each configuration 42 ( NEMO_VALIDATION_DIR/WCONFIG_NAME/WCOMPILER_NAME/TEST_NAME/REVISION_NUMBER(or DATE) ) 43 INPUT_DIR : directory in which store input files (DO NOT CHANGE IT) 44 # 45 # misc. 46 SVN_CMD : svn command use to do svn info (default svn). Could be useful if you are using git svn 47 : Reference directory for result comparison will be NEMO_VALIDATION_REF/COMPILER/NEMO_REV_REF 69 48 ADD_NODIGNEDZERO : set "yes" if you need key_nosignedzero to run nemo 70 As a lot of this settings are user dependent or computer dependant, it is possible to set by default most of these value in your .bashrc71 (it facilitates the installation of sette for a new branch)72 49 73 * in fcm_job.sh : 74 IMPORTANT : control that name of tar file in FORCING directory 75 is the same of that in input_REF_CONFIG.cfg, if not change one. 50 USAGE of main scripts: 51 * sette.sh : it is a simple wrapper to run tests on the reference configurations, test-cases and generate the report at the end. 52 - if no argument is given, sette run all the reference configurations and test cases 53 - if a list of argument is provided (sette.sh -t "CFG1 CFG2 CFG3"), sette will only run these configurations. 54 The list of available configurations is available in running ./sette_list_avail_cfg.sh. 55 - user need to set up the param.cfg file correctly and download input file before running the script. 56 * sette_rpt.sh : it generates the sette report. 57 - if no argument is given, the report will be generated on the last changed revision. 58 - if an argument is given (revision number) the report will be generated for this revision only 59 - if 'old' is given as argument the former behavior is applied (the latest revision is check whatever the current revision 60 - it is possible to retreive all the available revision test using sette_list_avail_rev.sh 61 - it is NOT possible to run sette_rpt.sh for a single configuration. 62 * sette_list_avail_rev.sh : generate a listing on all the revisions available in the validation directory for each configuration. 63 - no argument needed 64 * sette_list_avail_cfg.sh : generate a listing of all the available configurations. 65 - no argument needed 76 66 77 change in "RUN OPA" if [ ${INTERACT_FLAG} == "yes" ] mpirun command line (some examples are already written) 78 NOTE: if INTERACT_FLAG="no" it is needed a batch_file (see below to create a new one) 79 80 * in sette_rpt.sh : - if no argument are given, the report will be generated on the last changed revision. 81 - if an argument is given (revision number) the report will be generated for this revision 82 - if 'old' is given as argument the former behavior is applied (the latest revision is check whatever the current revision 83 - it is possible to retreive all the available revision test using sette_list_avail_rev.sh 84 85 NOTE: 86 in case of error you can remove you NEW_CONF directory and all files doing : 87 ./makenemo -n MY_CONFIG clean_config 88 89 if you want recompile nemo but before you want to clean compiled code do : 90 ./makenemo clean 91 92 and if you have already compiled you can re-run all sette.sh and compilation part will be by-passed. 93 94 NOTE 2: 95 if you want a completly verbose makenemo you can uncomment "set -x" in makenemo script 96 and then launch ./sette.sh 2>&1 | tee out.sette 67 NOTES: 68 * compilation issues: 69 - in case of error you can remove you NEW_CONF directory and all files doing : 70 ./makenemo -n MY_CONFIG clean_config 71 - if you want recompile nemo but before you want to clean compiled code do : 72 ./makenemo clean 73 - if you have already compiled you can re-run all sette.sh and compilation part will be by-passed. 74 * verbose sette output: 75 - if you want a completly verbose makenemo you can uncomment "set -x" in makenemo script 76 and then launch ./sette.sh 2>&1 | tee out.sette . This creates out.sette file in ${SETTE_DIR} 97 77 98 78 TO ADD NEW CONFIGURATION: … … 100 80 2. makenemo -n NEW_CONFIG -r REF_CONFIG (if it is based on another reference configuration) 101 81 3. for fcm_job.sh creates a new input_NEW_CONFIG.cfg if you need tar file (if you use same tar file of GYRE, ORCA2_LIM or ORCA2_LIM_PISCES you can use it) 82 4. add a bloc in one of the sette_reference-configuration.sh or sette_test-cases.sh script 83 5. add your configuration to the list in param.cfg 102 84 103 85 TO ADD NEW MACHINE: … … 110 92 2. create you own batch file like: batch-${COMPILER} 111 93 (note: sette_test-cases.sh and sette_cfg-ref.sh will copy it in job_batch_template if you run tests in INTERACT_FLAG="no") 94 95 96 97 98 99 USER MODIFICATIONS (sequential/parrallel (default), interacive or not (default) and mpi (default) or not ) 100 * in sette_cfg-ref.sh and sette_test-cases.sh script : 101 - BATCH_COMMAND_PAR is the command for job submission in batch mode parallel (MMIÂ) specified in param.cfg 102 - BATCH_COMMAND_SEQ is the command for job submission in batch mode sequential (NB_PROC = 1) 103 the default value is the BATCH_COMMAND_PAR value 104 - INTERACT_FLAG : "yes" if you want to run in interactive mode 105 "no" if you want to run in batch mode (default) 106 - MPIRUN_FLAG : "yes" if you want to run in parallel (MPI) (default) 107 "no" if you want to run in sequential mode (NB_PROC = 1) 108 109 USAGE of internal function/script 110 * makenemo -m : machine (compilator); see ./makenemo -m help (in NEMOGCM/CONFIG directory) 111 -n : config name (in examples config_name_nproci_nprocj), see ./makenemo -n help 112 -r : reference configuration (if you don't give it you have to choise directories to install) 113 add_key : to add list of keys 114 del_key : to del list of keys 115 and for more details and options of makenemo see ./makenemo -h 116 * set_namelist : to set namelist values 117 (example : to set 75 time steps for one test : namelist nn_itend 75) 118 ( : to change cn_icerst_in name : namelist_ice cn_icerst_in \"test_00101231_restart_icemod\") 119 ( : to activate control print : namelist ln_ctl false ) 120 * prepare_exe_dir.sh : creates execution directory for each test 121 - for EACH JOB EXECUTION it is needed to set TEST_NAME variable 122 - TEST_NAME variable gives name to execution directory (created in prepare_exe_dir.sh) 123 - It is mandatory to have different TEST_NAME for each job run if not existing files are re-written (for example namelist) 124 * prepare_job.sh : generates the script to be executed 125 - input arguments are INPUT_FILE_CONFIG_NAME NUMBER_PROC TEST_NAME MPIRUN_FLAG JOB_FILE 126 * fcm_job.sh : input arguments are TOTAL_NUMBER_OF_PROCS_used JOB_FILE_NAME INTERACT_FLAG MPIRUN_FLAG 127 (for example if you run with 4 procs JOB_FILE batch parallel 128 and only run short test ===> ./fcm_job 4 SHORT no yes ) 129 - IMPORTANT control that name of tar file in FORCING directory 130 is the same of that in input_REF_CONFIG.cfg, if not change one. 131 - change in "RUN OPA" if [ ${INTERACT_FLAG} == "yes" ] mpirun command line (some examples are already written) 132 - if INTERACT_FLAG="no" it is needed a batch_file (see below to create a new one) -
NEMO/branches/2019/fix_sette_ticket2239/param.cfg
r10769 r10819 1 # Computer dependant variable needed to run sette1 # Computer/user dependant variable needed to run sette 2 2 # 3 3 # Compiler among those in NEMOGCM/ARCH … … 9 9 NEMO_REV_REF=0000 10 10 # 11 # IF SETTE_* variable mentioned below are exported from your initialisation, nothing to do 12 # otherwise, please update the default iniitialisation or overwrite it 11 # ------------------------------------------------------------------------------------------ 12 # IMPORTANT: 13 # variables (BATCH_CMD, BATCH_STAT, BATCH_NAME, FORCING_DIR, SVN_CMD, ADD_NOSIGNEDZERO) 14 # can be exported from your shell startup files. 15 # If it is the case, nothing to do 16 # otherwise, you need to update the default initialisation 17 # 18 # ADVISE: setup the default value in your startup file if it doesn't fit your need. 13 19 # 14 20 # BATCH command … … 17 23 # command to check job status 18 24 BATCH_STAT=${SETTE_BATCH_STAT:-llq} 19 # generic sette job name 25 # generic sette job name (as it appears with $BATCH_STAT command) 20 26 BATCH_NAME=${SETTE_BATCH_NAME:-sette} 21 # List of test configurations22 TEST_CONFIGS=${SETTE_TEST_CONFIGS:-"ORCA2_ICE_PISCES ORCA2_OFF_PISCES AMM12 AGRIF SPITZ12 GYRE_PISCES SAS ORCA2_ICE_OBS ICE_AGRIF OVERFLOW LOCK_EXCHANGE VORTEX ISOMIP"}23 27 # 24 #- FORCING files storing 28 # 29 # FILES STORING 30 # forcing files storing 25 31 FORCING_DIR=${SETTE_FORCING_DIR:-$WORKDIR/FORCING} 32 # validation files storing 33 NEMO_VALIDATION_DIR=${WORKDIR:-$MAIN_DIR}/NEMO_VALIDATION 34 # input files storing (namelist, iodef ...) (DO NOT CHANGE) 35 INPUT_DIR=${CONFIG_DIR}/${NEW_CONF}/EXP00 26 36 # 27 # CMD for svn (some people use git svn) 37 # 38 # MISCELENIOUS 39 # command for svn (some people use git svn) 28 40 # used in sette_rpt and in all_function to display revision and build correct repository name 29 41 SVN_CMD=${SETTE_SVN_CMD:-svn} 30 # 42 # add key key_nosignedzero at the compilation step 43 ADD_NOSIGNEDZERO=${SETTE_ADD_NOSIGNEDZERO:-"yes"} 31 44 #- only for IBM 32 45 #TMPDIR=${CONFIG_DIR}/${NEW_CONF}/EXP00 33 46 # 34 #- key_nosignedzero35 ADD_NOSIGNEDZERO=${SETTE_ADD_NOSIGNEDZERO:-"yes"}36 47 # 37 #- INPUT files storing 38 INPUT_DIR=${CONFIG_DIR}/${NEW_CONF}/EXP00 39 # 40 #- VALIDATION files storing 41 NEMO_VALIDATION_DIR=${WORKDIR:-$MAIN_DIR}/NEMO_VALIDATION 42 # 48 # LIST OF CONFIGURATIONS 49 # to be updated if you added an new configuration in sette_test-cases.sh or sette_reference-configuration.sh 50 TEST_CONFIGS=${SETTE_TEST_CONFIGS:-"ORCA2_ICE_PISCES ORCA2_OFF_PISCES AMM12 AGRIF SPITZ12 GYRE_PISCES SAS ORCA2_ICE_OBS ICE_AGRIF OVERFLOW LOCK_EXCHANGE VORTEX ISOMIP"}
Note: See TracChangeset
for help on using the changeset viewer.
