|
|
@1225
|
9 years |
sdipsl |
- metrics NEMO submission
- enable NEMO_VertLevels group
|
|
|
@1224
|
9 years |
sdipsl |
- fix a typo
|
|
|
@1223
|
9 years |
sdipsl |
- Some metrics for NEMO. Quick and dirty at this stage. NEMO_VertLevels …
|
|
|
@1222
|
9 years |
sdipsl |
- use PCMDI_MP v2.12
|
|
|
@1221
|
9 years |
sdipsl |
- compute some more metrics groups for LMDz
|
|
|
@1220
|
9 years |
sdipsl |
- Any post-procesing errors will stop the computing jobs when …
|
|
|
@1219
|
9 years |
sdipsl |
- dirty fix for ncra memory issue
|
|
|
@1218
|
9 years |
sdipsl |
- Make sure every jobs have a different name for ensemble case
NL
|
|
|
@1216
|
9 years |
sdipsl |
- Dont send those messages for now. No effect on results or robustess. …
|
|
|
@1215
|
9 years |
jgipsl |
Ticket #263
Only return for the case nonblocker if the file do not exist.
|
|
|
@1214
|
9 years |
sdipsl |
- clearer information message
|
|
|
@1213
|
9 years |
sdipsl |
- hotfix. The wrong environment setup was sourced
|
|
|
@1210
|
9 years |
sdipsl |
- we don't want the module list when running checker like tools
|
|
|
@1209
|
9 years |
sdipsl |
- bugfix introduced by r1202
|
|
|
@1208
|
9 years |
sdipsl |
- Print list of loaded modules
|
|
|
@1207
|
9 years |
sdipsl |
- IGCM_debug_BigBro_Finalize and IGCM_debug_Verif_Exit are in charge of …
|
|
|
@1206
|
9 years |
sdipsl |
- Remove IGCM_debug_Verif_Exit_Post. Only IGCM_debug_Verif_Exit will …
|
|
|
@1205
|
9 years |
sdipsl |
- Bugfix
|
|
|
@1204
|
9 years |
sdipsl |
- Add an optional flag for the [Post] section of config.card : …
|
|
|
@1203
|
9 years |
sdipsl |
- bugfix
|
|
|
@1202
|
9 years |
sdipsl |
- send job duration in seconds. see #253
- send project name (aka …
|
|
|
@1201
|
9 years |
sdipsl |
- add jobWarningDelay in seconds to supervision messages see #253 …
|
|
|
@1200
|
9 years |
sdipsl |
... useless cosmetics
|
|
|
@1199
|
9 years |
sdipsl |
- submit metrics job for configuration involving LMDz.
|
|
|
@1198
|
9 years |
sdipsl |
- Supervize post-processing jobs see #252
- Homogenize post-processing …
|
|
|
@1197
|
9 years |
sdipsl |
cleanup
|
|
|
@1196
|
9 years |
sdipsl |
- Fix a typo
|
|
|
@1195
|
9 years |
sdipsl |
- IDRIS said the fix is not needed nowadays
|
|
|
@1194
|
9 years |
sdipsl |
- Optimize nco calls. Need nco > 4.2.0
|
|
|
@1193
|
9 years |
sdipsl |
- First move towards netcdf4.3.3.1_hdf5_parallel : have the correct module …
|
|
|
@1192
|
9 years |
sdipsl |
Bugfix caused by r1189 side effect. On Curie job begins their execution in …
|
|
|
@1191
|
9 years |
aclsce |
Cores assignment on Curie : back to the previous "standard" method.
|
|
|
@1190
|
9 years |
sdipsl |
- bugfix with IGCM_comp_modifyDefFile to allow "clean error" so to say
|
|
|
@1189
|
9 years |
sdipsl |
- Extend Big Brother scope to also watch post-processing jobs. Only the …
|
|
|
@1188
|
9 years |
sdipsl |
- /usr/bin/qsub is the right path. ciclad-ng removed /usr/local/bin/qsub …
|
|
|
@1187
|
9 years |
sdipsl |
be more specific. we only need modules …
|
|
|
@1186
|
9 years |
sdipsl |
Ada bugfix. House keeping side effect when calling scipt interactively …
|
|
|
@1185
|
9 years |
sdipsl |
Bugfix. It failed if run.card contained PeriodState?="Fatal" and …
|
|
|
@1184
|
9 years |
sdipsl |
Add ciclad to the system. Episode II
|
|
|
@1183
|
9 years |
sdipsl |
cosmetics
|
|
|
@1182
|
9 years |
sdipsl |
Add ciclad to the system. see #248
|
|
|
@1181
|
9 years |
sdipsl |
house keeping in libIGCM_sys (fourth pass). Will ease #248 and #251
|
|
|
@1180
|
9 years |
sdipsl |
house keeping in libIGCM_sys (third pass). Will ease #248 and #251
|
|
|
@1179
|
9 years |
sdipsl |
house keeping in libIGCM_sys (second pass). Will ease #248 and #251
|
|
|
@1178
|
9 years |
sdipsl |
house keeping in libIGCM_sys (first pass). Will ease #248
|
|
|
@1177
|
9 years |
sdipsl |
Slightly adjust template editing
|
|
|
@1176
|
9 years |
sdipsl |
Some adjustements to properly configure the metrics package.
It works now
|
|
|
@1175
|
9 years |
sdipsl |
update the metrics package version
|
|
|
@1174
|
9 years |
aclsce |
- Patch to be able to run with next TGCC-curie bullxmpi modification.
|
|
|
@1173
|
9 years |
aclsce |
MPMD + MPI-OpenMP mode
- Modified to fit with new mpirun assignment …
|
|
|
@1172
|
9 years |
sdipsl |
- this ins_job dont need w_i_h
- simplify
|
|
|
@1171
|
9 years |
sdipsl |
- Bugfix and simplify calls to test file existence
|
|
|
@1169
|
9 years |
aclsce |
Modified to take into account modifications related to the use of sync …
|
|
|
@1167
|
9 years |
sdipsl |
run rebuild on standard node using iRAM high speed disk partition
|
|
|
@1166
|
9 years |
sdipsl |
the metrics template need more information
create IGCM_comp_modifyFile …
|
|
|
@1165
|
9 years |
sdipsl |
Add a /bin/sync after the rebuild sequence (only active for curie). Should …
|
|
|
@1164
|
9 years |
sdipsl |
Add ins_job in libIGCM. see #238 Dont change the name for now.
|
|
|
@1163
|
9 years |
sdipsl |
- Add IGCM_sys_sync function. Do a /bin/sync on Curie and a dummy call …
|
|
|
@1162
|
9 years |
sdipsl |
Send metrics results to a central place
|
|
|
@1161
|
9 years |
sdipsl |
- this part is really ugly ... but let's continue to add to it for now …
|
|
|
@1160
|
9 years |
sdipsl |
- better this way …
|
|
|
@1159
|
9 years |
sdipsl |
- metrics first guess
|
|
|
@1158
|
9 years |
sdipsl |
- metrics path
|
|
|
@1157
|
9 years |
sdipsl |
- whitespace cleanup
|
|
|
@1156
|
9 years |
sdipsl |
- homogenize indentation
|
|
|
@1155
|
9 years |
sdipsl |
- enhanced comments
- whitespace clean up
|
|
|
@1154
|
9 years |
sdipsl |
- Add a pattern for node type. See #241
|
|
|
@1153
|
9 years |
sdipsl |
- r1150 missed this bit. Need to flush messages when something wrong …
|
|
|
@1152
|
9 years |
sdipsl |
- Avoid tmpfile usage. See #244
|
|
|
@1151
|
9 years |
sdipsl |
adressing #237
|
|
|
@1150
|
9 years |
sdipsl |
adressing #242
message 2000 and 3000 (pushstack and popstack) has been …
|
|
|
@1149
|
9 years |
sdipsl |
adressing #236
|
|
|
@1148
|
9 years |
sdipsl |
Adapt to mfls recent modifications
|
|
|
@1147
|
10 years |
sdipsl |
broken ensemble. Tentative fix for Agathe. Ensemble are broken since …
|
|
|
@1146
|
10 years |
aclsce |
Added functionality to add the attribute (if it does not exist) and its …
|
|
|
@1145
|
10 years |
aclsce |
Small bug fixed.
|
|
|
@1144
|
10 years |
acosce |
Correct comment on rebuild header
|
|
|
@1143
|
10 years |
sdipsl |
This trick is not any more mandatory. Rsh/rcp have been modified to reduce …
|
|
|
@1142
|
10 years |
sdipsl |
- Simplify. Also needed to work with netcdf4 files. See #233
|
|
|
@1140
|
10 years |
sdipsl |
- May be the end of the /tmp issue. Tests ongoing.…
|
|
|
@1138
|
10 years |
sdipsl |
- Dump one day …
|
|
|
@1136
|
10 years |
sdipsl |
- This is dumb idea …
|
|
|
@1135
|
10 years |
sdipsl |
- Cosmetics. It was backward compatible but it's nicer like this
|
|
|
@1134
|
10 years |
sdipsl |
- Bugfix due to a side effect of the IDRIS /tmp workaround
|
|
|
@1131
|
10 years |
sdipsl |
- Bugfix Ada to avoid collision (permission denied getting file) between …
|
|
|
@1130
|
10 years |
jgipsl |
Corrected bug introduced in rev [1028]
|
|
|
@1129
|
10 years |
sdipsl |
# Nasty hack to overcome /tmp hunting side effect on Ada …
|
|
|
@1128
|
10 years |
jgipsl |
Shortend standalone default value for CARD_DIR to correspond to case …
|
|
|
@1127
|
10 years |
sdipsl |
- Easier for now not to fill the stack for checking tasks. see #232
|
|
|
@1126
|
10 years |
jgipsl |
Changed default queue for main job at obelix. Now use mediump which is …
|
|
|
@1125
|
10 years |
sdipsl |
- Correct a typo again affecting only verbosity "messages"
|
|
|
@1124
|
10 years |
jgipsl |
Added -x in typset SUBMIT_DIR for obelix et default
Added function …
|
|
|
@1123
|
10 years |
sdipsl |
- Correct a typo
|
|
|
@1122
|
10 years |
sdipsl |
- Hunting /tmp to overcome the Ada /tmp issue
|
|
|
@1121
|
10 years |
sdipsl |
- Hunting /tmp usage on Ada
|
|
|
@1120
|
10 years |
sdipsl |
- Hunting /tmp usage …
|
|
|
@1119
|
10 years |
sdipsl |
- Hunting /tmp usage …
|
|
|
@1118
|
10 years |
sdipsl |
- Avoid usage of /tmp. See #232
|
|
|
@1117
|
10 years |
sdipsl |
- Need to have access to IGCM_config_CommonConfiguration to know where to …
|
|
|
@1116
|
10 years |
sdipsl |
- Cosmetics. Just because it's nicer that way
|
|
|