Changes between Version 30 and Version 31 of Doc/env/TgccCurie
- Timestamp:
- 05/17/16 17:10:56 (8 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Doc/env/TgccCurie
v30 v31 47 47 PARTITION STATUS TOTAL DOWN USED FREE TOTAL DOWN USED FREE MpC CpN SpN CpS TpC 48 48 --------- ------ ------ ------ ------ ------ ------ ------ ------ ------ ---- --- --- --- --- 49 standard up 80368 32 77083 3253 5023 1 4824 198 4000 16 2 8 1 50 xlarge up 10112 128 1546 8438 79 1 14 64 4000 128 16 8 1 51 hybrid up 1144 0 264 880 143 0 33 110 2900 8 2 4 1 49 standard up 80352 0 76161 4191 5022 0 4724 298 4000 16 2 8 1 50 hybrid up 264 0 0 264 33 0 0 33 2900 8 2 4 1 52 51 }}} 53 52 * detail of a running job. One command per line ccc_mprun : … … 137 136 }}} 138 137 139 # Fat nodes / Thin nodes # 140 141 As gencmip6 project has resources on standard node only, ins_job (libIGCM/ins_job since libIGCM_v2/7) ask you questions to know you project and where your post-processing jobs will run. Read carefully questions and default answers (frequently appropriate). 142 143 Thin nodes are two times faster than fat nodes for computations; they are as fast as fat nodes for post processing. 144 145 We decided to use thin nodes for computations and fat nodes for post processing. 146 138 # Thin nodes # 139 140 Since April 2016, only thin nodes are available at TGCCC. 147 141 The job header must include #MSUB -q standard to use thin nodes. 148 149 The job header must include #MSUB -q xlarge to use fat nodes.150 142 151 143 ## SSD on standard node : how to use it for rebuild job ## … … 156 148 #MSUB -x # exclusive node 157 149 RUN_DIR_PATH=/tmp/REBUILD_DIR_MR_$$ 158 }}}159 160 ## RAM on fat node : how to use it for rebuild job ##161 If /tmp of standard node (64GB) is too small for you, you can use fat node. You have to know the /tmp of fat node is also of 64GB. You have to use the RAM disk which is 250GB/node. You have only to change header and RUN_DIR_PATH in rebuild.job. Take care you will run faster but cost will be multiplied by a factor of 128 because fat node ie 128 cpus are dedicated.162 163 {{{164 #MSUB -q xlarge # fat nodes165 #MSUB -x # exclusive node166 RUN_DIR_PATH=/dev/shm/REBUILD_DIR_MR_$$167 150 }}} 168 151