本人在企业里也在Linux计算机集群上做计算但用得不是很熟练
分享下我用的脚本 一起交流
#PBS -N test
#PBS -l nodes=4:ppn=16
#PBS -j oe
#PBS -l walltime=24:0:0
cd /public/home/yangjm/fluent_template
echo my job id is $PBS_JOBID | tee test.log
echo run nodes is following: | tee test.log
cat $PBS_NODEFILE | tee -a test.log
sort $PBS_NODEFILE | uniq > ./host
echo begin time is `date` | tee -a test.log
id=`echo $PBS_JOBID|awk -F. '{print $1}' `
NP=`cat $PBS_NODEFILE|wc -l`
fluent 3ddp -g -t$NP -cnf=./host -mpi=pcmpi -ssh -i ./inputfile 2>&1 | tee -a test.log
echo end time is `date` | tee -a test.log
然后inputfile 是
rc fluent7.cas
rd fluent7.dat
/solve/initialize/initialize-flow
it 1000
wd fluent2D1.dat
exit
yes
或者
rc fluent.cas
rd fluent.dat
solve/set/ri 1
file/auto/data 100
solve/set/time-step 1.0e-5
solve/dual 400000 30
exit
yes
根据需要改一些参数，