python脚本在超算上提交作业时出错

[weixin_45907705]

用python在天河二号上面提交作业的时候遇到了如下错误，自己没有搞明白问题出在哪里，来请教大家。可以有偿解决，谢谢！

Traceback (most recent call last):
File "/BIGDATA1/hwrf/hwrfrun/scripts/exhwrf_init.py", line 89, in <module>
main()
File "/BIGDATA1/hwrf/hwrfrun/scripts/exhwrf_init.py", line 73, in main
init.run_through_anl()
File "/BIGDATA1/hwrf/hwrfrun/ush/hwrf/init.py", line 559, in run_through_anl
self.geogrid.run()
File "/BIGDATA1/hwrf/hwrfrun/ush/hwrf/wps.py", line 637, in run
produtil.run.checkrun(cmd,logger=logger)
File "/BIGDATA1/hwrf/hwrfrun/ush/produtil/run.py", line 415, in checkrun
raise ExitStatusException('%s: non-zero exit status'%(repr(arg),),r)
ExitStatusException: exe('/usr/bin/srun')['--export=ALL','--cpu_bind=core','--distribution=block:block','/BIGDATA1/hwrf/hwrfrun/sorc/WPSV3/geogrid.exe'].out('geogrid.log',append=False): non-zero exit status (returncode=1)
11/20 20:53:40Z NONAME-CRITICAL: HWRF init is aborting: exe('/usr/bin/srun')['--export=ALL','--cpu_bind=core','--distribution=block:block','/BIGDATA1/hwrf/hwrfrun/sorc/WPSV3/geogrid.exe'].out('geogrid.log',append=False): non-zero exit status (returncode=1)

run.py :
def checkrun(arg,logger=None,**kwargs):
"""!This is a simple wrapper round run that raises
ExitStatusException if the program exit status is non-zero.

@param arg the produtil.prog.Runner to execute (output of
exe(), bigexe() or mpirun()
@param logger a logging.Logger to log messages
@param kwargs The optional run=[] argument can provide a different
list of acceptable exit statuses."""
r=run(arg,logger=logger)
if kwargs is not None and 'ret' in kwargs:
if not r in kwargs['ret']:
raise ExitStatusException('%s: unexpected exit status'%(repr(arg),),r)
elif not r==0:
raise ExitStatusException('%s: non-zero exit status'%(repr(arg),),r) #第415行
return r

wps.py:
class Geogrid(WPSTask):
.....
def run(self):
"""!Copies inputs, links fix files, runs geogrid and delivers
results."""
logger=self.log()
try:
produtil.fileop.makedirs(self.outdir)
with NamedDir(self.location) as dir:
logger.info('Geogrid running in directory: '+os.getcwd())
assert(not re.match('\A/tmp',os.getcwd()))

for f in glob.glob('geo*'):
try:
produtil.fileop.remove_file(f,logger=logger)
except(EnvironmentError) as e:
logger.warning('%s: did not remove file, but '
'continuing anyway'%(f,))
self.link_fix(geog_data=True,table='GEOGRID.TBL')

with open('namelist.wps', 'w') as f:
f.write(self.make_namelist())

prog = self.getexe('hwrf_geogrid')
log = self._section + '.log'
cmd = produtil.run.mpirun(produtil.run.mpi(prog),
allranks=True)
if self.redirect: cmd=cmd > log
logger.info('%s command: %s'%(self.taskname, repr(cmd),))
produtil.run.checkrun(cmd,logger=logger) #第635行
findme="Successful completion"
geogrid_log=None
for glog in ( 'geogrid.log', 'geogrid.log.0000',
'geogrid.log.00000' ):
if os.path.exists(glog):
geogrid_log=glog
else:
logger.info('%s: does not exist.'%(glog,))
if geogrid_log is None:
msg='WPS Geogrid failed: could not find geogrid log file.'
logger.error(msg)
self.state=FAILED
raise GeogridNoLog(msg)
logger.info('%s: will check for %s'%(geogrid_log,findme))
if not check_last_lines(geogrid_log,findme):
raise WPSError('%s: did not find "%s"'
%(geogrid_log,findme))
self.deliver_products(keep=False,relink=True)

init.py:
class HWRFInit(HWRFTask):
......
def run_through_anl(self):
"""!Runs the following jobs, if they are enabled: geogrid,
ungrib, metgrid, init-length prep_hybrid, init-length real_nmm
and wrfanl."""
self.log().warning('run_through_anl')
if 'geogrid' in self.__dict__:
set_ecflow_label('status','[geogrid] process terrain')
self.geogrid.run() #第559行
set_ecflow_label('status','[ungrib] process parent GRIB')
self.ungrib.run()
set_ecflow_label('status','[metgrid] interpolate horizontally')
self.metgrid.run()
if 'prep' in self.__dict__:
set_ecflow_label('status','[prep] process parent spectral')
for i in xrange(len(self.ibdytimes)):
t=self.ibdytimes[i]
if t>self.initlen: break
self.prep.run_ipiece(i)
set_ecflow_label('status','[realinit] make wrfinput file')
self.realinit.run()
set_ecflow_label('status','[runwrfanl] make wrfanl files')
self.runwrfanl.run()
set_ecflow_label('status','Passed through anl.')

[weixin_45907705]

脚本太长了，有没有办法上传整个脚本呢？
run.py

def checkrun(arg,logger=None,**kwargs):
    """!This is a simple wrapper round run that raises
    ExitStatusException if the program exit status is non-zero.

    @param arg the produtil.prog.Runner to execute (output of
      exe(), bigexe() or mpirun()
    @param logger a logging.Logger to log messages
    @param kwargs The optional run=[] argument can provide a different
    list of acceptable exit statuses."""
    r=run(arg,logger=logger)
    if kwargs is not None and 'ret' in kwargs:
        if not r in kwargs['ret']:
            raise ExitStatusException('%s: unexpected exit status'%(repr(arg),),r)
    elif not r==0:
        raise ExitStatusException('%s: non-zero exit status'%(repr(arg),),r)   #第415行
    return r

wps.py

class Geogrid(WPSTask):
    .....
    def run(self):
        """!Copies inputs, links fix files, runs geogrid and delivers
        results."""
        logger=self.log()
        try:
            produtil.fileop.makedirs(self.outdir)
            with NamedDir(self.location) as dir:
                logger.info('Geogrid running in directory: '+os.getcwd())
                assert(not re.match('\A/tmp',os.getcwd()))

                for f in glob.glob('geo*'):
                    try:
                        produtil.fileop.remove_file(f,logger=logger)
                    except(EnvironmentError) as e:
                        logger.warning('%s: did not remove file, but '
                                       'continuing anyway'%(f,))
                self.link_fix(geog_data=True,table='GEOGRID.TBL')

                with open('namelist.wps', 'w') as f:
                    f.write(self.make_namelist())

                prog = self.getexe('hwrf_geogrid')
                log = self._section + '.log'
                cmd = produtil.run.mpirun(produtil.run.mpi(prog),
                                          allranks=True)
                if self.redirect: cmd=cmd > log
                logger.info('%s command: %s'%(self.taskname, repr(cmd),))
                produtil.run.checkrun(cmd,logger=logger)   #第635行
                findme="Successful completion"
                geogrid_log=None
                for glog in ( 'geogrid.log', 'geogrid.log.0000',
                              'geogrid.log.00000' ):
                    if os.path.exists(glog):
                        geogrid_log=glog
                    else:
                        logger.info('%s: does not exist.'%(glog,))
                if geogrid_log is None:
                    msg='WPS Geogrid failed: could not find geogrid log file.'
                    logger.error(msg)
                    self.state=FAILED
                    raise GeogridNoLog(msg)
                logger.info('%s: will check for %s'%(geogrid_log,findme))
                if not check_last_lines(geogrid_log,findme):
                    raise WPSError('%s: did not find "%s"'
                                   %(geogrid_log,findme))
                self.deliver_products(keep=False,relink=True)

init.py

class HWRFInit(HWRFTask):
    ......
    def run_through_anl(self):
        """!Runs the following jobs, if they are enabled: geogrid,
        ungrib, metgrid, init-length prep_hybrid, init-length real_nmm
        and wrfanl."""
        self.log().warning('run_through_anl')
        if 'geogrid' in self.__dict__:
            set_ecflow_label('status','[geogrid] process terrain')
            self.geogrid.run()    #第559行
        set_ecflow_label('status','[ungrib] process parent GRIB')
        self.ungrib.run()
        set_ecflow_label('status','[metgrid] interpolate horizontally')
        self.metgrid.run()
        if 'prep' in self.__dict__:
            set_ecflow_label('status','[prep] process parent spectral')
            for i in xrange(len(self.ibdytimes)):
                t=self.ibdytimes[i]
                if t>self.initlen: break
                self.prep.run_ipiece(i)
        set_ecflow_label('status','[realinit] make wrfinput file')
        self.realinit.run()
        set_ecflow_label('status','[runwrfanl] make wrfanl files')
        self.runwrfanl.run()
        set_ecflow_label('status','Passed through anl.')

[WilliamHuangSt]

csdn有个东西叫代码框，直接复制还是python这种语言，很难看清的。

[WilliamHuangSt]

引用 1 楼 WilliamHuangSt的回复:

csdn有个东西叫代码框，直接复制还是python这种语言，很难看清的。

而且我在用手机刷，看的更辛苦。