1 |
"""
|
2 |
Implements the Remote Glidein scheduler
|
3 |
"""
|
4 |
|
5 |
from SchedulerGrid import SchedulerGrid
|
6 |
from crab_exceptions import CrabException
|
7 |
from crab_util import runCommand
|
8 |
from ServerConfig import *
|
9 |
from WMCore.SiteScreening.BlackWhiteListParser import SEBlackWhiteListParser
|
10 |
import Scram
|
11 |
|
12 |
import common
|
13 |
import os
|
14 |
import socket
|
15 |
import re
|
16 |
import commands
|
17 |
|
18 |
# FUTURE: for python 2.4 & 2.6
|
19 |
try:
|
20 |
from hashlib import sha1
|
21 |
except:
|
22 |
from sha import sha as sha1
|
23 |
|
24 |
class SchedulerRemoteglidein(SchedulerGrid) :
|
25 |
"""
|
26 |
Class to implement the vanilla remote Condor scheduler
|
27 |
for a glidein frontend, the created JDL will not work
|
28 |
on valilla local condor.
|
29 |
Naming convention: Methods starting with 'ws' provide
|
30 |
the corresponding part of the job script
|
31 |
('ws' stands for 'write script').
|
32 |
"""
|
33 |
|
34 |
def __init__(self):
|
35 |
SchedulerGrid.__init__(self,"REMOTEGLIDEIN")
|
36 |
|
37 |
self.datasetPath = None
|
38 |
self.selectNoInput = None
|
39 |
self.OSBsize = 50*1000*1000 # 50 MB
|
40 |
|
41 |
self.environment_unique_identifier = None
|
42 |
self.submissionDay = time.strftime("%y%m%d",time.localtime())
|
43 |
|
44 |
return
|
45 |
|
46 |
|
47 |
def configure(self, cfg_params):
|
48 |
"""
|
49 |
Configure the scheduler with the config settings from the user
|
50 |
"""
|
51 |
|
52 |
SchedulerGrid.configure(self, cfg_params)
|
53 |
|
54 |
self.proxyValid=0
|
55 |
self.dontCheckProxy=int(cfg_params.get("GRID.dont_check_proxy",'0'))
|
56 |
self.space_token = cfg_params.get("USER.space_token",None)
|
57 |
self.proxyServer= 'myproxy.cern.ch'
|
58 |
self.group = cfg_params.get("GRID.group", None)
|
59 |
self.role = cfg_params.get("GRID.role", None)
|
60 |
self.VO = cfg_params.get('GRID.virtual_organization','cms')
|
61 |
self.allowOverflow = cfg_params.get('GRID.allow_overflow', '1')
|
62 |
self.max_rss = cfg_params.get('GRID.max_rss','2000')
|
63 |
|
64 |
self.checkProxy()
|
65 |
|
66 |
|
67 |
try:
|
68 |
tmp = cfg_params['CMSSW.datasetpath']
|
69 |
if tmp.lower() == 'none':
|
70 |
self.datasetPath = None
|
71 |
self.selectNoInput = 1
|
72 |
else:
|
73 |
self.datasetPath = tmp
|
74 |
self.selectNoInput = 0
|
75 |
except KeyError:
|
76 |
msg = "Error: datasetpath not defined "
|
77 |
raise CrabException(msg)
|
78 |
|
79 |
if cfg_params.get('GRID.ce_black_list', None) or \
|
80 |
cfg_params.get('GRID.ce_white_list', None) :
|
81 |
msg="BEWARE: scheduler REMOTEGLIDEIN ignores CE black/white lists."
|
82 |
msg+="\n Remove them from crab configuration to proceed."
|
83 |
msg+="\n Use GRID.se_white_list and/or GRID.se_black_list instead"
|
84 |
raise CrabException(msg)
|
85 |
|
86 |
|
87 |
# make sure proxy FQAN has not changed since last time
|
88 |
command = "voms-proxy-info -identity -fqan 2>/dev/null"
|
89 |
command += " | head -2"
|
90 |
identity = runCommand(command)
|
91 |
idfile = common.work_space.shareDir() + "GridIdentity"
|
92 |
if os.access(idfile, os.F_OK) :
|
93 |
# identity file exists from previous commands
|
94 |
f=open(idfile, 'r')
|
95 |
idFromFile=f.read()
|
96 |
f.close()
|
97 |
else :
|
98 |
# create it
|
99 |
f=open(idfile, 'w')
|
100 |
f.write(identity)
|
101 |
f.close()
|
102 |
idFromFile = identity
|
103 |
|
104 |
if identity != idFromFile:
|
105 |
msg = "Wrong Grid Credentials:\n%s" % identity
|
106 |
msg += "\nMake sure you have "
|
107 |
msg += " DN, FQAN =\n%s" % idFromFile
|
108 |
raise CrabException(msg)
|
109 |
|
110 |
return
|
111 |
|
112 |
def userName(self):
|
113 |
""" return the user name """
|
114 |
tmp=runCommand("voms-proxy-info -identity 2>/dev/null")
|
115 |
return tmp.strip()
|
116 |
|
117 |
def envUniqueID(self):
|
118 |
taskHash = sha1(common._db.queryTask('name')).hexdigest()
|
119 |
id = "https://" + socket.gethostname() + '/' + taskHash + "/${NJob}"
|
120 |
return id
|
121 |
|
122 |
def sched_parameter(self, i, task):
|
123 |
"""
|
124 |
Return scheduler-specific parameters. Used at crab -submit time
|
125 |
by $CRABPYTHON/Scheduler.py
|
126 |
"""
|
127 |
|
128 |
#SB paste from crab SchedulerGlidein
|
129 |
|
130 |
jobParams = ""
|
131 |
|
132 |
(self.remoteHost,self.remoteUserHost) = self.pickRemoteSubmissionHost(task)
|
133 |
|
134 |
seDest = task.jobs[i-1]['dlsDestination']
|
135 |
|
136 |
if seDest == [''] :
|
137 |
seDest = self.blackWhiteListParser.expandList("T") # all of SiteDB
|
138 |
|
139 |
seString=self.blackWhiteListParser.cleanForBlackWhiteList(seDest)
|
140 |
|
141 |
jobParams += '+DESIRED_SEs = "'+seString+'"; '
|
142 |
|
143 |
scram = Scram.Scram(None)
|
144 |
cmsVersion = scram.getSWVersion()
|
145 |
scramArch = scram.getArch()
|
146 |
|
147 |
cmsver=re.split('_', cmsVersion)
|
148 |
numericCmsVersion = "%s%.2d%.2d" %(cmsver[1], int(cmsver[2]), int(cmsver[3]))
|
149 |
|
150 |
jobParams += '+DESIRED_CMSVersion ="' +cmsVersion+'";'
|
151 |
jobParams += '+DESIRED_CMSVersionNr ="' +numericCmsVersion+'";'
|
152 |
jobParams += '+DESIRED_CMSScramArch ="' +scramArch+'";'
|
153 |
|
154 |
myscheddName = self.remoteHost
|
155 |
jobParams += '+Glidein_MonitorID = "https://'+ myscheddName + \
|
156 |
'//' + self.submissionDay + '//$(Cluster).$(Process)"; '
|
157 |
|
158 |
if (self.EDG_clock_time):
|
159 |
glideinTime = "%d" % (int(self.EDG_clock_time)+5) # 5 min to wrapup
|
160 |
jobParams += '+MaxWallTimeMins = '+ glideinTime + '; '
|
161 |
else:
|
162 |
jobParams += '+MaxWallTimeMins = %d; ' % (60*22 - 5) # 22h default in glidein, 5min to wrap
|
163 |
|
164 |
if self.max_rss :
|
165 |
jobParams += 'request_memory = '+self.max_rss+';'
|
166 |
|
167 |
if self.allowOverflow == "0":
|
168 |
jobParams += '+CMS_ALLOW_OVERFLOW = False; '
|
169 |
|
170 |
if self.EDG_addJdlParam:
|
171 |
if self.EDG_addJdlParam[-1] == '':
|
172 |
self.EDG_addJdlParam = self.EDG_addJdlParam[:-1]
|
173 |
for p in self.EDG_addJdlParam:
|
174 |
jobParams += p.strip()+';\n'
|
175 |
|
176 |
common._db.updateTask_({'jobType':jobParams})
|
177 |
|
178 |
|
179 |
return jobParams
|
180 |
|
181 |
|
182 |
def realSchedParams(self, cfg_params):
|
183 |
"""
|
184 |
Return dictionary with specific parameters, to use with real scheduler
|
185 |
is called when scheduler is initialized in Boss, i.e. at each crab command
|
186 |
"""
|
187 |
#SB this method is used to pass directory names to Boss Scheduler
|
188 |
# via params dictionary
|
189 |
|
190 |
jobDir = common.work_space.jobDir()
|
191 |
taskDir=common.work_space.topDir().split('/')[-2]
|
192 |
shareDir = common.work_space.shareDir()
|
193 |
|
194 |
params = {'shareDir':shareDir,
|
195 |
'jobDir':jobDir,
|
196 |
'taskDir':taskDir,
|
197 |
'submissionDay':self.submissionDay}
|
198 |
|
199 |
return params
|
200 |
|
201 |
|
202 |
def listMatch(self, seList, full):
|
203 |
"""
|
204 |
Check the compatibility of available resources
|
205 |
"""
|
206 |
|
207 |
return [True]
|
208 |
|
209 |
|
210 |
def decodeLogInfo(self, fileName):
|
211 |
"""
|
212 |
Parse logging info file and return main info
|
213 |
"""
|
214 |
|
215 |
import CondorGLoggingInfo
|
216 |
loggingInfo = CondorGLoggingInfo.CondorGLoggingInfo()
|
217 |
reason = loggingInfo.decodeReason(fileName)
|
218 |
return reason
|
219 |
|
220 |
|
221 |
# def wsCopyOutput(self):
|
222 |
# """
|
223 |
# Write a CopyResults part of a job script, e.g.
|
224 |
# to copy produced output into a storage element.
|
225 |
# """
|
226 |
# txt = self.wsCopyOutput()
|
227 |
# return txt
|
228 |
|
229 |
|
230 |
def wsExitFunc(self):
|
231 |
"""
|
232 |
Returns the part of the job script which runs prior to exit
|
233 |
"""
|
234 |
|
235 |
txt = '\n'
|
236 |
txt += '#\n'
|
237 |
txt += '# EXECUTE THIS FUNCTION BEFORE EXIT \n'
|
238 |
txt += '#\n\n'
|
239 |
|
240 |
txt += 'func_exit() { \n'
|
241 |
txt += self.wsExitFunc_common()
|
242 |
txt += '#Check for stdout/err in new location as of condor 7.7\n'
|
243 |
txt += ' if [ -s _condor_stdout ]; then\n'
|
244 |
txt += ' echo "Found _condor_stdout/err, rename for OSB"\n'
|
245 |
txt += ' cp -pfv _condor_stdout CMSSW_${NJob}.stdout\n'
|
246 |
txt += ' cp -pfv _condor_stderr CMSSW_${NJob}.stderr\n'
|
247 |
txt += ' fi\n'
|
248 |
txt += ' tar zcvf ${out_files}.tgz ${final_list}\n'
|
249 |
txt += ' tmp_size=`ls -gGrta ${out_files}.tgz | awk \'{ print $3 }\'`\n'
|
250 |
txt += ' rm ${out_files}.tgz\n'
|
251 |
txt += ' size=`expr $tmp_size`\n'
|
252 |
txt += ' echo "Total Output dimension: $size"\n'
|
253 |
txt += ' limit='+str(self.OSBsize) +' \n'
|
254 |
txt += ' echo "WARNING: output files size limit is set to: $limit"\n'
|
255 |
txt += ' if [ "$limit" -lt "$size" ]; then\n'
|
256 |
txt += ' exceed=1\n'
|
257 |
txt += ' job_exit_code=70000\n'
|
258 |
txt += ' echo "Output Sanbox too big. Produced output is lost "\n'
|
259 |
txt += ' else\n'
|
260 |
txt += ' exceed=0\n'
|
261 |
txt += ' echo "Total Output dimension $size is fine."\n'
|
262 |
txt += ' fi\n'
|
263 |
|
264 |
txt += ' echo "JOB_EXIT_STATUS = $job_exit_code"\n'
|
265 |
txt += ' echo "JobExitCode=$job_exit_code" >> $RUNTIME_AREA/$repo\n'
|
266 |
txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
|
267 |
txt += ' if [ $exceed -ne 1 ]; then\n'
|
268 |
txt += ' tar zcvf ${out_files}.tgz ${final_list}\n'
|
269 |
txt += ' else\n'
|
270 |
txt += ' tar zcvf ${out_files}.tgz CMSSW_${NJob}.stdout CMSSW_${NJob}.stderr\n'
|
271 |
txt += ' fi\n'
|
272 |
txt += ' python $RUNTIME_AREA/fillCrabFjr.py $RUNTIME_AREA/crab_fjr_$NJob.xml --errorcode $job_exit_code \n'
|
273 |
|
274 |
txt += ' exit $job_exit_code\n'
|
275 |
txt += '}\n'
|
276 |
|
277 |
return txt
|
278 |
|
279 |
|
280 |
def sched_fix_parameter(self):
|
281 |
"""
|
282 |
Returns string with requirements and scheduler-specific parameters
|
283 |
"""
|
284 |
|
285 |
if self.EDG_requirements:
|
286 |
req = self.EDG_requirements
|
287 |
taskReq = {'commonRequirements':req}
|
288 |
common._db.updateTask_(taskReq)
|
289 |
|
290 |
def pickRemoteSubmissionHost(self, task):
|
291 |
|
292 |
task = common._db.getTask()
|
293 |
|
294 |
if task['serverName']!=None and task['serverName']!="":
|
295 |
# remoteHost is already defined and stored for this task
|
296 |
# so pick it from DB
|
297 |
# cast to string to avoid issues with unicode :-(
|
298 |
remoteUserHost=str(task['serverName'])
|
299 |
common.logger.info("serverName from Task DB is %s" %
|
300 |
remoteUserHost)
|
301 |
if '@' in remoteUserHost:
|
302 |
remoteHost = remoteUserHost.split('@')[1]
|
303 |
else:
|
304 |
remoteHost = remoteUserHost
|
305 |
else:
|
306 |
if self.cfg_params.has_key('CRAB.submit_host'):
|
307 |
# get a remote submission host from crab config file
|
308 |
srvCfg=ServerConfig(self.cfg_params['CRAB.submit_host']).config()
|
309 |
remoteHost=srvCfg['serverName']
|
310 |
common.logger.info("remotehost from crab.cfg = %s" % remoteHost)
|
311 |
else:
|
312 |
# pick from Available Servers List
|
313 |
srvCfg=ServerConfig('default').config()
|
314 |
remoteHost = srvCfg['serverName']
|
315 |
common.logger.info("remotehost from Avail.List = %s" % remoteHost)
|
316 |
|
317 |
if not remoteHost:
|
318 |
raise CrabException('FATAL ERROR: remoteHost not defined')
|
319 |
|
320 |
#common.logger.info("try to find out username for remote Host via uberftp ...")
|
321 |
#command="uberftp %s pwd|grep User|awk '{print $3}'" % remoteHost
|
322 |
#(status, output) = commands.getstatusoutput(command)
|
323 |
#if status == 0:
|
324 |
# remoteUser = output
|
325 |
# common.logger.info("remoteUser set to %s" % remoteUser)
|
326 |
# if remoteUser==None:
|
327 |
# raise CrabException('FATAL ERROR: REMOTE USER not defined')
|
328 |
|
329 |
#remoteUserHost = remoteUser + '@' + remoteHost
|
330 |
remoteUserHost = remoteHost
|
331 |
|
332 |
common._db.updateTask_({'serverName':remoteUserHost})
|
333 |
|
334 |
return (remoteHost, remoteUserHost)
|