CRAB/python/JobDB.py

from WorkSpace import WorkSpace
from crab_exceptions import *
import common

import os, string

class dbEntry:
    def __init__(self):
        self.status = 'X'       # job status
        self.exitStatus = ''    # job exit status
        self.jid = ''           # scheduler job id
        self.bossid = ''        # BOSS job id
        self.collections = []   # EvCollection to be analyzed in this job
        self.inputSandbox = []  # InputSandbox
        self.outputSandbox = [] # OutputSandbox
        self.taskId = ''        # Task job belongs to
        self.block = ''         # Task job belongs to
        self.arguments = []     # abstract job_type parameters
        self.dest = []          # Destination for this job according to DLS
        return

    def __str__(self):
        txt  = 'Status <' + self.status + '>; '
        if self.exitStatus!='':
            txt += 'exitStatus <' + str(self.exitStatus) + '>\n'
        txt += 'Job Id <' + self.jid + '> Block <'+self.block+'>\n'
        if self.arguments:
            txt += 'Job Type Arguments <' + str(self.arguments) + '>\n'
        if self.dest:
            txt += 'Destination <' + str(self.dest) + '>\n'

        return txt

class JobDB:
    def __init__(self):
        self._dir = common.work_space.shareDir() + 'db/'
        self._db_fname = 'jobs'
        self._jobs = []        # list of dbEntry's
        return

    def __str__(self):
        njobs = self.nJobs()
        if njobs == 1: plural = ''
        else:          plural = 's'
        txt = 'Total of %d job%s:\n' % (njobs, plural)
        for i in range(njobs):
            txt += ('Job %03d' % (i+1)) + ': '
            txt += str(self._jobs[i])
            pass
        return txt

    def dump(self, jobs):
        njobs = len(jobs)
        if njobs == 1: plural = ''
        else:          plural = 's'
        print 'Listing %d job%s:\n' % (njobs, plural)
        for job in jobs:
            print ('Job %03d' % (job)) + ': ' + str(self._jobs[job-1])
            pass

    def nJobs(self):
        return len(self._jobs)

    def nSubmittedJobs(self):
        n = 0
        for i in range(self.nJobs()):
           if self.status(i) in ['S', 'R']:
               n += 1
        return n

    def create(self, njobs):

        if os.path.exists(self._dir):
            msg = 'Cannot create Job DB: already exists.'
            raise CrabException(msg)

        os.mkdir(self._dir)

        for i in range(njobs):
            self._jobs.append(dbEntry())
            pass

        common.logger.debug(5,'Created DB for '+str(njobs)+' jobs')

        self.save()
        return

    def save(self):
        try:
            db_file = open(self._dir+self._db_fname, 'w')
        except IOError:
            raise CrabException("Cannot open "+self._dir+self._db_fname+"\n")
        for i in range(len(self._jobs)):
            db_file.write(str(i)+'|')
            db_file.write(self._jobs[i].status+'|')
            db_file.write(self._jobs[i].exitStatus+'|')
            db_file.write(self._jobs[i].jid+'|')
            db_file.write(self._jobs[i].bossid+'|')
            db_file.write(string.join(self._jobs[i].collections)+'|')
            db_file.write(string.join(self._jobs[i].inputSandbox)+'|')
            db_file.write(string.join(self._jobs[i].outputSandbox)+'|')
            db_file.write(str(self._jobs[i].taskId)+'|')
            db_file.write(str(self._jobs[i].block)+'|')
            db_file.write(string.join(self._jobs[i].arguments)+'|')
            db_file.write(string.join(self._jobs[i].dest)+'|')
            db_file.write('\n')
            pass
        db_file.close()
        return

    def load(self):
        self._jobs = []
        try:
            db_file = open(self._dir+self._db_fname, 'r')
        except IOError:
            raise DBException("Something really serious! no JobDB is present!!!")

        for line in db_file:
            db_entry = dbEntry()
            (n, db_entry.status, db_entry.exitStatus, db_entry.jid, db_entry.bossid, collectionsTMP,  inputSandboxTMP , outputSandboxTMP , db_entry.taskId, db_entry.block, argumentsTMP, destTMP, rest) = string.split(line, '|')
            db_entry.collections = string.split(collectionsTMP)
            db_entry.inputSandbox = string.split(inputSandboxTMP)
            db_entry.outputSandbox = string.split(outputSandboxTMP)
            db_entry.arguments = string.split(argumentsTMP)
            db_entry.dest = string.split(destTMP)
            self._jobs.append(db_entry)
            pass
        db_file.close()
        return
    
    
    def setStatus(self, nj, status):
        self.check(nj)
        self._jobs[int(nj)].status = status
        return
    
    def status(self, nj):
        return self._jobs[int(nj)].status
    
    def setExitStatus(self, nj, exitStatus):
        self.check(nj)
        self._jobs[int(nj)].exitStatus = exitStatus
        return
    
    def exitStatus(self, nj):
        return self._jobs[int(nj)].exitStatus
    
    def setJobId(self, nj, jid):
        self.check(nj)
        self._jobs[int(nj)].jid = jid
        return
    
    def jobId(self, nj):
        return self._jobs[int(nj)].jid

    def setBossId(self, nj, bossid):
        self.check(nj)
        self._jobs[int(nj)].bossid = bossid
        return
    
    def bossId(self, nj):
        return self._jobs[int(nj)].bossid

    def setArguments(self, nj, args):
        self.check(nj)
        self._jobs[int(nj)].arguments = args
        return
    
    def arguments(self, nj):
        return self._jobs[int(nj)].arguments

    def setCollections(self, nj, Collections):
        self.check(nj)
        self._jobs[int(nj)].Collections = Collections
        return
    
    def collections(self, nj):
        return self._jobs[int(nj)].collections

    def setInputSandbox(self, nj, InputSandbox):
        self.check(nj)
        self._jobs[int(nj)].inputSandbox = InputSandbox
        return
    
    def inputSandbox(self, nj):
        return self._jobs[int(nj)].inputSandbox

    def setOutputSandbox(self, nj, OutputSandbox):
        self.check(nj)
        self._jobs[int(nj)].outputSandbox = OutputSandbox
        return
    
    def outputSandbox(self, nj):
        return self._jobs[int(nj)].outputSandbox
        
    def setTaskId(self, nj, taskId):
        self.check(nj)
        self._jobs[int(nj)].taskId = taskId

    def taskId(self, nj):
        return self._jobs[int(nj)].taskId

    def setBlock(self, nj, block):
        self.check(nj)
        self._jobs[int(nj)].block = block

    def block(self, nj):
        return self._jobs[int(nj)].block

    def setDestination(self, nj, args):
        self.check(nj)
        self._jobs[int(nj)].dest = args
        return
    
    def destination(self, nj):
        return self._jobs[int(nj)].dest

    def check(self, nj):
        """ Check if the job nj is already present in DB (first job is 0) and create it if not """
        if (int(nj) >= self.nJobs()): self._jobs.append(dbEntry())
Revision:	1.26
Committed:	Mon Mar 3 18:56:39 2008 UTC (17 years, 2 months ago) by spiga
Content type:	text/x-python
Branch:	MAIN
CVS Tags:	HEAD
Changes since 1.25:	+0 -0 lines
State:	*FILE REMOVED*
Log Message:	jobdb and taskdb are obsolete with bosslite
#	User	Rev	Content
1	nsmirnov	1.1	from WorkSpace import WorkSpace
2			from crab_exceptions import *
3			import common
4
5	nsmirnov	1.2	import os, string
6	nsmirnov	1.1
7			class dbEntry:
8			def __init__(self):
9	slacapra	1.5	self.status = 'X' # job status
10	slacapra	1.8	self.exitStatus = '' # job exit status
11	slacapra	1.5	self.jid = '' # scheduler job id
12	corvo	1.10	self.bossid = '' # BOSS job id
13	slacapra	1.5	self.collections = [] # EvCollection to be analyzed in this job
14			self.inputSandbox = [] # InputSandbox
15			self.outputSandbox = [] # OutputSandbox
16	corvo	1.10	self.taskId = '' # Task job belongs to
17	slacapra	1.22	self.block = '' # Task job belongs to
18			self.arguments = [] # abstract job_type parameters
19	gutsche	1.18	self.dest = [] # Destination for this job according to DLS
20	nsmirnov	1.1	return
21
22	nsmirnov	1.2	def __str__(self):
23			txt = 'Status <' + self.status + '>; '
24	slacapra	1.8	if self.exitStatus!='':
25			txt += 'exitStatus <' + str(self.exitStatus) + '>\n'
26	slacapra	1.22	txt += 'Job Id <' + self.jid + '> Block <'+self.block+'>\n'
27	gutsche	1.13	if self.arguments:
28			txt += 'Job Type Arguments <' + str(self.arguments) + '>\n'
29	slacapra	1.19	if self.dest:
30			txt += 'Destination <' + str(self.dest) + '>\n'
31	slacapra	1.4
32	nsmirnov	1.1	return txt
33
34			class JobDB:
35			def __init__(self):
36			self._dir = common.work_space.shareDir() + 'db/'
37			self._db_fname = 'jobs'
38			self._jobs = [] # list of dbEntry's
39			return
40
41	nsmirnov	1.2	def __str__(self):
42	nsmirnov	1.1	njobs = self.nJobs()
43			if njobs == 1: plural = ''
44			else: plural = 's'
45			txt = 'Total of %d job%s:\n' % (njobs, plural)
46			for i in range(njobs):
47	nsmirnov	1.3	txt += ('Job %03d' % (i+1)) + ': '
48	nsmirnov	1.1	txt += str(self._jobs[i])
49			pass
50			return txt
51
52	slacapra	1.4	def dump(self, jobs):
53			njobs = len(jobs)
54			if njobs == 1: plural = ''
55			else: plural = 's'
56			print 'Listing %d job%s:\n' % (njobs, plural)
57			for job in jobs:
58	slacapra	1.17	print ('Job %03d' % (job)) + ': ' + str(self._jobs[job-1])
59	slacapra	1.4	pass
60
61	nsmirnov	1.1	def nJobs(self):
62			return len(self._jobs)
63
64	corvo	1.23	def nSubmittedJobs(self):
65			n = 0
66			for i in range(self.nJobs()):
67	corvo	1.24	if self.status(i) in ['S', 'R']:
68	corvo	1.23	n += 1
69			return n
70
71	nsmirnov	1.1	def create(self, njobs):
72
73			if os.path.exists(self._dir):
74			msg = 'Cannot create Job DB: already exists.'
75			raise CrabException(msg)
76
77			os.mkdir(self._dir)
78
79			for i in range(njobs):
80			self._jobs.append(dbEntry())
81			pass
82
83	slacapra	1.4	common.logger.debug(5,'Created DB for '+str(njobs)+' jobs')
84
85	nsmirnov	1.1	self.save()
86			return
87
88			def save(self):
89	slacapra	1.25	try:
90			db_file = open(self._dir+self._db_fname, 'w')
91			except IOError:
92			raise CrabException("Cannot open "+self._dir+self._db_fname+"\n")
93	nsmirnov	1.1	for i in range(len(self._jobs)):
94	gutsche	1.21	db_file.write(str(i)+'\|')
95	slacapra	1.20	db_file.write(self._jobs[i].status+'\|')
96			db_file.write(self._jobs[i].exitStatus+'\|')
97			db_file.write(self._jobs[i].jid+'\|')
98			db_file.write(self._jobs[i].bossid+'\|')
99			db_file.write(string.join(self._jobs[i].collections)+'\|')
100			db_file.write(string.join(self._jobs[i].inputSandbox)+'\|')
101			db_file.write(string.join(self._jobs[i].outputSandbox)+'\|')
102			db_file.write(str(self._jobs[i].taskId)+'\|')
103	slacapra	1.22	db_file.write(str(self._jobs[i].block)+'\|')
104	slacapra	1.20	db_file.write(string.join(self._jobs[i].arguments)+'\|')
105			db_file.write(string.join(self._jobs[i].dest)+'\|')
106	nsmirnov	1.1	db_file.write('\n')
107			pass
108			db_file.close()
109			return
110
111			def load(self):
112			self._jobs = []
113	slacapra	1.6	try:
114			db_file = open(self._dir+self._db_fname, 'r')
115			except IOError:
116	slacapra	1.7	raise DBException("Something really serious! no JobDB is present!!!")
117	slacapra	1.6
118	nsmirnov	1.1	for line in db_file:
119	nsmirnov	1.2	db_entry = dbEntry()
120	slacapra	1.22	(n, db_entry.status, db_entry.exitStatus, db_entry.jid, db_entry.bossid, collectionsTMP, inputSandboxTMP , outputSandboxTMP , db_entry.taskId, db_entry.block, argumentsTMP, destTMP, rest) = string.split(line, '\|')
121	slacapra	1.16	db_entry.collections = string.split(collectionsTMP)
122			db_entry.inputSandbox = string.split(inputSandboxTMP)
123			db_entry.outputSandbox = string.split(outputSandboxTMP)
124			db_entry.arguments = string.split(argumentsTMP)
125	gutsche	1.18	db_entry.dest = string.split(destTMP)
126	nsmirnov	1.1	self._jobs.append(db_entry)
127			pass
128			db_file.close()
129			return
130
131	slacapra	1.5
132	nsmirnov	1.1	def setStatus(self, nj, status):
133	slacapra	1.20	self.check(nj)
134	gutsche	1.11	self._jobs[int(nj)].status = status
135	nsmirnov	1.1	return
136
137			def status(self, nj):
138	gutsche	1.11	return self._jobs[int(nj)].status
139	nsmirnov	1.1
140	slacapra	1.8	def setExitStatus(self, nj, exitStatus):
141	slacapra	1.20	self.check(nj)
142	gutsche	1.11	self._jobs[int(nj)].exitStatus = exitStatus
143	slacapra	1.8	return
144
145			def exitStatus(self, nj):
146	gutsche	1.11	return self._jobs[int(nj)].exitStatus
147	slacapra	1.8
148	nsmirnov	1.1	def setJobId(self, nj, jid):
149	slacapra	1.20	self.check(nj)
150	gutsche	1.11	self._jobs[int(nj)].jid = jid
151	nsmirnov	1.1	return
152
153			def jobId(self, nj):
154	gutsche	1.11	return self._jobs[int(nj)].jid
155	slacapra	1.4
156	fanzago	1.9	def setBossId(self, nj, bossid):
157	slacapra	1.20	self.check(nj)
158	gutsche	1.11	self._jobs[int(nj)].bossid = bossid
159	fanzago	1.9	return
160
161			def bossId(self, nj):
162	gutsche	1.11	return self._jobs[int(nj)].bossid
163	fanzago	1.9
164	gutsche	1.13	def setArguments(self, nj, args):
165	slacapra	1.20	self.check(nj)
166	gutsche	1.13	self._jobs[int(nj)].arguments = args
167	slacapra	1.4	return
168
169	gutsche	1.13	def arguments(self, nj):
170			return self._jobs[int(nj)].arguments
171	slacapra	1.4
172			def setCollections(self, nj, Collections):
173	slacapra	1.20	self.check(nj)
174	gutsche	1.11	self._jobs[int(nj)].Collections = Collections
175	slacapra	1.4	return
176
177			def collections(self, nj):
178	gutsche	1.11	return self._jobs[int(nj)].collections
179	slacapra	1.5
180			def setInputSandbox(self, nj, InputSandbox):
181	slacapra	1.20	self.check(nj)
182	gutsche	1.11	self._jobs[int(nj)].inputSandbox = InputSandbox
183	slacapra	1.5	return
184
185			def inputSandbox(self, nj):
186	gutsche	1.11	return self._jobs[int(nj)].inputSandbox
187	slacapra	1.5
188			def setOutputSandbox(self, nj, OutputSandbox):
189	slacapra	1.20	self.check(nj)
190	gutsche	1.11	self._jobs[int(nj)].outputSandbox = OutputSandbox
191	slacapra	1.5	return
192
193			def outputSandbox(self, nj):
194	gutsche	1.11	return self._jobs[int(nj)].outputSandbox
195	corvo	1.10
196			def setTaskId(self, nj, taskId):
197	slacapra	1.20	self.check(nj)
198	gutsche	1.11	self._jobs[int(nj)].taskId = taskId
199	gutsche	1.18
200	slacapra	1.22	def taskId(self, nj):
201	slacapra	1.20	return self._jobs[int(nj)].taskId
202
203	slacapra	1.22	def setBlock(self, nj, block):
204			self.check(nj)
205			self._jobs[int(nj)].block = block
206
207			def block(self, nj):
208			return self._jobs[int(nj)].block
209
210	gutsche	1.18	def setDestination(self, nj, args):
211	slacapra	1.20	self.check(nj)
212	gutsche	1.18	self._jobs[int(nj)].dest = args
213			return
214
215			def destination(self, nj):
216			return self._jobs[int(nj)].dest
217	slacapra	1.20
218			def check(self, nj):
219			""" Check if the job nj is already present in DB (first job is 0) and create it if not """
220			if (int(nj) >= self.nJobs()): self._jobs.append(dbEntry())