def aggregateLogSummary():
"""
Create an aggregate of log summary files, write it in an aggregate log
summary file and in the end return the values
"""
global monitorAggregatorConfig
# initialize global counters
global_total={'Current':{},'Entered':{},'Exited':{},'CompletedCounts':{'Sum':{},'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{}}}
for s in ('Wait','Idle','Running','Held'):
for k in ['Current','Entered','Exited']:
global_total[k][s]=0
for s in ('Completed','Removed'):
for k in ['Entered']:
global_total[k][s]=0
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
el={}
for t in glideFactoryMonitoring.getAllMillRanges():
el[t]=0
global_total['CompletedCounts'][w][k]=el
el={}
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t]=0
global_total['CompletedCounts']['Lasted']=el
el={}
for t in glideFactoryMonitoring.getAllJobRanges():
el[t]=0
global_total['CompletedCounts']['JobsNr']=el
el={}
# KEL - why is the same el used twice (see above)
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t]=0
global_total['CompletedCounts']['JobsDuration']=el
global_total['CompletedCounts']['Sum']={'Glideins':0,
'Lasted':0,
'FailedNr':0,
'JobsNr':0,
'JobsLasted':0,
'JobsGoodput':0,
'JobsTerminated':0,
'CondorLasted':0}
fe_total=copy.deepcopy(global_total) # same as above but for frontend totals
#
status={'entries':{},'total':global_total}
status_fe={'frontends':{}} #analogous to above but for frontend totals
nr_entries=0
nr_feentries={} #dictionary for nr entries per fe
for entry in monitorAggregatorConfig.entries:
# load entry log summary file
status_fname=os.path.join(os.path.join(monitorAggregatorConfig.monitor_dir,'entry_'+entry),
monitorAggregatorConfig.logsummary_relname)
try:
entry_data=xmlParse.xmlfile2dict(status_fname,always_singular_list=['Fraction','TimeRange','Range'])
except IOError:
continue # file not found, ignore
# update entry
out_data={}
for frontend in entry_data['frontends'].keys():
fe_el=entry_data['frontends'][frontend]
out_fe_el={}
for k in ['Current','Entered','Exited']:
out_fe_el[k]={}
for s in fe_el[k].keys():
out_fe_el[k][s]=int(fe_el[k][s])
out_fe_el['CompletedCounts']={'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{},'Sum':{}}
for tkey in fe_el['CompletedCounts']['Sum'].keys():
out_fe_el['CompletedCounts']['Sum'][tkey]=int(fe_el['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
out_fe_el['CompletedCounts'][w][k]={}
for t in glideFactoryMonitoring.getAllMillRanges():
out_fe_el['CompletedCounts'][w][k][t]=int(fe_el['CompletedCounts'][w][k][t]['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['Lasted'][t]=int(fe_el['CompletedCounts']['Lasted'][t]['val'])
out_fe_el['CompletedCounts']['JobsDuration']={}
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['JobsDuration'][t]=int(fe_el['CompletedCounts']['JobsDuration'][t]['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
out_fe_el['CompletedCounts']['JobsNr'][t]=int(fe_el['CompletedCounts']['JobsNr'][t]['val'])
out_data[frontend]=out_fe_el
status['entries'][entry]={'frontends':out_data}
# update total
if entry_data.has_key('total'):
nr_entries+=1
local_total={}
for k in ['Current','Entered','Exited']:
local_total[k]={}
for s in global_total[k].keys():
local_total[k][s]=int(entry_data['total'][k][s])
global_total[k][s]+=int(entry_data['total'][k][s])
local_total['CompletedCounts']={'Sum':{},'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{}}
for tkey in entry_data['total']['CompletedCounts']['Sum'].keys():
local_total['CompletedCounts']['Sum'][tkey]=int(entry_data['total']['CompletedCounts']['Sum'][tkey])
global_total['CompletedCounts']['Sum'][tkey]+=int(entry_data['total']['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
local_total['CompletedCounts'][w][k]={}
for t in glideFactoryMonitoring.getAllMillRanges():
local_total['CompletedCounts'][w][k][t]=int(entry_data['total']['CompletedCounts'][w][k][t]['val'])
global_total['CompletedCounts'][w][k][t]+=int(entry_data['total']['CompletedCounts'][w][k][t]['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['Lasted'][t]=int(entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
global_total['CompletedCounts']['Lasted'][t]+=int(entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
local_total['CompletedCounts']['JobsDuration']={}
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['JobsDuration'][t]=int(entry_data['total']['CompletedCounts']['JobsDuration'][t]['val'])
global_total['CompletedCounts']['JobsDuration'][t]+=int(entry_data['total']['CompletedCounts']['JobsDuration'][t]['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
local_total['CompletedCounts']['JobsNr'][t]=int(entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
global_total['CompletedCounts']['JobsNr'][t]+=int(entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
status['entries'][entry]['total']=local_total
# update frontends
for fe in out_data:
#compare each to the list of fe's accumulated so far
if not (fe in status_fe['frontends']):
status_fe['frontends'][fe]={}
if not (fe in nr_feentries):
nr_feentries[fe]=1 #already found one
else:
nr_feentries[fe]+=1
# sum them up
sumDictInt(out_data[fe],status_fe['frontends'][fe])
# Write xml files
# To do - Igor: Consider adding status_fe to the XML file
updated=time.time()
xml_str=('<?xml version="1.0" encoding="ISO-8859-1"?>\n\n'+
'<glideFactoryLogSummary>\n'+
xmlFormat.time2xml(updated,"updated", indent_tab=xmlFormat.DEFAULT_TAB,leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
xmlFormat.dict2string(status["entries"],dict_name="entries",el_name="entry",
subtypes_params={"class":{"dicts_params":{"frontends":{"el_name":"frontend",
"subtypes_params":{"class":{'subclass_params':{'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()}}}}},
"subclass_params":{"total":{"subclass_params":{'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()}}}
}
},
leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
xmlFormat.class2string(status["total"],inst_name="total",subclass_params={'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()},leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
"</glideFactoryLogSummary>\n")
glideFactoryMonitoring.monitoringConfig.write_file(monitorAggregatorConfig.logsummary_relname,xml_str)
# Write rrds
writeLogSummaryRRDs("total",status["total"])
# Frontend total rrds across all factories
for fe in status_fe['frontends']:
writeLogSummaryRRDs("total/%s"%("frontend_"+fe),status_fe['frontends'][fe])
return status
def verifyRRD(fix_rrd=False):
"""
Go through all known monitoring rrds and verify that they
match existing schema (could be different if an upgrade happened)
If fix_rrd is true, then also attempt to add any missing attributes.
"""
global rrd_problems_found
global monitorAggregatorConfig
dir=monitorAggregatorConfig.monitor_dir
total_dir=os.path.join(dir,"total")
status_dict={}
completed_stats_dict={}
completed_waste_dict={}
counts_dict={}
# initialize the RRD dictionaries to match the current schema for verification
for tp in status_attributes.keys():
if tp in type_strings.keys():
tp_str=type_strings[tp]
attributes_tp=status_attributes[tp]
for a in attributes_tp:
status_dict["%s%s"%(tp_str,a)]=None
for jobrange in glideFactoryMonitoring.getAllJobRanges():
completed_stats_dict["JobsNr_%s"%(jobrange)]=None
for timerange in glideFactoryMonitoring.getAllTimeRanges():
completed_stats_dict["Lasted_%s"%(timerange)]=None
completed_stats_dict["JobsLasted_%s"%(timerange)]=None
for jobtype in glideFactoryMonitoring.getAllJobTypes():
for timerange in glideFactoryMonitoring.getAllMillRanges():
completed_waste_dict["%s_%s"%(jobtype,timerange)]=None
for jobtype in ('Entered','Exited','Status'):
for jobstatus in ('Wait','Idle','Running','Held'):
counts_dict["%s%s"%(jobtype,jobstatus)]=None
for jobstatus in ('Completed','Removed'):
counts_dict["%s%s"%('Entered',jobstatus)]=None
verifyHelper(os.path.join(total_dir,
"Status_Attributes.rrd"),status_dict, fix_rrd)
verifyHelper(os.path.join(total_dir,
"Log_Completed.rrd"),
glideFactoryMonitoring.getLogCompletedDefaults(), fix_rrd)
verifyHelper(os.path.join(total_dir,
"Log_Completed_Stats.rrd"),completed_stats_dict, fix_rrd)
verifyHelper(os.path.join(total_dir,
"Log_Completed_WasteTime.rrd"),completed_waste_dict, fix_rrd)
verifyHelper(os.path.join(total_dir,
"Log_Counts.rrd"),counts_dict, fix_rrd)
for filename in os.listdir(dir):
if filename[:6]=="entry_":
entrydir=os.path.join(dir,filename)
for subfilename in os.listdir(entrydir):
if subfilename[:9]=="frontend_":
current_dir=os.path.join(entrydir,subfilename)
verifyHelper(os.path.join(current_dir,
"Status_Attributes.rrd"),status_dict, fix_rrd)
verifyHelper(os.path.join(current_dir,
"Log_Completed.rrd"),
glideFactoryMonitoring.getLogCompletedDefaults(),fix_rrd)
verifyHelper(os.path.join(current_dir,
"Log_Completed_Stats.rrd"),completed_stats_dict,fix_rrd)
verifyHelper(os.path.join(current_dir,
"Log_Completed_WasteTime.rrd"),
completed_waste_dict,fix_rrd)
verifyHelper(os.path.join(current_dir,
"Log_Counts.rrd"),counts_dict,fix_rrd)
return not rrd_problems_found
def verifyRRD(fix_rrd=False):
"""
Go through all known monitoring rrds and verify that they
match existing schema (could be different if an upgrade happened)
If fix_rrd is true, then also attempt to add any missing attributes.
"""
global rrd_problems_found
global monitorAggregatorConfig
mon_dir = monitorAggregatorConfig.monitor_dir
status_dict = {}
completed_stats_dict = {}
completed_waste_dict = {}
counts_dict = {}
# initialize the RRD dictionaries to match the current schema for verification
for tp in status_attributes.keys():
if tp in type_strings.keys():
tp_str = type_strings[tp]
attributes_tp = status_attributes[tp]
for a in attributes_tp:
status_dict["%s%s" % (tp_str, a)] = None
for jobrange in glideFactoryMonitoring.getAllJobRanges():
completed_stats_dict["JobsNr_%s" % (jobrange, )] = None
for timerange in glideFactoryMonitoring.getAllTimeRanges():
completed_stats_dict["Lasted_%s" % (timerange, )] = None
completed_stats_dict["JobsLasted_%s" % (timerange, )] = None
for jobtype in glideFactoryMonitoring.getAllJobTypes():
for timerange in glideFactoryMonitoring.getAllMillRanges():
completed_waste_dict["%s_%s" % (jobtype, timerange)] = None
for jobtype in ('Entered', 'Exited', 'Status'):
for jobstatus in ('Wait', 'Idle', 'Running', 'Held'):
counts_dict["%s%s" % (jobtype, jobstatus)] = None
for jobstatus in ('Completed', 'Removed'):
counts_dict["%s%s" % ('Entered', jobstatus)] = None
# FROM: lib2to3.fixes.fix_ws_comma
# completed_waste_dict["%s_%s"%(jobtype, timerange)]=None
#
# for jobtype in ('Entered', 'Exited', 'Status'):
# for jobstatus in ('Wait', 'Idle', 'Running', 'Held'):
# counts_dict["%s%s"%(jobtype, jobstatus)]=None
# for jobstatus in ('Completed', 'Removed'):
# counts_dict["%s%s"%('Entered', jobstatus)]=None
#
# verifyHelper(os.path.join(total_dir,
# "Status_Attributes.rrd"), status_dict, fix_rrd)
# verifyHelper(os.path.join(total_dir,
# "Log_Completed.rrd"),
# glideFactoryMonitoring.getLogCompletedDefaults(), fix_rrd)
# verifyHelper(os.path.join(total_dir,
# "Log_Completed_Stats.rrd"), completed_stats_dict, fix_rrd)
# verifyHelper(os.path.join(total_dir,
# "Log_Completed_WasteTime.rrd"), completed_waste_dict, fix_rrd)
# verifyHelper(os.path.join(total_dir,
# "Log_Counts.rrd"), counts_dict, fix_rrd)
# for filename in os.listdir(dir):
# if filename[:6]=="entry_":
# entrydir=os.path.join(dir, filename)
# for subfilename in os.listdir(entrydir):
# if subfilename[:9]=="frontend_":
# current_dir=os.path.join(entrydir, subfilename)
# verifyHelper(os.path.join(current_dir,
# "Status_Attributes.rrd"), status_dict, fix_rrd)
# verifyHelper(os.path.join(current_dir,
# "Log_Completed.rrd"),
# glideFactoryMonitoring.getLogCompletedDefaults(), fix_rrd)
# verifyHelper(os.path.join(current_dir,
# "Log_Completed_Stats.rrd"), completed_stats_dict, fix_rrd)
# verifyHelper(os.path.join(current_dir,
# "Log_Completed_WasteTime.rrd"),
# completed_waste_dict, fix_rrd)
# verifyHelper(os.path.join(current_dir,
# "Log_Counts.rrd"), counts_dict, fix_rrd)
# return not rrd_problems_found
completed_dict = glideFactoryMonitoring.getLogCompletedDefaults()
rrdict = {
"Status_Attributes.rrd": status_dict,
"Log_Completed.rrd": completed_dict,
"Log_Completed_Stats.rrd": completed_stats_dict,
"Log_Completed_WasteTime.rrd": completed_waste_dict,
"Log_Counts.rrd": counts_dict,
}
for dir_name, sdir_name, f_list in os.walk(mon_dir):
for file_name in f_list:
if file_name in rrdict.keys():
verifyHelper(os.path.join(dir_name, file_name),
rrdict[file_name], fix_rrd)
return not rrd_problems_found
def aggregateLogSummary():
"""
Create an aggregate of log summary files, write it in an aggregate log
summary file and in the end return the values
"""
global monitorAggregatorConfig
# initialize global counters
global_total = {
'Current': {},
'Entered': {},
'Exited': {},
'CompletedCounts': {
'Sum': {},
'Waste': {},
'WasteTime': {},
'Lasted': {},
'JobsNr': {},
'JobsDuration': {}
}
}
for s in ('Wait', 'Idle', 'Running', 'Held'):
for k in ['Current', 'Entered', 'Exited']:
global_total[k][s] = 0
for s in ('Completed', 'Removed'):
for k in ['Entered']:
global_total[k][s] = 0
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ('Waste', 'WasteTime'):
el = {}
for t in glideFactoryMonitoring.getAllMillRanges():
el[t] = 0
global_total['CompletedCounts'][w][k] = el
el = {}
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t] = 0
global_total['CompletedCounts']['Lasted'] = el
el = {}
for t in glideFactoryMonitoring.getAllJobRanges():
el[t] = 0
global_total['CompletedCounts']['JobsNr'] = el
el = {}
# KEL - why is the same el used twice (see above)
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t] = 0
global_total['CompletedCounts']['JobsDuration'] = el
global_total['CompletedCounts']['Sum'] = {
'Glideins': 0,
'Lasted': 0,
'FailedNr': 0,
'JobsNr': 0,
'JobsLasted': 0,
'JobsGoodput': 0,
'JobsTerminated': 0,
'CondorLasted': 0
}
fe_total = copy.deepcopy(
global_total) # same as above but for frontend totals
#
status = {'entries': {}, 'total': global_total}
status_fe = {'frontends': {}} # analogous to above but for frontend totals
nr_entries = 0
nr_feentries = {} # dictionary for nr entries per fe
for entry in monitorAggregatorConfig.entries:
# load entry log summary file
status_fname = os.path.join(
os.path.join(monitorAggregatorConfig.monitor_dir,
'entry_' + entry),
monitorAggregatorConfig.logsummary_relname)
try:
entry_data = xmlParse.xmlfile2dict(
status_fname,
always_singular_list=['Fraction', 'TimeRange', 'Range'])
except IOError:
continue # file not found, ignore
# update entry
out_data = {}
for frontend in entry_data['frontends'].keys():
fe_el = entry_data['frontends'][frontend]
out_fe_el = {}
for k in ['Current', 'Entered', 'Exited']:
out_fe_el[k] = {}
for s in fe_el[k].keys():
out_fe_el[k][s] = int(fe_el[k][s])
out_fe_el['CompletedCounts'] = {
'Waste': {},
'WasteTime': {},
'Lasted': {},
'JobsNr': {},
'JobsDuration': {},
'Sum': {}
}
for tkey in fe_el['CompletedCounts']['Sum'].keys():
out_fe_el['CompletedCounts']['Sum'][tkey] = int(
fe_el['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste", "WasteTime"):
out_fe_el['CompletedCounts'][w][k] = {}
for t in glideFactoryMonitoring.getAllMillRanges():
out_fe_el['CompletedCounts'][w][k][t] = int(
fe_el['CompletedCounts'][w][k][t]['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['Lasted'][t] = int(
fe_el['CompletedCounts']['Lasted'][t]['val'])
out_fe_el['CompletedCounts']['JobsDuration'] = {}
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['JobsDuration'][t] = int(
fe_el['CompletedCounts']['JobsDuration'][t]['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
out_fe_el['CompletedCounts']['JobsNr'][t] = int(
fe_el['CompletedCounts']['JobsNr'][t]['val'])
out_data[frontend] = out_fe_el
status['entries'][entry] = {'frontends': out_data}
# update total
if 'total' in entry_data:
nr_entries += 1
local_total = {}
for k in ['Current', 'Entered', 'Exited']:
local_total[k] = {}
for s in global_total[k].keys():
local_total[k][s] = int(entry_data['total'][k][s])
global_total[k][s] += int(entry_data['total'][k][s])
local_total['CompletedCounts'] = {
'Sum': {},
'Waste': {},
'WasteTime': {},
'Lasted': {},
'JobsNr': {},
'JobsDuration': {}
}
for tkey in entry_data['total']['CompletedCounts']['Sum'].keys():
local_total['CompletedCounts']['Sum'][tkey] = int(
entry_data['total']['CompletedCounts']['Sum'][tkey])
global_total['CompletedCounts']['Sum'][tkey] += int(
entry_data['total']['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ('Waste', 'WasteTime'):
local_total['CompletedCounts'][w][k] = {}
for t in glideFactoryMonitoring.getAllMillRanges():
local_total['CompletedCounts'][w][k][t] = int(
entry_data['total']['CompletedCounts'][w][k][t]
['val'])
global_total['CompletedCounts'][w][k][t] += int(
entry_data['total']['CompletedCounts'][w][k][t]
['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['Lasted'][t] = int(
entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
global_total['CompletedCounts']['Lasted'][t] += int(
entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
local_total['CompletedCounts']['JobsDuration'] = {}
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['JobsDuration'][t] = int(
entry_data['total']['CompletedCounts']['JobsDuration'][t]
['val'])
global_total['CompletedCounts']['JobsDuration'][t] += int(
entry_data['total']['CompletedCounts']['JobsDuration'][t]
['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
local_total['CompletedCounts']['JobsNr'][t] = int(
entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
global_total['CompletedCounts']['JobsNr'][t] += int(
entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
status['entries'][entry]['total'] = local_total
# update frontends
for fe in out_data:
# compare each to the list of fe's accumulated so far
if not (fe in status_fe['frontends']):
status_fe['frontends'][fe] = {}
if not (fe in nr_feentries):
nr_feentries[fe] = 1 # already found one
else:
nr_feentries[fe] += 1
# sum them up
sumDictInt(out_data[fe], status_fe['frontends'][fe])
# Write xml files
# To do - Igor: Consider adding status_fe to the XML file
updated = time.time()
xml_str = ('<?xml version="1.0" encoding="ISO-8859-1"?>\n\n' +
'<glideFactoryLogSummary>\n' +
xmlFormat.time2xml(updated,
"updated",
indent_tab=xmlFormat.DEFAULT_TAB,
leading_tab=xmlFormat.DEFAULT_TAB) + '\n' +
xmlFormat.dict2string(
status["entries"],
dict_name="entries",
el_name="entry",
subtypes_params={
"class": {
"dicts_params": {
"frontends": {
"el_name": "frontend",
"subtypes_params": {
"class": {
'subclass_params': {
'CompletedCounts':
glideFactoryMonitoring.
get_completed_stats_xml_desc()
}
}
}
}
},
"subclass_params": {
"total": {
"subclass_params": {
'CompletedCounts':
glideFactoryMonitoring.
get_completed_stats_xml_desc()
}
}
}
}
},
leading_tab=xmlFormat.DEFAULT_TAB) + '\n' +
xmlFormat.class2string(
status["total"],
inst_name="total",
subclass_params={
'CompletedCounts':
glideFactoryMonitoring.get_completed_stats_xml_desc()
},
leading_tab=xmlFormat.DEFAULT_TAB) + '\n' +
"</glideFactoryLogSummary>\n")
glideFactoryMonitoring.monitoringConfig.write_file(
monitorAggregatorConfig.logsummary_relname, xml_str)
# Write rrds
writeLogSummaryRRDs("total", status["total"])
# Frontend total rrds across all factories
for fe in status_fe['frontends']:
writeLogSummaryRRDs("total/%s" % ("frontend_" + fe),
status_fe['frontends'][fe])
return status
def aggregateLogSummary():
"""
Create an aggregate of log summary files, write it in an aggregate log
summary file and in the end return the values
"""
global monitorAggregatorConfig
# initialize global counters
global_total={'Current':{},'Entered':{},'Exited':{},'CompletedCounts':{'Sum':{},'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{}}}
for s in ('Wait','Idle','Running','Held'):
for k in ['Current','Entered','Exited']:
global_total[k][s]=0
for s in ('Completed','Removed'):
for k in ['Entered']:
global_total[k][s]=0
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
el={}
for t in glideFactoryMonitoring.getAllMillRanges():
el[t]=0
global_total['CompletedCounts'][w][k]=el
el={}
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t]=0
global_total['CompletedCounts']['Lasted']=el
el={}
for t in glideFactoryMonitoring.getAllJobRanges():
el[t]=0
global_total['CompletedCounts']['JobsNr']=el
el={}
# KEL - why is the same el used twice (see above)
for t in glideFactoryMonitoring.getAllTimeRanges():
el[t]=0
global_total['CompletedCounts']['JobsDuration']=el
global_total['CompletedCounts']['Sum']={'Glideins':0,
'Lasted':0,
'FailedNr':0,
'JobsNr':0,
'JobsLasted':0,
'JobsGoodput':0,
'JobsTerminated':0,
'CondorLasted':0}
#
status={'entries':{},'total':global_total}
nr_entries=0
for entry in monitorAggregatorConfig.entries:
# load entry log summary file
status_fname=os.path.join(os.path.join(monitorAggregatorConfig.monitor_dir,'entry_'+entry),
monitorAggregatorConfig.logsummary_relname)
try:
entry_data=xmlParse.xmlfile2dict(status_fname,always_singular_list=['Fraction','TimeRange','Range'])
except IOError:
continue # file not found, ignore
# update entry
out_data={}
for frontend in entry_data['frontends'].keys():
fe_el=entry_data['frontends'][frontend]
out_fe_el={}
for k in ['Current','Entered','Exited']:
out_fe_el[k]={}
for s in fe_el[k].keys():
out_fe_el[k][s]=int(fe_el[k][s])
out_fe_el['CompletedCounts']={'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{},'Sum':{}}
for tkey in fe_el['CompletedCounts']['Sum'].keys():
out_fe_el['CompletedCounts']['Sum'][tkey]=int(fe_el['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
out_fe_el['CompletedCounts'][w][k]={}
for t in glideFactoryMonitoring.getAllMillRanges():
out_fe_el['CompletedCounts'][w][k][t]=int(fe_el['CompletedCounts'][w][k][t]['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['Lasted'][t]=int(fe_el['CompletedCounts']['Lasted'][t]['val'])
out_fe_el['CompletedCounts']['JobsDuration']={}
for t in glideFactoryMonitoring.getAllTimeRanges():
out_fe_el['CompletedCounts']['JobsDuration'][t]=int(fe_el['CompletedCounts']['JobsDuration'][t]['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
out_fe_el['CompletedCounts']['JobsNr'][t]=int(fe_el['CompletedCounts']['JobsNr'][t]['val'])
out_data[frontend]=out_fe_el
status['entries'][entry]={'frontends':out_data}
# update total
if entry_data.has_key('total'):
nr_entries+=1
local_total={}
for k in ['Current','Entered','Exited']:
local_total[k]={}
for s in global_total[k].keys():
local_total[k][s]=int(entry_data['total'][k][s])
global_total[k][s]+=int(entry_data['total'][k][s])
local_total['CompletedCounts']={'Sum':{},'Waste':{},'WasteTime':{},'Lasted':{},'JobsNr':{},'JobsDuration':{}}
for tkey in entry_data['total']['CompletedCounts']['Sum'].keys():
local_total['CompletedCounts']['Sum'][tkey]=int(entry_data['total']['CompletedCounts']['Sum'][tkey])
global_total['CompletedCounts']['Sum'][tkey]+=int(entry_data['total']['CompletedCounts']['Sum'][tkey])
for k in glideFactoryMonitoring.getAllJobTypes():
for w in ("Waste","WasteTime"):
local_total['CompletedCounts'][w][k]={}
for t in glideFactoryMonitoring.getAllMillRanges():
local_total['CompletedCounts'][w][k][t]=int(entry_data['total']['CompletedCounts'][w][k][t]['val'])
global_total['CompletedCounts'][w][k][t]+=int(entry_data['total']['CompletedCounts'][w][k][t]['val'])
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['Lasted'][t]=int(entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
global_total['CompletedCounts']['Lasted'][t]+=int(entry_data['total']['CompletedCounts']['Lasted'][t]['val'])
local_total['CompletedCounts']['JobsDuration']={}
for t in glideFactoryMonitoring.getAllTimeRanges():
local_total['CompletedCounts']['JobsDuration'][t]=int(entry_data['total']['CompletedCounts']['JobsDuration'][t]['val'])
global_total['CompletedCounts']['JobsDuration'][t]+=int(entry_data['total']['CompletedCounts']['JobsDuration'][t]['val'])
for t in glideFactoryMonitoring.getAllJobRanges():
local_total['CompletedCounts']['JobsNr'][t]=int(entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
global_total['CompletedCounts']['JobsNr'][t]+=int(entry_data['total']['CompletedCounts']['JobsNr'][t]['val'])
status['entries'][entry]['total']=local_total
# Write xml files
updated=time.time()
xml_str=('<?xml version="1.0" encoding="ISO-8859-1"?>\n\n'+
'<glideFactoryLogSummary>\n'+
xmlFormat.time2xml(updated,"updated", indent_tab=xmlFormat.DEFAULT_TAB,leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
xmlFormat.dict2string(status["entries"],dict_name="entries",el_name="entry",
subtypes_params={"class":{"dicts_params":{"frontends":{"el_name":"frontend",
"subtypes_params":{"class":{'subclass_params':{'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()}}}}},
"subclass_params":{"total":{"subclass_params":{'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()}}}
}
},
leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
xmlFormat.class2string(status["total"],inst_name="total",subclass_params={'CompletedCounts':glideFactoryMonitoring.get_completed_stats_xml_desc()},leading_tab=xmlFormat.DEFAULT_TAB)+"\n"+
"</glideFactoryLogSummary>\n")
glideFactoryMonitoring.monitoringConfig.write_file(monitorAggregatorConfig.logsummary_relname,xml_str)
# Write rrds
fe_dir="total"
sdata=status["total"]['Current']
glideFactoryMonitoring.monitoringConfig.establish_dir(fe_dir)
val_dict_counts={}
val_dict_counts_desc={}
val_dict_completed={}
val_dict_stats={}
val_dict_waste={}
val_dict_wastetime={}
for s in ('Wait','Idle','Running','Held','Completed','Removed'):
if not (s in ('Completed','Removed')): # I don't have their numbers from inactive logs
count=sdata[s]
val_dict_counts["Status%s"%s]=count
val_dict_counts_desc["Status%s"%s]={'ds_type':'GAUGE'}
exited=-status["total"]['Exited'][s]
val_dict_counts["Exited%s"%s]=exited
val_dict_counts_desc["Exited%s"%s]={'ds_type':'ABSOLUTE'}
entered=status["total"]['Entered'][s]
val_dict_counts["Entered%s"%s]=entered
val_dict_counts_desc["Entered%s"%s]={'ds_type':'ABSOLUTE'}
if s=='Completed':
completed_counts=status["total"]['CompletedCounts']
count_entered_times=completed_counts['Lasted']
count_jobnrs=completed_counts['JobsNr']
count_jobs_duration=completed_counts['JobsDuration']
count_waste_mill=completed_counts['Waste']
time_waste_mill=completed_counts['WasteTime']
# save run times
for timerange in count_entered_times.keys():
val_dict_stats['Lasted_%s'%timerange]=count_entered_times[timerange]
# they all use the same indexes
val_dict_stats['JobsLasted_%s'%timerange]=count_jobs_duration[timerange]
# save jobsnr
for jobrange in count_jobnrs.keys():
val_dict_stats['JobsNr_%s'%jobrange]=count_jobnrs[jobrange]
# save simple vals
for tkey in completed_counts['Sum'].keys():
val_dict_completed[tkey]=completed_counts['Sum'][tkey]
# save waste_mill
for w in count_waste_mill.keys():
count_waste_mill_w=count_waste_mill[w]
for p in count_waste_mill_w.keys():
val_dict_waste['%s_%s'%(w,p)]=count_waste_mill_w[p]
for w in time_waste_mill.keys():
time_waste_mill_w=time_waste_mill[w]
for p in time_waste_mill_w.keys():
val_dict_wastetime['%s_%s'%(w,p)]=time_waste_mill_w[p]
# write the data to disk
glideFactoryMonitoring.monitoringConfig.write_rrd_multi_hetero("%s/Log_Counts"%fe_dir,
val_dict_counts_desc,updated,val_dict_counts)
glideFactoryMonitoring.monitoringConfig.write_rrd_multi("%s/Log_Completed"%fe_dir,
"ABSOLUTE",updated,val_dict_completed)
glideFactoryMonitoring.monitoringConfig.write_rrd_multi("%s/Log_Completed_Stats"%fe_dir,
"ABSOLUTE",updated,val_dict_stats)
# Disable Waste RRDs... WasteTime much more useful
#glideFactoryMonitoring.monitoringConfig.write_rrd_multi("%s/Log_Completed_Waste"%fe_dir,
# "ABSOLUTE",updated,val_dict_waste)
glideFactoryMonitoring.monitoringConfig.write_rrd_multi("%s/Log_Completed_WasteTime"%fe_dir,
"ABSOLUTE",updated,val_dict_wastetime)
return status