def main(argv):
pywrapfile.File.Init()
config = entconfig.EntConfig(argv[0])
if not config.Load(): sys.exit(__doc__)
# Collect logs only if active
state = install_utilities.install_state(config.var('VERSION'))
if not state in [ 'ACTIVE', 'SERVE' ]:
sys.exit(0)
# NO collection for sitesearches:
if config.var('SITESEARCH_INTERFACE'):
sys.exit(0)
# If I'm not a config replica I don't collect the logs..
replicas = config.var('CONFIG_REPLICAS')
crt_machine = E.getCrtHostName()
if not crt_machine in replicas:
logging.error('Not a replica')
sys.exit(0)
gws_log_dir = liblog.get_gws_log_dir(config)
collect_dir = liblog.get_collect_dir(config)
partition_dir = liblog.get_partition_dir(config)
apache_dir = liblog.get_apache_dir(config)
click_dir = liblog.get_click_dir(config)
directory_map_file = liblog.get_directory_map_file(config)
# in case cron job starts before adminrunner
liblog.MakeDir(collect_dir)
liblog.MakeDir(partition_dir)
liblog.MakeDir(apache_dir)
liblog.MakeDir(click_dir)
# Collect Logs from machines
all_machines = config.var('MACHINES')
CollectLogs(all_machines, gws_log_dir, collect_dir)
# Partition gwslogs by collections and convert to apache logs
preprocess_logs.PartitionLogs(config)
# Sanitize collection directory map
coll_directory_map = liblog.CollectionDirectoryMap(directory_map_file)
if coll_directory_map.sanitizeMap(partition_dir, apache_dir, click_dir):
coll_directory_map.saveToDisk()
# Send the OP logs to all machines
SyncOpLogs(all_machines, config.var('LOGDIR'))
logging.info('Done')
def __init__(self, cfg):
self.cfg = cfg # configurator object
self.entConfig = cfg.globalParams
# locks for updating the report lists
self.logreplock = threading.RLock()
self.logdir = self.cfg.getGlobalParam('LOGDIR')
liblog.MakeDir(liblog.get_click_dir(self.entConfig))
liblog.MakeDir(liblog.get_collect_dir(self.entConfig))
liblog.MakeDir(liblog.get_apache_dir(self.entConfig))
liblog.MakeDir(liblog.get_partition_dir(self.entConfig))
liblog.MakeGoogleDir(self.entConfig, liblog.get_report_dir(self.entConfig))
collection_dir_map_file = liblog.get_directory_map_file(self.entConfig)
if not os.path.exists(collection_dir_map_file):
open(collection_dir_map_file, 'w').close() # a trick to touch a file.
self.reportCount = { liblog.RAW_REPORT: 0,
liblog.SUMMARY_REPORT: 0, }
self.sanitizeReportList(liblog.RAW_REPORT)
self.sanitizeReportList(liblog.SUMMARY_REPORT)
self.joblock = threading.Lock()
self.runningJobs = {}
def main(argv):
argc = len(argv)
if argc < 10:
sys.exit(__doc__)
config = entconfig.EntConfig(argv[0])
if not config.Load(): sys.exit(__doc__)
pywrapbase.InitGoogleScript('', [
'foo',
'--gfs_aliases=%s' % config.var("GFS_ALIASES"),
'--bnsresolver_use_svelte=false', '--logtostderr'
], 0)
gfile.Init()
client = argv[1]
date_fields = string.split(argv[2], '_')
date_range = liblog.ParseDateRange(date_fields[0], date_fields[1:])
withResults = argv[3]
topCount = argv[4]
diagnosticTerms = argv[5]
html_file = argv[6]
valid_file = argv[7]
new_html_file = argv[8]
new_valid_file = argv[9]
if not date_range:
sys.exit(__doc__)
first_date, last_date, printable_date, file_date = date_range
if last_date.as_int() < first_date.as_int():
logging.fatal('invalid date range')
gws_log_dir = liblog.get_gws_log_dir(config)
collect_dir = liblog.get_collect_dir(config)
partition_dir = liblog.get_partition_dir(config)
directory_map_file = liblog.get_directory_map_file(config)
# Collect logs first from all gws nodes and preprocess
# logs to make sure logs are up to date.
all_machines = config.var('MACHINES')
collect_logs.CollectLogs(all_machines, gws_log_dir, collect_dir)
preprocess_logs.PartitionLogs(config)
gws_logs = liblog.FindClientLogFiles(partition_dir, directory_map_file,
client, first_date, last_date)
# note that collection (client) has been factored into gwslog_dir.
result = CreateLogReport(config, printable_date, gws_logs,
config.var('MAIN_GOOGLE3_DIR'), withResults,
topCount, diagnosticTerms, html_file, valid_file,
new_html_file, new_valid_file)
if result == liblog.FAILURE:
logging.error('CreateLogReport Failed')
sys.exit(result)
def main(argv):
argc = len(argv)
if argc < 10:
sys.exit(__doc__)
config = entconfig.EntConfig(argv[0])
if not config.Load(): sys.exit(__doc__)
pywrapbase.InitGoogleScript('', ['foo',
'--gfs_aliases=%s' % config.var("GFS_ALIASES"),
'--bnsresolver_use_svelte=false',
'--logtostderr'], 0)
gfile.Init()
client = argv[1]
date_fields = string.split(argv[2], '_')
date_range = liblog.ParseDateRange(date_fields[0], date_fields[1:])
withResults = argv[3]
topCount = argv[4]
diagnosticTerms = argv[5]
html_file = argv[6]
valid_file = argv[7]
new_html_file = argv[8]
new_valid_file = argv[9]
if not date_range:
sys.exit(__doc__)
first_date, last_date, printable_date, file_date = date_range
if last_date.as_int() < first_date.as_int():
logging.fatal('invalid date range')
gws_log_dir = liblog.get_gws_log_dir(config)
collect_dir = liblog.get_collect_dir(config)
partition_dir = liblog.get_partition_dir(config)
directory_map_file = liblog.get_directory_map_file(config)
# Collect logs first from all gws nodes and preprocess
# logs to make sure logs are up to date.
all_machines = config.var('MACHINES')
collect_logs.CollectLogs(all_machines, gws_log_dir, collect_dir)
preprocess_logs.PartitionLogs(config)
gws_logs = liblog.FindClientLogFiles(partition_dir, directory_map_file,
client, first_date, last_date)
# note that collection (client) has been factored into gwslog_dir.
result = CreateLogReport(config, printable_date, gws_logs,
config.var('MAIN_GOOGLE3_DIR'),
withResults, topCount, diagnosticTerms,
html_file, valid_file,
new_html_file, new_valid_file)
if result == liblog.FAILURE:
logging.error('CreateLogReport Failed')
sys.exit(result)
def main(argv):
argc = len(argv)
if argc < 6:
sys.exit(__doc__)
config = entconfig.EntConfig(argv[0])
if not config.Load():
sys.exit(__doc__)
pywrapbase.InitGoogleScript('', ['foo',
'--gfs_aliases=%s' % config.var("GFS_ALIASES"),
'--bnsresolver_use_svelte=false',
'--logtostderr'], 0)
gfile.Init()
client = argv[1]
date_arg = argv[2]
html_file = argv[3]
valid_file = argv[4]
new_valid_file = argv[5]
# extract tag and date_range from command line args
date_fields = string.split(date_arg, '_')
date_range = liblog.ParseDateRange(date_fields[0], date_fields[1:])
if not date_range:
sys.exit(__doc__)
first_date, last_date, printable_date, file_date = date_range
if last_date.as_int() < first_date.as_int():
sys.exit(__doc__)
gws_log_dir = liblog.get_gws_log_dir(config)
click_dir = liblog.get_click_dir(config)
collect_dir = liblog.get_collect_dir(config)
apache_dir = liblog.get_apache_dir(config)
directory_map_file = liblog.get_directory_map_file(config)
# we need to collect logs first from all gws nodes and preprocess
# logs first to make sure logs are up to date.
all_machines = config.var('MACHINES')
collect_logs.CollectLogs(all_machines, gws_log_dir, collect_dir)
preprocess_logs.PartitionLogs(config)
# make a vector of Log objects for all apache_logs and click_logs matching
# the given date range and client.
apache_logs = liblog.FindClientLogFiles(apache_dir, directory_map_file,
client, first_date, last_date)
click_logs = liblog.FindClientLogFiles(click_dir, directory_map_file,
client, first_date, last_date)
# If we have valid file and report file, we check to see if the data in
# apache_dir has been changed and if the report is still valid.
if (gfile.Exists(html_file) and gfile.Exists(valid_file) and
liblog.checkValid(html_file, valid_file, apache_logs)):
logging.info('%s still valid.' % html_file)
sys.exit(liblog.STILL_VALID)
# if there is no valid report, we create a new one
DumpApacheAndClickLogs(apache_logs, click_logs)
if not liblog.makeValid(new_valid_file, apache_logs):
logging.error('Error validating %s' % html_file)
sys.exit(liblog.FAILURE)
logging.info('done apache_log, new_valid_file: %s' % new_valid_file)
sys.exit(liblog.SUCCESS)
def main(argv):
argc = len(argv)
if argc < 6:
sys.exit(__doc__)
config = entconfig.EntConfig(argv[0])
if not config.Load():
sys.exit(__doc__)
pywrapbase.InitGoogleScript('', [
'foo',
'--gfs_aliases=%s' % config.var("GFS_ALIASES"),
'--bnsresolver_use_svelte=false', '--logtostderr'
], 0)
gfile.Init()
client = argv[1]
date_arg = argv[2]
html_file = argv[3]
valid_file = argv[4]
new_valid_file = argv[5]
# extract tag and date_range from command line args
date_fields = string.split(date_arg, '_')
date_range = liblog.ParseDateRange(date_fields[0], date_fields[1:])
if not date_range:
sys.exit(__doc__)
first_date, last_date, printable_date, file_date = date_range
if last_date.as_int() < first_date.as_int():
sys.exit(__doc__)
gws_log_dir = liblog.get_gws_log_dir(config)
click_dir = liblog.get_click_dir(config)
collect_dir = liblog.get_collect_dir(config)
apache_dir = liblog.get_apache_dir(config)
directory_map_file = liblog.get_directory_map_file(config)
# we need to collect logs first from all gws nodes and preprocess
# logs first to make sure logs are up to date.
all_machines = config.var('MACHINES')
collect_logs.CollectLogs(all_machines, gws_log_dir, collect_dir)
preprocess_logs.PartitionLogs(config)
# make a vector of Log objects for all apache_logs and click_logs matching
# the given date range and client.
apache_logs = liblog.FindClientLogFiles(apache_dir, directory_map_file,
client, first_date, last_date)
click_logs = liblog.FindClientLogFiles(click_dir, directory_map_file,
client, first_date, last_date)
# If we have valid file and report file, we check to see if the data in
# apache_dir has been changed and if the report is still valid.
if (gfile.Exists(html_file) and gfile.Exists(valid_file)
and liblog.checkValid(html_file, valid_file, apache_logs)):
logging.info('%s still valid.' % html_file)
sys.exit(liblog.STILL_VALID)
# if there is no valid report, we create a new one
DumpApacheAndClickLogs(apache_logs, click_logs)
if not liblog.makeValid(new_valid_file, apache_logs):
logging.error('Error validating %s' % html_file)
sys.exit(liblog.FAILURE)
logging.info('done apache_log, new_valid_file: %s' % new_valid_file)
sys.exit(liblog.SUCCESS)
