os-loganalyze/os_loganalyze/filter.py

216 lines
6.6 KiB
Python

#!/usr/bin/env python
#
# Copyright (c) 2013 IBM Corp.
# Copyright (c) 2014 Hewlett-Packard Development Company, L.P.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
# License for the specific language governing permissions and limitations
# under the License.
import re
import os_loganalyze.generator as generator
import os_loganalyze.util as util
# which logs support severity
SUPPORTS_SEV = re.compile(
r'/' # this uses an re.search so anchor the string
r'((screen-)?(n-|c-|g-|h-|ir-|ironic-|m-|o-|df-|placement-api|'
r'q-|neutron-|' # support both lib/neutron and lib/neutron-legacy logs
r'ceil|key|sah|des|tr)' # openstack logs
r'|(devstack\@)' # systemd logs
# other things we understand
r'|(keystone|tempest)\.txt|syslog)')
SYSLOGDATE = '\w+\s+\d+\s+\d{2}:\d{2}:\d{2}((\.|\,)\d{3,6})?'
DATEFMT = '\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}((\.|\,)\d{3,6})?'
STATUSFMT = '(DEBUG|INFO|WARNING|ERROR|TRACE|AUDIT|CRITICAL)'
OSLO_LOGMATCH = '^(?P<date>%s)(?P<line>(?P<pid> \d+)? (?P<status>%s).*)' % \
(DATEFMT, STATUSFMT)
SYSLOG_MATCH = ('^(?P<date>%s)(?P<line> (?P<host>[\w\-]+) '
'(?P<service>[^\[\s]+):.*)' %
(SYSLOGDATE))
SYSTEMD_MATCH = (
'^(?P<date>%s)(?P<line> (?P<host>\S+) \S+\[\d+\]\: (?P<status>%s)?.*)' %
(SYSLOGDATE, STATUSFMT))
CONSOLE_MATCH = '^(?P<date>%s)(?P<line>.*)' % DATEFMT
OSLORE = re.compile(OSLO_LOGMATCH)
SYSLOGRE = re.compile(SYSLOG_MATCH)
CONSOLERE = re.compile(CONSOLE_MATCH)
SYSTEMDRE = re.compile(SYSTEMD_MATCH)
SEVS = {
'NONE': 0,
'DEBUG': 1,
'INFO': 2,
'AUDIT': 3,
'TRACE': 4,
'WARNING': 5,
'ERROR': 6,
'CRITICAL': 7,
}
class LogLine(object):
status = "NONE"
line = ""
date = ""
pid = ""
service = ""
def __init__(self, line, old_sev="NONE"):
self._parse(line, old_sev)
def _syslog_status(self, service):
if service in ('tgtd', 'proxy-server'):
return 'DEBUG'
else:
return 'INFO'
def safe_date(self):
return '_' + re.sub('[\s\:\.\,]', '_', self.date)
def _parse(self, line, old_sev):
m = OSLORE.match(line)
if m:
self.status = m.group('status')
self.line = m.group('line')
self.date = m.group('date')
self.pid = m.group('pid')
return
m = SYSTEMDRE.match(line)
if m:
self.status = m.group('status') or "NONE"
self.line = m.group('line')
self.date = m.group('date')
self.host = m.group('host')
return
m = CONSOLERE.match(line)
if m:
self.date = m.group('date')
self.status = old_sev
self.line = m.group('line')
return
m = SYSLOGRE.match(line)
if m:
self.service = m.group('service')
self.line = m.group('line')
self.date = m.group('date')
self.status = self._syslog_status(self.service)
return
self.status = old_sev
self.line = line.rstrip()
class SevFilter(object):
def __init__(self, file_generator, minsev="NONE", limit=None):
self.minsev = minsev
self.file_generator = file_generator
self.supports_sev = \
SUPPORTS_SEV.search(file_generator.logname) is not None
self.limit = limit
self.strip_control = False
def strip(self, line):
return re.sub('\x1b\[(([03]\d)|\;)+m', '', line)
def __iter__(self):
old_sev = "NONE"
lineno = 1
for line in self.file_generator:
# bail early for limits
if self.limit and lineno > int(self.limit):
raise StopIteration()
# strip control chars in case the console is ascii colored
if self.strip_control:
line = self.strip(line)
logline = LogLine(line, old_sev)
# Some log lines come without severity. Treat those as
# belonging to the previous non NONE severity so that we
# get those log lines associated to the proper level.
if logline.status != "NONE":
old_sev = logline.status
else:
logline.status = old_sev
if self.supports_sev and self.skip_by_sev(logline.status):
continue
lineno += 1
yield logline
def skip_by_sev(self, sev):
"""should we skip this line?
If the line severity is less than our minimum severity,
yes we should.
"""
minsev = self.minsev
return SEVS.get(sev, 0) < SEVS.get(minsev, 0)
class Line(object):
date = ''
def __init__(self, line):
self.line = line
class NoFilter(object):
supports_sev = False
def __init__(self, file_generator):
self.file_generator = file_generator
def __iter__(self):
for line in self.file_generator:
yield Line(line)
def get_filter_generator(file_generator, environ, root_path, config):
"""Return the filter to use as per the config."""
# Check if the generator is an index page. If so, we don't want to apply
# any filters
if isinstance(file_generator, generator.IndexIterableBuffer):
return NoFilter(file_generator)
# Check file specific conditions first
filter_selected = util.get_file_conditions('filter', file_generator,
environ, root_path, config)
# Otherwise use the defaults in the config
if not filter_selected:
if config.has_section('general'):
if config.has_option('general', 'filter'):
filter_selected = config.get('general', 'filter')
minsev = util.parse_param(environ, 'level', default="NONE")
limit = util.parse_param(environ, 'limit')
if filter_selected:
if filter_selected.lower() in ['sevfilter', 'sev']:
return SevFilter(file_generator, minsev, limit)
elif filter_selected.lower() in ['nofilter', 'no']:
return NoFilter(file_generator)
# Otherwise guess
if util.use_passthrough_view(file_generator.file_headers):
return NoFilter(file_generator)
return SevFilter(file_generator, minsev, limit)