summaryrefslogtreecommitdiffstats
path: root/doctor_tests/main.py
blob: 8051ad6e756a6b088a218feef8cd311708105861 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
##############################################################################
# Copyright (c) 2017 ZTE Corporation and others.
#
# All rights reserved. This program and the accompanying materials
# are made available under the terms of the Apache License, Version 2.0
# which accompanies this distribution, and is available at
# http://www.apache.org/licenses/LICENSE-2.0
##############################################################################
import os
from os.path import isfile, join
import random
import signal
import sys
import time

from doctor_tests.alarm import Alarm
from doctor_tests.common.constants import Host
from doctor_tests.common.utils import match_rep_in_file
from doctor_tests import config
from doctor_tests.consumer import get_consumer
from doctor_tests.identity_auth import get_identity_auth
from doctor_tests.identity_auth import get_session
from doctor_tests.image import Image
from doctor_tests.instance import Instance
from doctor_tests.inspector import get_inspector
from doctor_tests.installer import get_installer
import doctor_tests.logger as doctor_log
from doctor_tests.network import Network
from doctor_tests.monitor import get_monitor
from doctor_tests.os_clients import nova_client
from doctor_tests.profiler_poc import main as profiler_main
from doctor_tests.scenario.common import calculate_notification_time
from doctor_tests.scenario.network_failure import NetworkFault
from doctor_tests.user import User


Logger = doctor_log.Logger('doctor')
LOG = Logger.getLogger()
LogFile = Logger.getLogFilename()


class DoctorTest(object):

    def __init__(self, conf):
        self.conf = conf
        self.image = Image(self.conf, LOG)
        self.user = User(self.conf, LOG)
        self.network = Network(self.conf, LOG)
        self.instance = Instance(self.conf, LOG)
        self.alarm = Alarm(self.conf, LOG)
        self.installer = get_installer(self.conf, LOG)
        self.inspector = get_inspector(self.conf, LOG)
        self.monitor = get_monitor(self.conf,
                                   self.inspector.get_inspector_url(),
                                   LOG)
        self.consumer = get_consumer(self.conf, LOG)
        self.fault = NetworkFault(self.conf, self.installer, LOG)
        auth = get_identity_auth(project=self.conf.doctor_project)
        self.nova = nova_client(self.conf.nova_version,
                                get_session(auth=auth))
        self.down_host = None

    def setup(self):
        # prepare the cloud env
        self.installer.setup()

        # preparing VM image...
        self.image.create()

        # creating test user...
        self.user.create()

    def setup_fault_management(self):
        # user settings...
        self.user.update_quota()

        # creating VM...
        self.network.create()
        self.instance.create()
        self.instance.wait_for_vm_launch()

        # creating alarm...
        self.alarm.create()

        # starting doctor sample components...
        # tbd tojuvone: move inspector and consumer to common setup
        # when they support updating VMs via instance.create and
        # instance.delete alarm

        self.inspector.start()
        self.consumer.start()
        self.down_host = self.get_host_info_for_random_vm()
        self.monitor.start(self.down_host)

    def test_fault_management(self):
        try:
            LOG.info('doctor fault management test starting.......')

            # prepare test env
            self.setup_fault_management()

            # wait for aodh alarms are updated in caches for event evaluator,
            # sleep time should be larger than event_alarm_cache_ttl
            # (default 60)
            # (tojuvone) Fraser currently needs 120
            time.sleep(120)

            # injecting host failure...
            # NOTE (umar) add INTERFACE_NAME logic to host injection

            self.fault.start(self.down_host)
            time.sleep(10)

            # verify the test results
            # NOTE (umar) copy remote monitor.log file when monitor=collectd
            self.check_host_status(self.down_host.name, 'down')

            notification_time = calculate_notification_time(LogFile)
            if notification_time < 1 and notification_time > 0:
                LOG.info('doctor fault management test successfully, '
                         'notification_time=%s' % notification_time)
            else:
                LOG.error('doctor fault management test failed, '
                          'notification_time=%s' % notification_time)
                sys.exit(1)

            if self.conf.profiler_type:
                LOG.info('doctor fault management test begin to run '
                         'profile.......')
                self.collect_logs()
                self.run_profiler()
        except Exception as e:
            LOG.error('doctor fault management test failed, '
                      'Exception=%s' % e)
            sys.exit(1)
        finally:
            self.cleanup_fault_management()

    def _amount_compute_nodes(self):
        services = self.nova.services.list(binary='nova-compute')
        return len(services)

    def test_maintenance(self):
        cnodes = self._amount_compute_nodes()
        if cnodes < 3:
            # need 2 compute for redundancy and one spare to migrate
            LOG.info('not enough compute nodes, skipping doctor '
                     'maintenance test')
            return
        try:
            LOG.info('doctor maintenance test starting.......')
            # TODO (tojuvone) test setup and actual test
        except Exception as e:
            LOG.error('doctor maintenance test failed, Exception=%s' % e)
            sys.exit(1)
        # TODO (tojuvone) finally: test case specific cleanup

    def run(self):
        """run doctor tests"""
        try:
            LOG.info('doctor test starting.......')
            # prepare common test env
            self.setup()
            if self.conf.test_case == 'all':
                self.test_fault_management()
                self.test_maintenance()
            else:
                function = 'test_%s' % self.conf.test_case
                if hasattr(self, function):
                    getattr(self, function)()
                else:
                    raise Exception('Can not find function <%s> in'
                                    'DoctorTest, see config manual'
                                    % function)
        except Exception as e:
            LOG.error('doctor test failed, Exception=%s' % e)
            sys.exit(1)
        finally:
            self.cleanup()

    def get_host_info_for_random_vm(self):
        num = random.randint(0, self.conf.instance_count - 1)
        vm_name = "%s%d" % (self.conf.instance_basename, num)

        servers = {getattr(server, 'name'): server
                   for server in self.nova.servers.list()}
        server = servers.get(vm_name)
        if not server:
            raise Exception('Can not find instance: vm_name(%s)' % vm_name)
        host_name = server.__dict__.get('OS-EXT-SRV-ATTR:hypervisor_hostname')
        host_ip = self.installer.get_host_ip_from_hostname(host_name)

        LOG.info('Get host info(name:%s, ip:%s) which vm(%s) launched at'
                 % (host_name, host_ip, vm_name))
        return Host(host_name, host_ip)

    def check_host_status(self, hostname, state):
        service = self.nova.services.list(host=hostname,
                                          binary='nova-compute')
        host_state = service[0].__dict__.get('state')
        assert host_state == state

    def unset_forced_down_hosts(self):
        if self.down_host:
            self.nova.services.force_down(self.down_host.name,
                                          'nova-compute', False)
            time.sleep(2)
            self.check_host_status(self.down_host.name, 'up')

    def collect_logs(self):
        self.fault.get_disable_network_log()

    def run_profiler(self):

        net_down_log_file = self.fault.get_disable_network_log()
        reg = '(?<=doctor set link down at )\d+.\d+'
        linkdown = float(match_rep_in_file(reg, net_down_log_file).group(0))

        reg = '(.* doctor mark vm.* error at )(\d+.\d+)'
        vmdown = float(match_rep_in_file(reg, LogFile).group(2))

        reg = '(.* doctor mark host.* down at )(\d+.\d+)'
        hostdown = float(match_rep_in_file(reg, LogFile).group(2))

        reg = '(?<=doctor monitor detected at )\d+.\d+'
        detected = float(match_rep_in_file(reg, LogFile).group(0))

        reg = '(?<=doctor consumer notified at )\d+.\d+'
        notified = float(match_rep_in_file(reg, LogFile).group(0))

        # TODO(yujunz) check the actual delay to verify time sync status
        # expected ~1s delay from $trigger to $linkdown
        relative_start = linkdown
        os.environ['DOCTOR_PROFILER_T00'] = (
            str(int((linkdown - relative_start) * 1000)))
        os.environ['DOCTOR_PROFILER_T01'] = (
            str(int((detected - relative_start) * 1000)))
        os.environ['DOCTOR_PROFILER_T03'] = (
            str(int((vmdown - relative_start) * 1000)))
        os.environ['DOCTOR_PROFILER_T04'] = (
            str(int((hostdown - relative_start) * 1000)))
        os.environ['DOCTOR_PROFILER_T09'] = (
            str(int((notified - relative_start) * 1000)))

        profiler_main(log=LOG)

    def cleanup_fault_management(self):
        self.unset_forced_down_hosts()
        self.inspector.stop()
        self.monitor.stop()
        self.consumer.stop()
        self.alarm.delete()
        self.instance.delete()
        self.network.delete()
        self.fault.cleanup()

    def cleanup(self):
        self.installer.cleanup()
        self.image.delete()
        self.user.delete()
        # Kill possible hanging subprocess
        os.killpg(0, signal.SIGKILL)


def main():
    """doctor main"""
    # TODO (tojuvone): JIRA DOCTOR-123: Test cases have some issue to always
    # kill all subprocesses. To ensure they are killed this group is done so
    # all processes can be killed without knowing what they are.
    os.setpgrp()
    test_dir = os.path.split(os.path.realpath(__file__))[0]
    doctor_root_dir = os.path.dirname(test_dir)

    config_file_dir = '{0}/{1}'.format(doctor_root_dir, 'etc/')
    config_files = [join(config_file_dir, f)
                    for f in os.listdir(config_file_dir)
                    if isfile(join(config_file_dir, f))]

    conf = config.prepare_conf(args=sys.argv[1:],
                               config_files=config_files)

    doctor = DoctorTest(conf)
    doctor.run()