35d648c9cf
In some part in the code we import objects. In the Openstack style guidelines they recommend to import only modules. http://docs.openstack.org/developer/hacking/#imports Change-Id: Icae231f06f3c4fd15256f06a464d3ba3e2845e33
183 lines
7.8 KiB
Python
183 lines
7.8 KiB
Python
# 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 Connmon
|
|
import datetime
|
|
import glob
|
|
import Grafana
|
|
import logging
|
|
import os
|
|
import shutil
|
|
import subprocess
|
|
import time
|
|
import Tools
|
|
import WorkloadBase
|
|
|
|
|
|
class PerfKit(WorkloadBase.WorkloadBase):
|
|
|
|
def __init__(self, config):
|
|
self.logger = logging.getLogger('browbeat.PerfKit')
|
|
self.config = config
|
|
self.error_count = 0
|
|
self.tools = Tools.Tools(self.config)
|
|
self.connmon = Connmon.Connmon(self.config)
|
|
self.grafana = Grafana.Grafana(self.config)
|
|
self.test_count = 0
|
|
self.scenario_count = 0
|
|
self.pass_count = 0
|
|
|
|
def _log_details(self):
|
|
self.logger.info(
|
|
"Current number of Perkit scenarios executed: {}".format(self.scenario_count))
|
|
self.logger.info(
|
|
"Current number of Perfkit test(s) executed: {}".format(self.test_count))
|
|
self.logger.info(
|
|
"Current number of Perfkit test(s) succeeded: {}".format(self.pass_count))
|
|
self.logger.info(
|
|
"Current number of Perfkit test failures: {}".format(self.error_count))
|
|
|
|
def update_tests(self):
|
|
self.test_count += 1
|
|
|
|
def update_pass_tests(self):
|
|
self.pass_count += 1
|
|
|
|
def update_fail_tests(self):
|
|
self.error_count += 1
|
|
|
|
def update_scenarios(self):
|
|
self.scenario_count += 1
|
|
|
|
def run_benchmark(self, benchmark_config, result_dir, test_name, cloud_type="OpenStack"):
|
|
self.logger.debug("--------------------------------")
|
|
self.logger.debug("Benchmark_config: {}".format(benchmark_config))
|
|
self.logger.debug("result_dir: {}".format(result_dir))
|
|
self.logger.debug("test_name: {}".format(test_name))
|
|
self.logger.debug("--------------------------------")
|
|
|
|
# Build command to run
|
|
if 'enabled' in benchmark_config:
|
|
del benchmark_config['enabled']
|
|
cmd = ("source /home/stack/overcloudrc; source {0}; "
|
|
"/home/stack/perfkit-venv/PerfKitBenchmarker/pkb.py "
|
|
"--cloud={1} --run_uri=browbeat".format(self.config['perfkit']['venv'], cloud_type))
|
|
# Add default parameters as necessary
|
|
for default_item, value in self.config['perfkit']['default'].iteritems():
|
|
if default_item not in benchmark_config:
|
|
benchmark_config[default_item] = value
|
|
for parameter, value in benchmark_config.iteritems():
|
|
if not parameter == 'name':
|
|
self.logger.debug(
|
|
"Parameter: {}, Value: {}".format(parameter, value))
|
|
cmd += " --{}={}".format(parameter, value)
|
|
|
|
# Remove any old results
|
|
if os.path.exists("/tmp/perfkitbenchmarker/run_browbeat"):
|
|
shutil.rmtree("/tmp/perfkitbenchmarker/run_browbeat")
|
|
|
|
if self.config['connmon']['enabled']:
|
|
self.connmon.start_connmon()
|
|
|
|
self.logger.info("Running Perfkit Command: {}".format(cmd))
|
|
stdout_file = open("{}/pkb.stdout.log".format(result_dir), 'w')
|
|
stderr_file = open("{}/pkb.stderr.log".format(result_dir), 'w')
|
|
from_ts = time.time()
|
|
if 'sleep_before' in self.config['perfkit']:
|
|
time.sleep(self.config['perfkit']['sleep_before'])
|
|
process = subprocess.Popen(
|
|
cmd, shell=True, stdout=stdout_file, stderr=stderr_file)
|
|
process.communicate()
|
|
if 'sleep_after' in self.config['perfkit']:
|
|
time.sleep(self.config['perfkit']['sleep_after'])
|
|
to_ts = time.time()
|
|
|
|
# Stop connmon at end of perfkit task
|
|
if self.config['connmon']['enabled']:
|
|
self.connmon.stop_connmon()
|
|
try:
|
|
self.connmon.move_connmon_results(result_dir, test_name)
|
|
self.connmon.connmon_graphs(result_dir, test_name)
|
|
except Exception:
|
|
self.logger.error(
|
|
"Connmon Result data missing, Connmon never started")
|
|
|
|
workload = self.__class__.__name__
|
|
new_test_name = test_name.split('-')
|
|
new_test_name = new_test_name[2:]
|
|
new_test_name = '-'.join(new_test_name)
|
|
# Determine success
|
|
try:
|
|
with open("{}/pkb.stderr.log".format(result_dir), 'r') as stderr:
|
|
if any('SUCCEEDED' in line for line in stderr):
|
|
self.logger.info("Benchmark completed.")
|
|
self.update_pass_tests()
|
|
self.update_total_pass_tests()
|
|
self.get_time_dict(
|
|
to_ts, from_ts, benchmark_config[
|
|
'benchmarks'], new_test_name,
|
|
workload, "pass")
|
|
else:
|
|
self.logger.error("Benchmark failed.")
|
|
self.update_fail_tests()
|
|
self.update_total_fail_tests()
|
|
self.get_time_dict(
|
|
to_ts, from_ts, benchmark_config[
|
|
'benchmarks'], new_test_name,
|
|
workload, "fail")
|
|
except IOError:
|
|
self.logger.error(
|
|
"File missing: {}/pkb.stderr.log".format(result_dir))
|
|
|
|
# Copy all results
|
|
for perfkit_file in glob.glob("/tmp/perfkitbenchmarker/run_browbeat/*"):
|
|
shutil.move(perfkit_file, result_dir)
|
|
if os.path.exists("/tmp/perfkitbenchmarker/run_browbeat"):
|
|
shutil.rmtree("/tmp/perfkitbenchmarker/run_browbeat")
|
|
|
|
# Grafana integration
|
|
self.grafana.create_grafana_urls(
|
|
{'from_ts': int(from_ts * 1000),
|
|
'to_ts': int(to_ts * 1000)})
|
|
self.grafana.print_dashboard_url(test_name)
|
|
self.grafana.log_snapshot_playbook_cmd(
|
|
from_ts, to_ts, result_dir, test_name)
|
|
self.grafana.run_playbook(from_ts, to_ts, result_dir, test_name)
|
|
|
|
def start_workloads(self):
|
|
self.logger.info("Starting PerfKitBenchmarker Workloads.")
|
|
time_stamp = datetime.datetime.utcnow().strftime("%Y%m%d-%H%M%S")
|
|
self.logger.debug("Time Stamp (Prefix): {}".format(time_stamp))
|
|
benchmarks = self.config.get('perfkit')['benchmarks']
|
|
if (benchmarks is not None and len(benchmarks) > 0):
|
|
for benchmark in benchmarks:
|
|
if benchmark['enabled']:
|
|
self.logger.info("Benchmark: {}".format(benchmark['name']))
|
|
self.update_scenarios()
|
|
self.update_total_scenarios()
|
|
for run in range(self.config['browbeat']['rerun']):
|
|
self.update_tests()
|
|
self.update_total_tests()
|
|
result_dir = self.tools.create_results_dir(
|
|
self.config['browbeat']['results'], time_stamp, benchmark['name'], run)
|
|
test_name = "{}-{}-{}".format(time_stamp,
|
|
benchmark['name'], run)
|
|
workload = self.__class__.__name__
|
|
self.workload_logger(result_dir, workload)
|
|
self.run_benchmark(benchmark, result_dir, test_name)
|
|
self._log_details()
|
|
else:
|
|
self.logger.info(
|
|
"Skipping {} benchmark, enabled: false".format(benchmark['name']))
|
|
else:
|
|
self.logger.error("Config file contains no perfkit benchmarks.")
|