aboutsummaryrefslogtreecommitdiffhomepage
path: root/tools/run_tests/run_build_statistics.py
blob: 654cf95a38c51a126035fca8eec69d4599066681 (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
#!/usr/bin/env python2.7
# Copyright 2016, Google Inc.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
#     * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
#     * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following disclaimer
# in the documentation and/or other materials provided with the
# distribution.
#     * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

"""Tool to get build statistics from Jenkins and upload to BigQuery."""

import argparse
import jenkinsapi
from jenkinsapi.custom_exceptions import JenkinsAPIException
from jenkinsapi.jenkins import Jenkins
import json
import os
import re
import sys
import urllib


gcp_utils_dir = os.path.abspath(os.path.join(
    os.path.dirname(__file__), '../gcp/utils'))
sys.path.append(gcp_utils_dir)
import big_query_utils


_PROJECT_ID = 'grpc-testing'
_HAS_MATRIX = True
_BUILDS = {'gRPC_interop_master': not _HAS_MATRIX,
           'gRPC_master_linux': not _HAS_MATRIX,
           'gRPC_master_macos': not _HAS_MATRIX,
           'gRPC_master_windows': not _HAS_MATRIX,
           'gRPC_performance_master': not _HAS_MATRIX,
           'gRPC_portability_master_linux': not _HAS_MATRIX,
           'gRPC_portability_master_windows': not _HAS_MATRIX,
           'gRPC_master_asanitizer_c': not _HAS_MATRIX,
           'gRPC_master_asanitizer_cpp': not _HAS_MATRIX,
           'gRPC_master_msan_c': not _HAS_MATRIX,
           'gRPC_master_tsanitizer_c': not _HAS_MATRIX,
           'gRPC_master_tsan_cpp': not _HAS_MATRIX,
           'gRPC_interop_pull_requests': not _HAS_MATRIX,
           'gRPC_performance_pull_requests': not _HAS_MATRIX,
           'gRPC_portability_pull_requests_linux': not _HAS_MATRIX,
           'gRPC_portability_pr_win': not _HAS_MATRIX,
           'gRPC_pull_requests_linux': not _HAS_MATRIX,
           'gRPC_pull_requests_macos': not _HAS_MATRIX,
           'gRPC_pr_win': not _HAS_MATRIX,
           'gRPC_pull_requests_asan_c': not _HAS_MATRIX,
           'gRPC_pull_requests_asan_cpp': not _HAS_MATRIX,
           'gRPC_pull_requests_msan_c': not _HAS_MATRIX,
           'gRPC_pull_requests_tsan_c': not _HAS_MATRIX,
           'gRPC_pull_requests_tsan_cpp': not _HAS_MATRIX,
}
_URL_BASE = 'https://grpc-testing.appspot.com/job'

# This is a dynamic list where known and active issues should be added. 
# Fixed ones should be removed.
# Also try not to add multiple messages from the same failure.
_KNOWN_ERRORS = [
    'Failed to build workspace Tests with scheme AllTests',
    'Build timed out',
    'TIMEOUT: tools/run_tests/pre_build_node.sh',
    'TIMEOUT: tools/run_tests/pre_build_ruby.sh',
    'FATAL: Unable to produce a script file',
    'FAILED: build_docker_c\+\+',
    'cannot find package \"cloud.google.com/go/compute/metadata\"',
    'LLVM ERROR: IO failure on output stream.',
    'MSBUILD : error MSB1009: Project file does not exist.',
    'fatal: git fetch_pack: expected ACK/NAK',
    'Failed to fetch from http://github.com/grpc/grpc.git',
    ('hudson.remoting.RemotingSystemException: java.io.IOException: '
     'Backing channel is disconnected.'),
    'hudson.remoting.ChannelClosedException',
    'Could not initialize class hudson.Util',
    'Too many open files in system',
    'FAILED: bins/tsan/qps_openloop_test GRPC_POLL_STRATEGY=epoll',
    'FAILED: bins/tsan/qps_openloop_test GRPC_POLL_STRATEGY=legacy',
    'FAILED: bins/tsan/qps_openloop_test GRPC_POLL_STRATEGY=poll',
    ('tests.bins/asan/h2_proxy_test streaming_error_response '
     'GRPC_POLL_STRATEGY=legacy'),
]
_NO_REPORT_FILES_FOUND_ERROR = 'No test report files were found. Configuration error?'
_UNKNOWN_ERROR = 'Unknown error'
_DATASET_ID = 'build_statistics'


def _scrape_for_known_errors(html):
  error_list = []
  known_error_count = 0
  for known_error in _KNOWN_ERRORS:
    errors = re.findall(known_error, html)
    this_error_count = len(errors)
    if this_error_count > 0: 
      known_error_count += this_error_count
      error_list.append({'description': known_error,
                         'count': this_error_count})
      print('====> %d failures due to %s' % (this_error_count, known_error))
  return error_list, known_error_count


def _no_report_files_found(html):
  return _NO_REPORT_FILES_FOUND_ERROR in html


def _get_last_processed_buildnumber(build_name):
  query = 'SELECT max(build_number) FROM [%s:%s.%s];' % (
      _PROJECT_ID, _DATASET_ID, build_name)
  query_job = big_query_utils.sync_query_job(bq, _PROJECT_ID, query)
  page = bq.jobs().getQueryResults(
      pageToken=None,
      **query_job['jobReference']).execute(num_retries=3)
  if page['rows'][0]['f'][0]['v']:
    return int(page['rows'][0]['f'][0]['v'])
  return 0


def _process_matrix(build, url_base):
  matrix_list = []
  for matrix in build.get_matrix_runs():
    matrix_str = re.match('.*\\xc2\\xbb ((?:[^,]+,?)+) #.*', 
                          matrix.name).groups()[0]
    matrix_tuple = matrix_str.split(',')
    json_url = '%s/config=%s,language=%s,platform=%s/testReport/api/json' % (
        url_base, matrix_tuple[0], matrix_tuple[1], matrix_tuple[2])
    console_url = '%s/config=%s,language=%s,platform=%s/consoleFull' % (
        url_base, matrix_tuple[0], matrix_tuple[1], matrix_tuple[2])
    matrix_dict = {'name': matrix_str,
                   'duration': matrix.get_duration().total_seconds()}
    matrix_dict.update(_process_build(json_url, console_url))
    matrix_list.append(matrix_dict)

  return matrix_list 


def _process_build(json_url, console_url):
  build_result = {}
  error_list = []
  try:
    html = urllib.urlopen(json_url).read()
    test_result = json.loads(html)
    print('====> Parsing result from %s' % json_url)
    failure_count = test_result['failCount']
    build_result['pass_count'] = test_result['passCount']
    build_result['failure_count'] = failure_count
    build_result['no_report_files_found'] = _no_report_files_found(html)
    if failure_count > 0:
      error_list, known_error_count = _scrape_for_known_errors(html)
      unknown_error_count = failure_count - known_error_count
      # This can happen if the same error occurs multiple times in one test.
      if failure_count < known_error_count:
        print('====> Some errors are duplicates.')
        unknown_error_count = 0
      error_list.append({'description': _UNKNOWN_ERROR, 
                         'count': unknown_error_count})
  except Exception as e:
    print('====> Got exception for %s: %s.' % (json_url, str(e)))   
    print('====> Parsing errors from %s.' % console_url)
    html = urllib.urlopen(console_url).read()
    build_result['pass_count'] = 0  
    build_result['failure_count'] = 1
    error_list, _ = _scrape_for_known_errors(html)
    if error_list:
      error_list.append({'description': _UNKNOWN_ERROR, 'count': 0})
    else:
      error_list.append({'description': _UNKNOWN_ERROR, 'count': 1})
 
  if error_list:
    build_result['error'] = error_list

  return build_result 


# parse command line
argp = argparse.ArgumentParser(description='Get build statistics.')
argp.add_argument('-u', '--username', default='jenkins')
argp.add_argument('-b', '--builds', 
                  choices=['all'] + sorted(_BUILDS.keys()),
                  nargs='+',
                  default=['all'])
args = argp.parse_args()

J = Jenkins('https://grpc-testing.appspot.com', args.username, 'apiToken')
bq = big_query_utils.create_big_query()

for build_name in _BUILDS.keys() if 'all' in args.builds else args.builds:
  print('====> Build: %s' % build_name)
  # Since get_last_completed_build() always fails due to malformatted string
  # error, we use get_build_metadata() instead.
  job = None
  try:
    job = J[build_name]
  except Exception as e:
    print('====> Failed to get build %s: %s.' % (build_name, str(e)))
    continue
  last_processed_build_number = _get_last_processed_buildnumber(build_name)
  last_complete_build_number = job.get_last_completed_buildnumber()
  # To avoid processing all builds for a project never looked at. In this case,
  # only examine 10 latest builds.
  starting_build_number = max(last_processed_build_number+1, 
                              last_complete_build_number-9)
  for build_number in xrange(starting_build_number, 
                             last_complete_build_number+1):
    print('====> Processing %s build %d.' % (build_name, build_number))
    build = None
    try:
      build = job.get_build_metadata(build_number)
    except KeyError:
      print('====> Build %s is missing. Skip.' % build_number)
      continue
    build_result = {'build_number': build_number, 
                    'timestamp': str(build.get_timestamp())}
    url_base = json_url = '%s/%s/%d' % (_URL_BASE, build_name, build_number)
    if _BUILDS[build_name]:  # The build has matrix, such as gRPC_master.
      build_result['matrix'] = _process_matrix(build, url_base)
    else:
      json_url = '%s/testReport/api/json' % url_base
      console_url = '%s/consoleFull' % url_base
      build_result['duration'] = build.get_duration().total_seconds()
      build_result.update(_process_build(json_url, console_url))
    rows = [big_query_utils.make_row(build_number, build_result)]
    if not big_query_utils.insert_rows(bq, _PROJECT_ID, _DATASET_ID, build_name, 
                                       rows):
      print '====> Error uploading result to bigquery.'
      sys.exit(1)