Hide keyboard shortcuts

Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

1# This file is part of verify. 

2# 

3# Developed for the LSST Data Management System. 

4# This product includes software developed by the LSST Project 

5# (https://www.lsst.org). 

6# See the COPYRIGHT file at the top-level directory of this distribution 

7# for details of code ownership. 

8# 

9# This program is free software: you can redistribute it and/or modify 

10# it under the terms of the GNU General Public License as published by 

11# the Free Software Foundation, either version 3 of the License, or 

12# (at your option) any later version. 

13# 

14# This program is distributed in the hope that it will be useful, 

15# but WITHOUT ANY WARRANTY; without even the implied warranty of 

16# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 

17# GNU General Public License for more details. 

18# 

19# You should have received a copy of the GNU General Public License 

20# along with this program. If not, see <https://www.gnu.org/licenses/>. 

21"""Upload LSST Science Pipelines Verification `~lsst.verify.Job` datasets to 

22the SQUASH dashboard. 

23 

24Job JSON files can be created by `lsst.verify.Job.write` or 

25`lsst.verify.output_quantities`. A `~lsst.verify.Job` dataset consists of 

26metric measurements, associated blobs, and pipeline execution metadata. 

27Individual LSST Science Pipelines tasks typically write separate JSON datasets. 

28This command can collect and combine multiple Job JSON datasets into a single 

29Job upload. 

30 

31**Configuration** 

32 

33dispatch_verify.py is configurable from both the command line and environment 

34variables. See the argument documenation for environment variable equivalents. 

35Command line settings override environment variable configuration. 

36 

37**Metadata and environment** 

38 

39dispatch_verify.py can enrich Verification Job metadata with information 

40from the environment. Currently dispatch_verify.py supports the Jenkins CI 

41and the LSST Data Facility (LDF) execution environments. 

42 

43In the Jenkins CI execution environment (``--env=jenkins``) the 

44following environment variables are consumed: 

45 

46- ``BUILD_ID``: ID in the CI system 

47- ``BUILD_URL``: CI page with information about the build 

48- ``PRODUCT``: the name of the product built, e.g. 'validate_drp' 

49- ``dataset``: the name of the dataset processed, e.g. 'validation_data_cfht' 

50- ``label``: the name of the platform where it runs 

51- ``refs``: the branches run by Jenkins, e.g. 'tickets/DM-12345 master' 

52 

53If ``--lsstsw`` is used, additional Git branch information is included with 

54Science Pipelines package metadata. 

55 

56In the LSST Data Facility execution environment (``--env=ldf``) the following 

57environment variables are consumed: 

58 

59- ``DATASET``: the name of the dataset processed, e.g 'HSC RC2' 

60- ``DATASET_REPO_URL``: a reference URL with information about the dataset 

61- ``RUN_ID``: ID of the run in the LDF environment 

62- ``RUN_ID_URL``: a reference URL with information about the run 

63- ``VERSION_TAG``: the version tag of the LSST software used, e.g. 'w_2018_18' 

64 

65Note: currently it is not possible to gather Science Pipelines package metadata 

66in the LDF environment, thus if ``--env=ldf`` is used ``--ignore-lsstsw`` is 

67aslo used by default in this environment. 

68""" 

69# For determining what is documented in Sphinx 

70__all__ = ['build_argparser', 'main', 'insert_lsstsw_metadata', 

71 'insert_extra_package_metadata', 'insert_env_metadata', 

72 'validate_date_created', 'Configuration'] 

73 

74import argparse 

75import os 

76import json 

77import getpass 

78from dateutil import parser as date_parser 

79from datetime import datetime, timezone 

80 

81try: 

82 import git 

83except ImportError: 

84 # GitPython is not a standard Stack package; skip gracefully if unavailable 

85 git = None 

86 

87import lsst.log 

88from lsst.verify import Job 

89from lsst.verify.metadata.lsstsw import LsstswRepos 

90from lsst.verify.metadata.eupsmanifest import Manifest 

91from lsst.verify.metadata.jenkinsci import get_jenkins_env 

92from lsst.verify.metadata.ldf import get_ldf_env 

93 

94 

95def build_argparser(): 

96 parser = argparse.ArgumentParser( 

97 description=__doc__, 

98 formatter_class=argparse.RawDescriptionHelpFormatter, 

99 epilog='More information is available at https://pipelines.lsst.io.') 

100 

101 parser.add_argument( 

102 'json_paths', 

103 nargs='+', 

104 metavar='json', 

105 help='Verification job JSON file, or files. When multiple JSON ' 

106 'files are present, their measurements, blobs, and metadata ' 

107 'are merged.') 

108 parser.add_argument( 

109 '--test', 

110 default=False, 

111 action='store_true', 

112 help='Run this command without uploading to the SQUASH service. ' 

113 'The JSON payload is printed to standard out.') 

114 parser.add_argument( 

115 '--write', 

116 metavar='PATH', 

117 dest='output_filepath', 

118 help='Write the merged and enriched Job JSON dataset to the given ' 

119 'path.') 

120 parser.add_argument( 

121 '--show', 

122 dest='show_json', 

123 action='store_true', 

124 default=False, 

125 help='Print the assembled Job JSON to standard output.') 

126 parser.add_argument( 

127 '--ignore-blobs', 

128 dest='ignore_blobs', 

129 action='store_true', 

130 default=False, 

131 help='Ignore data blobs even if they are available in the verification' 

132 'job.') 

133 

134 env_group = parser.add_argument_group('Environment arguments') 

135 env_group.add_argument( 

136 '--env', 

137 dest='env_name', 

138 choices=Configuration.allowed_env, 

139 help='Name of the environment where the verification job is being ' 

140 'run. In some environments display_verify.py will gather ' 

141 'additional metadata automatically:\n' 

142 '\n' 

143 'jenkins\n' 

144 ' For the Jenkins CI (https://ci.lsst.codes)' 

145 ' environment.\n' 

146 'ldf\n' 

147 ' For the LSST Data Facility environment. \n' 

148 '\n' 

149 'Equivalent to the $VERIFY_ENV environment variable.') 

150 env_group.add_argument( 

151 '--lsstsw', 

152 dest='lsstsw', 

153 metavar='PATH', 

154 help='lsstsw directory path. If available, Stack package versions are ' 

155 'read from lsstsw. Equivalent to the ``$LSSTSW`` environment ' 

156 'variable. Disabled with ``--ignore-lsstsw.``') 

157 env_group.add_argument( 

158 '--package-repos', 

159 dest='extra_package_paths', 

160 nargs='*', 

161 metavar='PATH', 

162 help='Paths to additional Stack package Git repositories. These ' 

163 'packages are tracked in Job metadata, like lsstsw-based ' 

164 'packages.') 

165 env_group.add_argument( 

166 '--ignore-lsstsw', 

167 dest='ignore_lsstsw', 

168 action='store_true', 

169 default=False, 

170 help='Ignore lsstsw metadata even if it is available (for example, ' 

171 'the ``$LSSTSW`` variable is set).') 

172 

173 api_group = parser.add_argument_group('SQUASH API arguments') 

174 api_group.add_argument( 

175 '--url', 

176 dest='api_url', 

177 metavar='URL', 

178 help='Root URL of the SQUASH API. Equivalent to the ``$SQUASH_URL`` ' 

179 'environment variable.') 

180 api_group.add_argument( 

181 '--user', 

182 dest='api_user', 

183 metavar='USER', 

184 help='Username for SQUASH API. Equivalent to the $SQUASH_USER ' 

185 'environment variable.') 

186 api_group.add_argument( 

187 '--password', 

188 dest='api_password', 

189 metavar='PASSWORD', 

190 help='Password for SQUASH API. Equivalent to the ``$SQUASH_PASSWORD`` ' 

191 'environment variable. If neither is set, you will be prompted.') 

192 api_group.add_argument( 

193 '--date-created', 

194 dest='date_created', 

195 help='ISO8601 formatted datetime in UTC for the Job creation date, ' 

196 'e.g. 2021-06-30T22:28:25Z. If not provided the current ' 

197 'datetime is used.') 

198 return parser 

199 

200 

201def main(): 

202 """Entrypoint for the ``dispatch_verify.py`` command line executable. 

203 """ 

204 log = lsst.log.Log.getLogger('verify.bin.dispatchverify.main') 

205 

206 parser = build_argparser() 

207 args = parser.parse_args() 

208 config = Configuration(args) 

209 log.debug(str(config)) 

210 

211 # Parse all Job JSON 

212 jobs = [] 

213 for json_path in config.json_paths: 

214 log.info('Loading {0}'.format(json_path)) 

215 with open(json_path) as fp: 

216 json_data = json.load(fp) 

217 # Ignore blobs from the verification jobs 

218 if config.ignore_blobs: 

219 log.info('Ignoring blobs from Job JSON {0}'.format(json_path)) 

220 json_data = delete_blobs(json_data) 

221 job = Job.deserialize(**json_data) 

222 jobs.append(job) 

223 

224 # Merge all Jobs into one 

225 job = jobs.pop(0) 

226 if len(jobs) > 0: 

227 log.info('Merging verification Job JSON.') 

228 for other_job in jobs: 

229 job += other_job 

230 

231 # Ensure all measurements have a metric so that units are normalized 

232 log.info('Refreshing metric definitions from verify_metrics') 

233 job.reload_metrics_package('verify_metrics') 

234 

235 # Insert package metadata from lsstsw 

236 if not config.ignore_lsstsw: 

237 log.info('Inserting lsstsw package metadata from ' 

238 '{0}.'.format(config.lsstsw)) 

239 job = insert_lsstsw_metadata(job, config) 

240 

241 # Insert metadata from additional specified packages 

242 if config.extra_package_paths is not None: 

243 job = insert_extra_package_metadata(job, config) 

244 

245 # Add environment variable metadata from the Jenkins CI environment 

246 if config.env_name == 'jenkins': 

247 log.info('Inserting Jenkins CI environment metadata.') 

248 jenkins_metadata = get_jenkins_env() 

249 job = insert_env_metadata(job, 'jenkins', jenkins_metadata, 

250 config.date_created) 

251 elif config.env_name == 'ldf': 

252 log.info('Inserting LSST Data Facility environment metadata.') 

253 ldf_metadata = get_ldf_env() 

254 job = insert_env_metadata(job, 'ldf', ldf_metadata, 

255 config.date_created) 

256 

257 # Upload job 

258 if not config.test: 

259 log.info('Uploading Job JSON to {0}.'.format(config.api_url)) 

260 job.dispatch(api_user=config.api_user, 

261 api_password=config.api_password, 

262 api_url=config.api_url) 

263 

264 if config.show_json: 

265 print(json.dumps(job.json, 

266 sort_keys=True, indent=4, separators=(',', ': '))) 

267 

268 # Write a json file 

269 if config.output_filepath is not None: 

270 log.info('Writing Job JSON to {0}.'.format(config.output_filepath)) 

271 job.write(config.output_filepath) 

272 

273 

274def delete_blobs(json_data): 

275 """Delete data blobs from the Job JSON 

276 """ 

277 if 'blobs' in json_data: 

278 del json_data['blobs'] 

279 return json_data 

280 

281 

282def insert_lsstsw_metadata(job, config): 

283 """Insert metadata for lsstsw-based packages into ``Job.meta['packages']``. 

284 """ 

285 lsstsw_repos = LsstswRepos(config.lsstsw) 

286 

287 with open(lsstsw_repos.manifest_path) as fp: 

288 manifest = Manifest(fp) 

289 

290 packages = {} 

291 for package_name, manifest_item in manifest.items(): 

292 package_doc = { 

293 'name': package_name, 

294 'git_branch': lsstsw_repos.get_package_branch(package_name), 

295 'git_url': lsstsw_repos.get_package_repo_url(package_name), 

296 'git_sha': manifest_item.git_sha, 

297 'eups_version': manifest_item.version 

298 } 

299 packages[package_name] = package_doc 

300 

301 if 'packages' in job.meta: 

302 # Extend packages entry 

303 job.meta['packages'].update(packages) 

304 else: 

305 # Create new packages entry 

306 job.meta['packages'] = packages 

307 return job 

308 

309 

310def insert_extra_package_metadata(job, config): 

311 """Insert metadata for extra packages ('--package-repos') into 

312 ``Job.meta['packages']``. 

313 """ 

314 log = lsst.log.Log.getLogger( 

315 'verify.bin.dispatchverify.insert_extra_package_metadata') 

316 

317 if 'packages' not in job.meta: 

318 job.meta['packages'] = dict() 

319 

320 for package_path in config.extra_package_paths: 

321 log.info('Inserting extra package metadata: {0}'.format(package_path)) 

322 package_name = package_path.split(os.sep)[-1] 

323 

324 package = {'name': package_name} 

325 

326 if git is not None: 

327 git_repo = git.Repo(package_path) 

328 package['git_sha'] = git_repo.active_branch.commit.hexsha 

329 package['git_branch'] = git_repo.active_branch.name 

330 package['git_url'] = git_repo.remotes.origin.url 

331 

332 if package_name in job.meta['packages']: 

333 # Update pre-existing package metadata 

334 job.meta['packages'][package_name].update(package) 

335 else: 

336 # Create new package metadata 

337 job.meta['packages'][package_name] = package 

338 

339 return job 

340 

341 

342def insert_env_metadata(job, env_name, metadata, date_created): 

343 """Insert environment metadata into the Job. 

344 """ 

345 metadata.update({'env_name': env_name}) 

346 

347 if date_created is not None: 

348 date = date_created 

349 else: 

350 date = datetime.now(timezone.utc).isoformat() 

351 

352 metadata.update({'date': date}) 

353 

354 job.meta['env'] = metadata 

355 

356 return job 

357 

358 

359def validate_date_created(date_created): 

360 """Ensure date_created is a valid datetime string in UTC. 

361 """ 

362 try: 

363 date = date_parser.parse(date_created) 

364 except ValueError: 

365 return False 

366 

367 if date.tzname() == 'UTC': 

368 return True 

369 else: 

370 return False 

371 

372 

373class Configuration(object): 

374 """Configuration for dispatch_verify.py that reconciles command line and 

375 environment variable arguments. 

376 

377 Configuration is validated for completeness and certain errors. 

378 

379 Parameters 

380 ---------- 

381 args : `argparse.Namespace` 

382 Parsed command line arguments, produced by `parse_args`. 

383 """ 

384 

385 allowed_env = ('jenkins', 'ldf') 

386 

387 def __init__(self, args): 

388 self.json_paths = args.json_paths 

389 

390 self.test = args.test 

391 

392 self.output_filepath = args.output_filepath 

393 

394 self.show_json = args.show_json 

395 

396 self.env_name = args.env_name or os.getenv('VERIFY_ENV') 

397 if self.env_name is not None and self.env_name not in self.allowed_env: 

398 message = '$VERIFY_ENV not one of {0!s}'.format(self.allowed_env) 

399 raise RuntimeError(message) 

400 

401 self.ignore_blobs = args.ignore_blobs 

402 

403 self.ignore_lsstsw = args.ignore_lsstsw 

404 

405 # Make sure --ignore-lsstw is used in the LDF environment 

406 if self.env_name == 'ldf': 

407 self.ignore_lsstsw = True 

408 

409 self.lsstsw = args.lsstsw or os.getenv('LSSTSW') 

410 if self.lsstsw is not None: 

411 self.lsstsw = os.path.abspath(self.lsstsw) 

412 if not self.ignore_lsstsw and not self.lsstsw: 

413 message = 'lsstsw directory not found at {0}'.format(self.lsstsw) 

414 raise RuntimeError(message) 

415 

416 if args.extra_package_paths is not None: 

417 self.extra_package_paths = [os.path.abspath(p) 

418 for p in args.extra_package_paths] 

419 else: 

420 self.extra_package_paths = [] 

421 for path in self.extra_package_paths: 

422 if not os.path.isdir(path): 

423 message = 'Package directory not found: {0}'.format(path) 

424 raise RuntimeError(message) 

425 

426 default_url = 'https://squash.lsst.codes/dashboard/api' 

427 self.api_url = args.api_url or os.getenv('SQUASH_URL', default_url) 

428 

429 self.api_user = args.api_user or os.getenv('SQUASH_USER') 

430 if not self.test and self.api_user is None: 

431 message = '--user or $SQUASH_USER configuration required' 

432 raise RuntimeError(message) 

433 

434 self.api_password = (args.api_password 

435 or os.getenv('SQUASH_password')) 

436 if not self.test and self.api_password is None: 

437 # If password hasn't been set, prompt for it. 

438 self.api_password = getpass.getpass(prompt="SQuaSH password: ") 

439 

440 self.date_created = args.date_created 

441 

442 if self.date_created is not None: 

443 if not validate_date_created(self.date_created): 

444 message = 'Invalid datetime string, use a ISO8601 formatted ' \ 

445 'datetime in UTC, e.g. 2021-06-30T22:28:25Z.' 

446 raise RuntimeError(message) 

447 else: 

448 self.date_created = \ 

449 date_parser.parse(self.date_created).isoformat() 

450 

451 def __str__(self): 

452 configs = { 

453 'json_paths': self.json_paths, 

454 'test': self.test, 

455 'output_filepath': self.output_filepath, 

456 'show_json': self.show_json, 

457 'ignore_blobs': self.ignore_blobs, 

458 'env': self.env_name, 

459 'ignore_lsstsw': self.ignore_lsstsw, 

460 'lsstsw': self.lsstsw, 

461 'extra_package_paths': self.extra_package_paths, 

462 'api_url': self.api_url, 

463 'api_user': self.api_user, 

464 'date_created': self.date_created, 

465 } 

466 if self.api_password is None: 

467 configs['api_password'] = None 

468 else: 

469 configs['api_password'] = '*' * len(self.api_password) 

470 

471 return json.dumps(configs, 

472 sort_keys=True, indent=4, separators=(',', ': '))