def run_job(self, job): with tempfile.NamedTemporaryFile() as f: f.write(job.query()) f.flush() arglist = [load_hive_cmd(), '-f', f.name] if job.hiverc(): arglist += ['-i', job.hiverc()] logger.info(arglist) HadoopJobRunner.run_and_track_hadoop_job(arglist)
def run_job(self, job): # TODO(jcrobak): libjars, files, etc. Can refactor out of # hadoop.HadoopJobRunner if not job.jar() or not os.path.exists(job.jar()): logger.error("Can't find jar: {0}, full path {1}".format(job.jar(), os.path.abspath(job.jar()))) raise Exception("job jar does not exist") arglist = ['hadoop', 'jar', job.jar()] if job.main(): arglist.append(job.main()) jobconfs = job.jobconfs() for jc in jobconfs: arglist += ['-D' + jc] (tmp_files, job_args) = HadoopJarJobRunner._fix_paths(job) arglist += job_args HadoopJobRunner.run_and_track_hadoop_job(arglist) for a, b in tmp_files: a.move(b)