From 33818a3dcc35351012300ffdcb7de0c86349575d Mon Sep 17 00:00:00 2001 From: wangjian Date: Fri, 15 Dec 2023 09:49:59 +0000 Subject: [PATCH] =?UTF-8?q?=E5=8E=BB=E9=99=A4dateutil=E4=BE=9D=E8=B5=96?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: wangjian --- LSF-Script/job/query | 77 ++++++++++++++++++++++++++++---------------- 1 file changed, 50 insertions(+), 27 deletions(-) diff --git a/LSF-Script/job/query b/LSF-Script/job/query index 3435e9f..6655ffa 100644 --- a/LSF-Script/job/query +++ b/LSF-Script/job/query @@ -4,11 +4,6 @@ # An example script for query job from LSF cluster. # Method "main" needs to be declared in this script for job query. -import json -import os -import sys -import re -from dateutil import parser # expected output format: """ @@ -95,6 +90,16 @@ from dateutil import parser "code":"success" } """ + +import json +import os +import sys +import re +from datetime import datetime + +B_JOBS_CMD = 'source @SCHEDULER_PROFILE_PATH@; timeout 10 bjobs -UF {} 2>&1' +B_HIST_CMD = 'source @SCHEDULER_PROFILE_PATH@; timeout 10 bhist -UF {} 2>&1' + JOB_ID_LIST = list() if len(sys.argv) > 1: @@ -109,13 +114,10 @@ SUCCESS = 'successData' ERROR = 'errorData' DATA_MAP = { - 'successData': [], - 'errorData': dict() + SUCCESS: [], + ERROR: dict() } -B_JOBS_CMD = 'source @SCHEDULER_PROFILE_PATH@; timeout 10 bjobs -UF {} 2>&1' -B_HIST_CMD = 'source @SCHEDULER_PROFILE_PATH@; timeout 10 bhist -UF {} 2>&1' - # jobStateMap for donau STATE_MAP = { 'UNKWN': 'UNRECOGNIZED', @@ -138,12 +140,37 @@ def main(): else: for _job_id in JOB_ID_LIST: query_job_info_with_id(_job_id) + # 处理时间格式问题 + _parse_format_time() result['message'] = 'List job detail infos successfully.' result['code'] = 'success' result['data'] = DATA_MAP print(json.dumps(result)) +def _parse_format_time(): + success_data = DATA_MAP.get(SUCCESS) + if not success_data: + return + is_time_year_fmt = len(success_data[0]['createTime'].strip(':').split(' ')[-1]) == 4 + for data in DATA_MAP.get(SUCCESS): + datetime_ = datetime.today() + for key in ['endTime', 'startTime', 'createTime']: + if not data[key]: + data[key] = '-' + else: + if is_time_year_fmt: + data[key] = datetime.strftime(datetime.strptime(data[key], '%a %b %d %H:%M:%S %Y'), '%Y-%m-%dT%H:%M:%SZ') + else: + value_time = datetime.strptime(data[key], '%a %b %d %H:%M:%S') + value_time = value_time.replace(year=datetime_.year) + if value_time > datetime_: + value_time = value_time.replace(year=value_time.year-1) + datetime_ = value_time + data[key] = datetime.strftime(value_time, '%Y-%m-%dT%H:%M:%SZ') + + + def query_job_info_with_id(job_id): # 先使用bjobs查询，不存在则使用bhost查询。如果bhost依然不存在，则报错。 if not _query_job_by_bjobs(job_id): @@ -199,8 +226,7 @@ def _parse_job_infos(job_info): error_file = _safe_get_value('Error File <', line) if error_file: dct.get('logRedirectPath')['stderrRedirectPath'] = error_file - datetime_struct = parser.parse(line.split(': Submitted from host')[0]) - dct['createTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['createTime'] = line.split(': Submitted from host')[0] pre_cmd = _safe_get_value('Pre-execution Command <', line) if pre_cmd: dct.get('preHook')['cmd'] = pre_cmd @@ -210,23 +236,25 @@ def _parse_job_infos(job_info): continue if ': Dispatched ' in line: dct['taskExecNodes'] = _safe_get_value('on Host\\(s\\) <', line) + if ' Started ' in line: + match_str = ': Started ' + dct['startTime'] = line.split(match_str)[0] + continue if ' started ' in line: match_str = ' started ' if '[' in dct['jobId']: index_val = re.findall(r'(\[.*?\])', dct['jobId'])[0] match_str = ': ' + index_val + match_str - datetime_struct = parser.parse(line.split(match_str)[0]) - dct['startTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['startTime'] = line.split(match_str)[0] + continue # bjobs查询时会展示作业状态，不需要在此更新job的状态 if ': Done successfully.' in line: dct['exitCode'] = 0 dct['exitMessage'] = 'Done successfully.' - datetime_struct = parser.parse(line.split(': Done successfully.')[0]) - dct['endTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['endTime'] = line.split(': Done successfully.')[0] continue if ': Completed ' in line: - datetime_struct = parser.parse(line.split(': Completed ')[0]) - dct['endTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['endTime'] = line.split(': Completed ')[0] if ': Completed ;' in line: dct['exitMessage'] = line.split(': Completed ;')[1] if 'Exited with exit code' in line: @@ -316,8 +344,7 @@ def _query_job_by_bhist(job_id): error_file = _safe_get_value('Error File <', line) if error_file: dct.get('logRedirectPath')['stderrRedirectPath'] = error_file - datetime_struct = parser.parse(line.split(': Submitted from host')[0]) - dct['createTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['createTime'] = line.split(': Submitted from host')[0] pre_cmd = _safe_get_value('Pre-execution Command <', line) if pre_cmd: dct.get('preHook')['cmd'] = pre_cmd @@ -331,20 +358,17 @@ def _query_job_by_bhist(job_id): dct['taskExecNodes'] = _safe_get_value('on Host\\(s\\) <', line) if ': Starting ' in line: match_str = ': Starting' - datetime_struct = parser.parse(line.split(match_str)[0]) - dct['startTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['startTime'] = line.split(match_str)[0] # bjobs查询时会展示作业状态，不需要在此更新job的状态 if ': Done successfully.' in line: dct['exitCode'] = 0 dct['exitMessage'] = 'Done successfully.' - datetime_struct = parser.parse(line.split(': Done successfully.')[0]) - dct['endTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['endTime'] = line.split(': Done successfully.')[0] dct['state'] = 'SUCCEEDED' continue if ': Completed ' in line: dct['exitMessage'] = 'Done successfully' - datetime_struct = parser.parse(line.split(': Completed ')[0]) - dct['endTime'] = datetime_struct.strftime('%Y-%m-%dT%H:%M:%SZ') + dct['endTime'] = line.split(': Completed ')[0] dct['state'] = 'FAILED' if ': Completed ;' in line: dct['exitMessage'] = line.split(': Completed ;')[1] @@ -363,4 +387,3 @@ def _query_job_by_bhist(job_id): if __name__ == '__main__': main() - -- Gitee