def get_executions_data(flow_name):
ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"],
MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"])
try:
sql = "select id from merce_flow_execution where flow_name like '%s%%%%' order by create_time desc limit 1" % flow_name
flow_info = ms.ExecuQuery(sql.encode('utf-8'))
print(sql)
print('execution_Id:', flow_info[0]["id"])
except:
return
new_data = {
"fieldList": [{
"fieldName": "executionId",
"fieldValue": flow_info[0]["id"],
"comparatorOperator": "EQUAL",
"logicalOperator": "AND"
}],
"sortObject": {
"field": "lastModifiedTime",
"orderDirection": "DESC"
},
"offset":
0,
"limit":
8
}
return new_data
class GetLog(object):
def __init__(self, execution_id, host):
self.ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"], MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"])
self.execution_id = execution_id
self.host = host
def get_job_id(self):
sql = "select flow_id ,job_id from merce_flow_execution where id = '%s' " % self.execution_id
try:
flow_job_id = self.ms.ExecuQuery(sql)
flow_id = flow_job_id[0]["flow_id"]
job_id = flow_job_id[0]["job_id"]
# print(flow_id, job_id)
return flow_id, job_id
except Exception:
return
def get_log_url(self):
"""
根据flow id和job id，查询任务日志
1. 若任务生成application ID，则进入yarn直接查看日志
2. 若任务没有生成application ID,则进入系统页面查看是否有日志生成
"""
flow_id = self.get_job_id()[0]
job_id = self.get_job_id()[1]
detail_url = "%s/api/executions/query" % self.host
data = {"fieldList": [{"fieldName": "flowId", "fieldValue": flow_id, "comparatorOperator":"EQUAL","logicalOperator":"AND"},{"fieldName":"jobId","fieldValue":job_id,"comparatorOperator":"EQUAL"}],"sortObject":{"field":"lastModifiedTime","orderDirection":"DESC"},"offset":0,"limit":8}
response = requests.post(url=detail_url, headers=get_headers(self.host), json=data)
# print(response.text)
try:
content_info = dict_res(response.text)["content"][0]
# print(content_info)
content_info_list = list(content_info.keys())
# 优先判断返回的detail信息中是否包含APPId,存在就到yarn上查看对应的日志
if 'appId' in content_info_list:
appId = content_info["appId"]
yarn_url = "http://info2:8088/cluster/app/%s" % appId
# print(yarn_url)
return yarn_url
# 如果没有生成application id，就进入日志页面直接查看
else:
print('没有生成applicationId,需要进入详情页查看日志信息')
loginfo_url = "%s/#/design/executingDetail/%s/exection/%s/logInfo" % (self.host, flow_id, job_id)
# print(loginfo_url)
return loginfo_url
except Exception:
return
#
# g = GetLog('6bac071d-5cc7-42bc-b64e-12b368ff7099', 'http://192.168.1.76:8515').get_log_url()
# print(g)
def get_dataflow_data(flow_name):
ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"],
MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"])
try:
sql = "select id, name,flow_type from merce_flow where name like '%s%%%%' order by create_time desc limit 1" % flow_name
flow_info = ms.ExecuQuery(sql.encode('utf-8'))
print("flow_id,flow_name,flow_type", flow_info[0]["id"],
flow_info[0]["name"], flow_info[0]["flow_type"])
except:
return
if "auto_api_test" in flow_name:
data = {
"configurations": {
"arguments": [],
"properties": [{
"name": "all.debug",
"value": "false",
"input": "false"
}, {
"name": "all.dataset-nullable",
"value": "false",
"input": "false"
}, {
"name": "all.optimized.enable",
"value": "true",
"input": "true"
}, {
"name": "all.lineage.enable",
"value": "false",
"input": "false"
}, {
"name": "all.debug-rows",
"value": "20",
"input": "20"
}, {
"name": "all.runtime.cluster-id",
"value": "cluster1",
"input": ["cluster1"]
}, {
"name": "dataflow.master",
"value": "yarn",
"input": "yarn"
}, {
"name": "dataflow.deploy-mode",
"value": "client",
"input": ["client", "cluster"]
}, {
"name": "dataflow.queue",
"value": "default",
"input": ["default"]
}, {
"name": "dataflow.num-executors",
"value": "2",
"input": "2"
}, {
"name": "dataflow.driver-memory",
"value": "512M",
"input": "512M"
}, {
"name": "dataflow.executor-memory",
"value": "1G",
"input": "1G"
}, {
"name": "dataflow.executor-cores",
"value": "2",
"input": "2"
}, {
"name": "dataflow.verbose",
"value": "true",
"input": "true"
}, {
"name": "dataflow.local-dirs",
"value": "",
"input": ""
}, {
"name": "dataflow.sink.concat-files",
"value": "true",
"input": "true"
}, {
"name": "dataflow.tempDirectory",
"value": "/tmp/dataflow/spark",
"input": "/tmp/dataflow/spark"
}],
"dependencies":
"",
"extraConfigurations": {},
"startTime":
1593964800000
},
"flowId": flow_info[0]["id"],
"flowName": flow_info[0]["name"],
"flowType": flow_info[0]["flow_type"],
"name": "auto_api_test_随机数",
"schedulerId": "once",
"source": "rhinos"
}
return data
elif "ES-upsert" in flow_name:
data = {
"configurations": {
"startTime":
1600617600000,
"arguments": [],
"dependencies": [],
"extraConfigurations": {},
"properties": [{
"name": "all.debug",
"value": "false",
"input": "false"
}, {
"name": "all.dataset-nullable",
"value": "false",
"input": "false"
}, {
"name": "all.optimized.enable",
"value": "true",
"input": "true"
}, {
"name": "all.lineage.enable",
"value": "false",
"input": "false"
}, {
"name": "all.debug-rows",
"value": "20",
"input": "20"
}, {
"name": "all.runtime.cluster-id",
"value": "cluster1",
"input": ["cluster1"]
}, {
"name": "dataflow.master",
"value": "yarn",
"input": "yarn"
}, {
"name": "dataflow.deploy-mode",
"value": "client",
"input": ["client", "cluster"]
}, {
"name": "dataflow.queue",
"value": "default",
"input": ["default"]
}, {
"name": "dataflow.num-executors",
"value": "2",
"input": "2"
}, {
"name": "dataflow.driver-memory",
"value": "512M",
"input": "512M"
}, {
"name": "dataflow.executor-memory",
"value": "1G",
"input": "1G"
}, {
"name": "dataflow.executor-cores",
"value": "2",
"input": "2"
}, {
"name": "dataflow.verbose",
"value": "true",
"input": "true"
}, {
"name": "dataflow.local-dirs",
"value": "",
"input": ""
}, {
"name": "dataflow.sink.concat-files",
"value": "true",
"input": "true"
}, {
"name": "dataflow.tempDirectory",
"value": "/tmp/dataflow/spark",
"input": "/tmp/dataflow/spark"
}],
"retry": {
"enable": "false",
"limit": 1,
"timeInterval": 1,
"intervalUnit": "MINUTES"
}
},
"schedulerId": "once",
"ource": "rhinos",
"version": 1,
"flowId": flow_info[0]["id"],
"flowType": flow_info[0]["flow_type"],
"name": "ES-upsert_随机数",
"creator": "admin",
"oldName": flow_info[0]["name"]
}
return data
else:
return
from util.encrypt import parameter_ungzip from basic_info.setting import MySQL_CONFIG from util.Open_DB import MYSQL # 配置数据库连接 ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"], MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"]) sql = "select links from merce_flow where name = 'minus_0628_3_no_element_test'" result = ms.ExecuQuery(sql) end_result = result[0]["links"] print(result[0]["links"]) print(parameter_ungzip(end_result))
class GetCheckoutDataSet(object):
"""
批量提交execution任务
任务执行结束后根据任务执行结果获取到对应的sink数据集输出结果
并对结果进行比对是否正确
"""
def __init__(self):
"""初始化数据库连接"""
self.ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"], MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"])
self.host = host
self.table = "flow_dataset_info.xlsx"
self.table_sheet = 'flow_info'
def get_flow_id(self):
"""
获取flow id并以list形式返回
:return: flow_id_list
"""##
print("------读取flow id list------")
flow_table = load_workbook(abs_dir(self.table))
info_sheet = flow_table.get_sheet_by_name(self.table_sheet)
max_row = info_sheet.max_row # 获取行数
flow_id_list = []
for i in range(2, max_row+1):
if info_sheet.cell(row=i, column=2).value and len(info_sheet.cell(row=i, column=2).value) > 10:
flow_id_list.append(info_sheet.cell(row=i, column=2).value)
flow_id_list = list(set(flow_id_list))
return flow_id_list
def get_flow_id_not_distinct(self):
"""
获取flow id并以list形式返回
:return: flow_id_list
"""
print("------读取flow id list------")
flow_table = load_workbook(abs_dir(self.table))
info_sheet = flow_table.get_sheet_by_name(self.table_sheet)
max_row = info_sheet.max_row # 获取行数
flow_id_list = []
for i in range(2, max_row+1):
if info_sheet.cell(row=i, column=2).value and len(info_sheet.cell(row=i, column=2).value) > 10:
flow_id_list.append(info_sheet.cell(row=i, column=2).value)
return flow_id_list
def data_for_create_scheduler(self):
"""
1. 根据flow_id 查找flow_name等信息
2. 根据查询到的flow信息，拼装创建scheduler所需要使用的data
:return: data_list
"""
print("------组装创建任务所需要的data------\n")
data_list = []
flow_id_list = self.get_flow_id()
for flow_id in flow_id_list:
try:
sql = 'select name, flow_type, parameters from merce_flow where id = "%s"' % flow_id
flow_info = self.ms.ExecuQuery(sql)
print('flow_info:', flow_info)
except Exception as e:
return e
else:
try:
flow_name = flow_info[0]["name"]
flow_type = flow_info[0]["flow_type"]
flow_parameters = flow_info[0]["parameters"]
arguments_list = []
arguments = {}
if flow_parameters: # flow parameters存在的情况下
print('parameters没有解压缩时：', flow_parameters)
parameters_use = parameter_ungzip(flow_parameters) # 将加密后的参数进行解密和解压缩处理
print('parameters解压缩后：', parameters_use)
flow_parameters_list = dict_res(parameters_use)
if len(flow_parameters_list) > 0:
arguments["name"] = flow_parameters_list[0]["name"]
arguments["category"] = flow_parameters_list[0]["category"]
arguments["value"] = flow_parameters_list[0]["defaultVal"]
arguments["refs"] = flow_parameters_list[0]["refs"]
arguments["description"] = flow_parameters_list[0]["description"]
arguments_list.append(arguments)
# print('arguments:', arguments)
except KeyError as e:
raise e
except IndexError as T:
raise T
data = {
"configurations": {
"arguments": arguments_list,
"properties": [
{
"name": "all.debug",
"value": "false"
},
{
"name": "all.dataset-nullable",
"value": "false"
},
{
"name": "all.lineage.enable",
"value": "true"
},
{
"name": "all.notify-output",
"value": "false"
},
{
"name": "all.debug-rows",
"value": "20"
},
{
"name": "dataflow.master",
"value": "yarn"
},
{
"name": "dataflow.deploy-mode",
"value": "client"
},
{
"name": "dataflow.queue",
"value": "merce.normal"
},
{
"name": "dataflow.num-executors",
"value": "2"
},
{
"name": "dataflow.driver-memory",
"value": "512M"
},
{
"name": "dataflow.executor-memory",
"value": "1G"
},
{
"name": "dataflow.executor-cores",
"value": "2"
},
{
"name": "dataflow.verbose",
"value": "true"
},
{
"name": "dataflow.local-dirs",
"value": ""
},
{
"name": "dataflow.sink.concat-files",
"value": "true"
}
],
"startTime": get_time()
},
"flowId": flow_id,
"flowName": flow_name,
"flowType": flow_type,
"name": flow_name + 'scheduler' + str(random.randint(0, 9999))+str(random.randint(0, 9999)),
"schedulerId": "once",
"source": "rhinos"
}
data_list.append(data)
print("------返回创建任务的data------")
# print(data_list)
return data_list
def create_new_scheduler(self):
"""
批量创建scheduler，
并返回scheduler_id_list， 供get_execution_info(self)调用
:return: scheduler_id_list
"""
print("------开始创建任务------")
from basic_info.url_info import create_scheduler_url
scheduler_id_list = []
scheduler_number = 1
for data in self.data_for_create_scheduler():
res = requests.post(url=create_scheduler_url, headers=get_headers(host), json=data)
print('第%d 个scheduler' % scheduler_number)
scheduler_number += 1
time.sleep(2)
# print(res.status_code, res.text)
if res.status_code == 201 and res.text:
scheduler_id_format = dict_res(res.text)
try:
scheduler_id = scheduler_id_format["id"]
except KeyError as e:
print("scheduler_id_format中存在异常%s" % e)
else:
scheduler_id_list.append(scheduler_id)
else:
print("flow: %s scheduler创建失败" % data["flowid"])
# return None
print("------create_new_scheduler(self)执行结束, 返回scheduler_id_list------\n")
print('scheduler_id_list', scheduler_id_list)
return scheduler_id_list
def get_execution_info(self):
"""
根据schedulers id 查询出execution id, name, status，flow_id,多个任务的信息以list的形式存储。
创建scheduler后查询execution有延迟，需要加等待时间
:return: e_info_list
"""
print("------查询execution信息------")
scheduler_id_list = self.create_new_scheduler()
if scheduler_id_list:
e_info_list = []
count = 1
for scheduler_id in scheduler_id_list:
print('开始第%d 次查询，查询scheduler id 为 %s 的execution info' % (count, scheduler_id))
# 等待30S后查询
time.sleep(20)
# 若没有查到execution id， 需要再次查询
e_info = self.get_e_finial_status(scheduler_id)
e_info_list.append(e_info)
print('e_info_list:', e_info_list)
count += 1
# print('查询得到的e_info_list', e_info_list)
print("------查询 execution信息结束------\n")
return e_info_list
else:
print("返回的scheduler_id_list为空", scheduler_id_list)
return None
def get_e_finial_status(self, scheduler_id):
"""
根据get_execution_info(self)返回的scheduler id,
查询该scheduler的execution id, name, status,flow id
:return: e_info
"""
print("------查询execution最终执行状态------")
if scheduler_id:
execution_sql = 'select id, status_type, flow_id , flow_scheduler_id from merce_flow_execution where flow_scheduler_id = "%s" ' % scheduler_id
time.sleep(10)
select_result = self.ms.ExecuQuery(execution_sql)
if select_result:
e_info = {}
# 从查询结果中取值
try:
e_info["e_id"] = select_result[0]["id"]
e_info["flow_id"] = select_result[0]["flow_id"]
e_info["flow_scheduler_id"] = select_result[0]["flow_scheduler_id"]
e_info["e_final_status"] = select_result[0]["status_type"]
except IndexError as e:
print("取值时报错 %s" % e)
raise e
print("------查询execution最终执行状态结束，返回execution id和status------\n")
return e_info
else:
return
else:
return
def get_execution_out_put(self):
"""
最终返回execution执行成功后的dataset id
:return: dataset id
"""
print("------开始查询execution的输出------")
e_info_list = self.get_execution_info()
print("返回的e_info: %s " % e_info_list)
# 数据无缺失，进行后续的判断
if len(e_info_list) == len(self.get_flow_id()): # flow id 数量相同
sink_dataset_list = []
for i in range(len(e_info_list)):
sink_dataset_dict = {}
try:
e_id = e_info_list[i]["e_id"]
e_final_status = e_info_list[i]["e_final_status"]
e_scheduler_id = e_info_list[i]["flow_scheduler_id"]
e_flow_id = e_info_list[i]["flow_id"]
except Exception as e:
print('请确认该flow的e_info_list:')
else:
sink_dataset_dict["flow_id"] = e_flow_id
sink_dataset_dict["execution_id"] = e_id
sink_dataset_dict["flow_scheduler_id"] = e_scheduler_id
if e_id:
print("------开始对%s 进行状态的判断------\n" % e_id)
while e_final_status in ("READY", "RUNNING"):
print("------进入while循环------\n")
# 状态为 ready 或者 RUNNING时，再次查询e_final_status
print("------查询前等待5S------\n")
time.sleep(5)
# 调用get_e_finial_status(e_scheduler_id)再次查询状态
e_info = self.get_e_finial_status(e_scheduler_id)
# 对e_final_status 重新赋值
e_final_status = e_info["e_final_status"]
print("------再次查询后的e_final_status: %s------\n" % e_final_status)
# time.sleep(50)
if e_final_status in ("FAILED", "KILLED"):
print("execution %s 执行失败" % e_id)
sink_dataset_dict["e_final_status"] = e_final_status
sink_dataset_dict["o_dataset"] = ""
sink_dataset_list.append(sink_dataset_dict)
# continue
elif e_final_status == "SUCCEEDED":
# 成功后查询flow_execution_output表中的dataset, 即sink对应的输出dataset，取出dataset id 并返回该ID，后续调用预览接口查看数据
sink_dataset_dict["e_final_status"] = e_final_status
# print(e_final_status, e_id)
data_json_sql = 'select b.dataset_json from merce_flow_execution as a LEFT JOIN merce_flow_execution_output as b on a.id = b.execution_id where a.id ="%s"' % e_id
print(data_json_sql)
data_json = self.ms.ExecuQuery(data_json_sql)
print("打印data_json:", data_json)
for n in range(len(data_json)):
sink_dataset = data_json[n]["dataset_json"] # 返回结果为元祖
print('-----sink_dataset-----', sink_dataset)
if sink_dataset:
sink_dataset_id = dict_res(sink_dataset)["id"] # 取出json串中的dataset id
sink_dataset_dict["dataset_id"] = sink_dataset_id
d = json.loads(json.dumps(sink_dataset_dict))
sink_dataset_list.append(d)
else:
continue
else:
print("返回的execution 执行状态不正确")
return
else:
print("execution不存在")
return
return sink_dataset_list
else:
print("返回的scheduler_id_list值缺失")
return
def get_json(self):
from util.count_items import count_items
"""
1.通过dataset预览接口取得数据的预览json串 result.text
2.判断dataset id是否存在，存在则通过数据集预览接口取回预览结果并找到表中相等的dataset id，写入预览结果
3.Nokia环境代码使用的是旧版的预览接口，0.8.11及以后高版本代码使用的是新预览接口
4.flow id 相等时，将execution id ,yarn上的log URL 和执行状态写入表中
5.对比结果
备注：对于sink multiouput flow特殊处理，表中把flow id全部写入，但是dataset是任务执行后再写入
:return:
"""
global dataset_id
print("------开始执行get_json()------")
sink_output_info = self.get_execution_out_put()
flow_table = load_workbook(abs_dir(self.table))
flow_sheet = flow_table.get_sheet_by_name(self.table_sheet)
sheet_rows = flow_sheet.max_row # 获取行数
print(sink_output_info)
flow_id_list = self.get_flow_id_not_distinct() # 获取flow表中未去重的所有flow id
sink_multi_flow_id = count_items(flow_id_list)[0] # 重复的flow id
sink_multi_flow_sink_num = count_items(flow_id_list)[1] # 重复的flow id 所重复的次数
# 处理multi_sink
if sink_multi_flow_id:
multi_flow_id = sink_multi_flow_id[0]
multi_flow_sink_num = sink_multi_flow_sink_num[0]
for j in range(2, sheet_rows + 1):
if sink_multi_flow_id == flow_sheet.cell(row=j, column=2).value:
for i in range(0, len(sink_output_info)):
log_url = GetLog(sink_output_info[i]["execution_id"], self.host).get_log_url()
dataset_id = sink_output_info[i]["dataset_id"]
flow_id = sink_output_info[i]["flow_id"]
if flow_id == sink_multi_flow_id:
if '57' in self.host:
priview_url = "%s/api/datasets/%s/preview?rows=5000&tenant=%s&rows=50" % (
self.host, dataset_id, get_tenant(self.host))
res = requests.get(url=priview_url, headers=get_headers(self.host))
result = res.text
else:
statementID = statementId_flow_use(self.host, dataset_id, tenant_id_83)
result = preview_result_flow_use(self.host, dataset_id, get_tenant(self.host),statementID)
flow_sheet.cell(row=j+i, column=4, value=dataset_id)
flow_sheet.cell(row=j+i, column=8, value=str(result))
for i in range(0, len(sink_output_info)):
##print('sink_output_info:',sink_output_info)
if "dataset_id" in sink_output_info[i].keys():#
dataset_id = sink_output_info[i]["dataset_id"]
flow_id = sink_output_info[i]["flow_id"]
if '57' in self.host:
priview_url = "%s/api/datasets/%s/preview?rows=5000&tenant=%s&rows=50" % (
self.host, dataset_id, get_tenant(self.host))
res = requests.get(url=priview_url, headers=get_headers(self.host))
result = res.text
else:
if "dataset_id" in sink_output_info[i].keys():#
statementID = statementId_flow_use(self.host, dataset_id, tenant_id_83)
result = preview_result_flow_use(self.host, dataset_id, get_tenant(self.host), statementID)
for j in range(2, sheet_rows + 1): # 按照行数进行循环
log_url = GetLog(sink_output_info[i]["execution_id"], self.host).get_log_url()
# 如果 dataset id相等就写入实际结果，不相等就向下找
if "dataset_id" in sink_output_info[i].keys():#
if dataset_id == flow_sheet.cell(row=j, column=4).value:#向左前进1格
flow_sheet.cell(row=j, column=8, value=str(result)) # dataset id 相等，实际结果写入表格 #向左前进1格
# flow id 相等时，将execution id ,yarn上的log URL 和执行状态写入
if flow_id == flow_sheet.cell(row=j, column=2).value:
# print(sink_dataset[i]["flow_id"])
flow_sheet.cell(row=j, column=5, value=sink_output_info[i]["execution_id"])
flow_sheet.cell(row=j, column=12, value=log_url)
flow_sheet.cell(row=j, column=6, value=sink_output_info[i]["e_final_status"])
else:
for t in range(j, 2, -1):
if flow_id == flow_sheet.cell(row=t - 1, column=2).value:
flow_sheet.cell(row=j, column=5, value=sink_output_info[i]["execution_id"])
flow_sheet.cell(row=j, column=12, value=log_url)
flow_sheet.cell(row=j, column=6, value=sink_output_info[i]["e_final_status"])
# flow_sheet.cell(row=j, column=8, value=result.text) # 实际结果写入表格
break
flow_table.save(abs_dir(self.table))
table = load_workbook(abs_dir(self.table))
table_sheet = table.get_sheet_by_name(self.table_sheet)
c_rows = table_sheet.max_row
print('-----开始对比结果----')
for i in range(2, c_rows+1):
table_sheet.cell(row=i, column=1, value=i-1)
# 对sample step涉及的flow单独进行结果判断
if table_sheet.cell(row=i, column=2).value == '0822a0a2-ce58-42cb-82de-f2ec434b5d94': #
if table_sheet.cell(row=i, column=6).value == 'SUCCEEDED' and table_sheet.cell(row=i, column=8):
new_result = []
expect_result = list(eval(table_sheet.cell(row=i, column=7).value))
actual_result = list(eval(table_sheet.cell(row=i, column=8).value))
for item in actual_result:
if item in expect_result:
new_result.append(item)
if new_result == actual_result:
table_sheet.cell(row=i, column=9, value="pass")
# print('test_result:', table_sheet.cell(row=i, column=9).value)
table_sheet.cell(row=i, column=10, value="")
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="execution: %s 预期结果实际结果不一致 " %
(table_sheet.cell(row=i, column=5).value))
elif table_sheet.cell(row=i, column=6).value == 'SUCCEEDED' and table_sheet.cell(row=i, column=8) == "":
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="execution: %s 预期结果实际结果不一致,实际结果为空 " %
(table_sheet.cell(row=i, column=5).value))
elif table_sheet.cell(row=i, column=6).value == 'FAILED':
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="execution: %s 执行状态为 %s" % (
table_sheet.cell(row=i, column=5).value, table_sheet.cell(row=i, column=6).value))
else:
print('请确认flow_id: %s的执行状态' % table_sheet.cell(row=i, column=2).value)
else:
# 判断mode为overwrite
if table_sheet.cell(row=i, column=11).value == 'overwrite': # 判断mode
# 实际结果存在并且执行结果为succeeded
if table_sheet.cell(row=i, column=8).value and table_sheet.cell(row=i, column=6).value == "SUCCEEDED":
# va7为预期结果，va8为实际结果，将二者先排序后对比是否相等
va7 = list(eval(table_sheet.cell(row=i, column=7).value))
#print('va7:',va7)
#if len(list(eval(table_sheet.cell(row=i, column=8).value))) != 0:
va8 = list(eval(table_sheet.cell(row=i, column=8).value)) # 注意：是不是需要放在if 语句后面？？？？
#print('va8',va8)
if va7 != [] and eval(table_sheet.cell(row=i, column=8).value).__class__ == [].__class__ :
va7_k = va7[0].keys()
va7_key = list(va7_k)
# 安排不同的key进行排序，只要有其中一个key排序后相等，就认为两个结果相等
result = []
for t in range(len(va7_key)):
S_va7 = sorted(va7, key=lambda item: item[va7_key[t]], reverse=True) # 没有 id时候的排序
S_va8 = sorted(va8, key=lambda item: item[va7_key[t]], reverse=True)
result.append(S_va7 == S_va8)
print('-----确认结果------')
if True in result:
table_sheet.cell(row=i, column=9, value="pass")
print('test_result:', table_sheet.cell(row=i, column=9).value)
table_sheet.cell(row=i, column=10, value="")
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="flowname: %s --->预期结果实际结果不一致 \n" %
(table_sheet.cell(row=i, column=3).value))
elif va7 == [] and va8 == []:
table_sheet.cell(row=i, column=9, value="pass")
# print('test_result:', table_sheet.cell(row=i, column=9).value)
table_sheet.cell(row=i, column=10, value="")
else:
table_sheet.cell(row=i, column=9, value="")
table_sheet.cell(row=i, column=10, value="请确认预期结果和实际结果")
elif table_sheet.cell(row=i, column=6).value == "FAILED":
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="flowname: %s --->执行状态为 %s\n" % (table_sheet.cell(row=i, column=3).value, table_sheet.cell(row=i, column=6).value))
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="用例参数或datasetID填写错误")
# 判断mode为append
elif table_sheet.cell(row=i, column=11).value == 'append':
if table_sheet.cell(row=i, column=8).value and table_sheet.cell(row=i, column=6).value == "SUCCEEDED": # 实际结果存在
expect_result_list = list(eval(table_sheet.cell(row=i, column=7).value))
expect_len = len(expect_result_list)
actual_result_list = list(eval(table_sheet.cell(row=i, column=8).value))
if expect_result_list == actual_result_list[-expect_len:]: # 实际结果切片和预期结果长度一致的数据结果是否相等
table_sheet.cell(row=i, column=9, value="pass")
table_sheet.cell(row=i, column=10, value="")
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10,
value="execution: %s 预期结果实际结果不一致 \n预期结果: %s\n实际结果: %s" % (
table_sheet.cell(row=i, column=5).value,
table_sheet.cell(row=i, column=7).value,
table_sheet.cell(row=i, column=8).value))
elif table_sheet.cell(row=i, column=6).value == "FAILED": # execution执行失败
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10,
value="execution: %s 执行状态为 %s" % (
table_sheet.cell(row=i, column=5).value, table_sheet.cell(row=i, column=6).value))
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="用例参数或datasetID填写错误")
else:
table_sheet.cell(row=i, column=9, value="fail")
table_sheet.cell(row=i, column=10, value="请确认flow的mode")
table.save(abs_dir("flow_dataset_info.xlsx"))
print('结果保存结束')
class GetExecutionStatus(object):
"""
该类用来获取批量创建的scheduler对应的execution，
并查询到execution的最终执行状态，写入到一个CSV文件中
"""
def __init__(self):
"""初始化数据库连接"""
self.ms = MYSQL(mysql_config["HOST"], mysql_config["USER"], mysql_config["PASSWORD"], mysql_config["DB"])
def exec_sql(self, scheduler_id):
"""
根据schedulers id 查询出execution id
创建scheduler后查询execution有延迟，需要加等待时间
"""
time.sleep(3)
# if scheduler:
execution_sql = "select id, status_type, flow_name, flow_id from merce_flow_execution where flow_scheduler_id = '%s'" % scheduler_id
try:
result = self.ms.ExecuQuery(execution_sql)
return result[0]
except Exception:
return
def get_execution_status(self):
"""该方法用来循环查询execution的最终执行状态，并将execution信息写入指定表中"""
# csv表的header: scheduler_id,execution_id,status,flow_name,flow_id
# 读取jmeter脚本创建scheduler后保存scheduler id的CSV文件，依次查询到其对应的execution信息，并写入一个新的file中
header = ('scheduler_id', 'execution_id', 'status', 'flow_name', 'flow_id', 'log_url')
print("------开始执行get_execution_status(self)------\n")
with open(scheduler_file_path, "r") as f:
readers = csv.reader(f)
headers = next(readers) # 抛去header行
executions = []
executions.append(header)
for scheduler_id in readers:
# print('scheduler_id:', scheduler_id)
e_info = self.exec_sql(scheduler_id[0])
print(e_info)
while not e_info:
time.sleep(5)
e_info = self.exec_sql(scheduler_id[0])
print(e_info)
if e_info:
# 使用元祖存储execution信息，方便后续向csv中写数据
print('sql第一次查询后返回的execution信息', e_info)
count = 0
while e_info["status_type"] in ("PREPARE", "READY", "RUNNING"):
print("------进入while循环------\n")
# 状态为 ready 或者 RUNNING时，再次查询e_final_status
print("------查询前等待5S------\n")
time.sleep(5)
# 调用get_e_finial_status(e_scheduler_id)再次查询状态
e_info = self.exec_sql(scheduler_id[0])
# count += 1
# if count == 100:
# break
print('再次查询后的execution"status_type"]', e_info)
# 获取execution的最终执行状态后，将信息存入元祖，并依次append到列表executions中
if e_info["status_type"] in ("SUCCEEDED", "FAILED", "KILLED"):
e_log_url = GetLog(e_info["id"]).get_log_url()
tp_execution = (scheduler_id[0],) + (e_info["id"],) + (e_info["status_type"],) + (
e_info["flow_name"],) + (e_info["flow_id"],) + (e_log_url,)
# print("tp_execution", tp_execution)
executions.append(tp_execution)
print("executions", executions)
# 依次从列表executions中读取元祖数据，并写入一个新的csv表中
file = get_result_file_name(file_path, HOST)
self.write_execution(file, executions)
print('write over')
@staticmethod
def write_execution(file, executions): # 向csv文件中写数据
with open(file, 'a', newline='') as f:
new_file = csv.writer(f, dialect='excel')
m = len(executions)
for i in range(m):
print(executions[i])
new_file.writerow(executions[i])
def get_dataflow_data(flow_name):
print("开始执行get_dataflow_data(flow_name)")
ms = MYSQL(MySQL_CONFIG["HOST"], MySQL_CONFIG["USER"],
MySQL_CONFIG["PASSWORD"], MySQL_CONFIG["DB"])
try:
sql = 'select id, flow_type from merce_flow where name = "%s"' % flow_name
flow_info = ms.ExecuQuery(sql)
print(sql)
print('flow_info:', flow_info)
except Exception as e:
raise e
else:
try:
flow_id = flow_info[0]["id"]
flow_type = flow_info[0]["flow_type"]
# print(flow_name, flow_type)
except KeyError as e:
raise e
data = {
"configurations": {
"arguments": [],
"properties": [{
"name": "all.debug",
"value": "false"
}, {
"name": "all.dataset-nullable",
"value": "false"
}, {
"name": "all.lineage.enable",
"value": "true"
}, {
"name": "all.notify-output",
"value": "false"
}, {
"name": "all.debug-rows",
"value": "20"
}, {
"name": "dataflow.master",
"value": "yarn"
}, {
"name": "dataflow.deploy-mode",
"value": "client"
}, {
"name": "dataflow.queue",
"value": "merce.normal"
}, {
"name": "dataflow.num-executors",
"value": "2"
}, {
"name": "dataflow.driver-memory",
"value": "512M"
}, {
"name": "dataflow.executor-memory",
"value": "1G"
}, {
"name": "dataflow.executor-cores",
"value": "2"
}, {
"name": "dataflow.verbose",
"value": "true"
}, {
"name": "dataflow.local-dirs",
"value": ""
}, {
"name": "dataflow.sink.concat-files",
"value": "true"
}],
"startTime":
get_time()
},
"flowId": flow_id,
"flowName": flow_name,
"flowType": flow_type,
"name": flow_name + str(random.randint(0, 99999)),
"schedulerId": "once",
"source": "rhinos"
}
return data
