Question

我写了一段简短的python代码来捕获一些运行时统计信息，这些信息可以帮助我们评估性能。它使用一个csv查询，运行它们，并捕获执行时间。我想捕获我运行的每个查询的查询概要文件，而不仅仅是执行时间，然后将它们转储到日志或表中。这可能吗？

import pandas as pd 
import pymysql
import time
import numpy as np

connec = pymysql.connect(host="xxxx",user="xxxx",password="xxxx",database="xxxx")

csv_location = r'/Users/xxx/Downloads/test_data.csv'



def run_query(q,conn):
    #time.sleep(1)
    print ('running query: ', q)
    start_time = time.time()
    pd.read_sql(q,conn)
    t = (time.time() - start_time)
    return t

# capture queries to run
df_queries = pd.read_csv(csv_location,names=['query'])

# multiple connections possible using if required
df_queries["time1"] = np.vectorize(run_query)(df_queries["query"],connec)
df_queries["time2"] = np.vectorize(run_query)(df_queries["query"],connec)
df_queries["time3"] = np.vectorize(run_query)(df_queries["query"],connec)
df_queries["time4"] = np.vectorize(run_query)(df_queries["query"],connec)
df_queries["time5"] = np.vectorize(run_query)(df_queries["query"],connec)

connec.close()

# stats
df_queries['median'] = df_queries.median(axis=1)
df_queries['percentile_90th'] = df_queries.quantile(axis=1,q=0.9)
df_queries['stdev'] = df_queries.std(axis=1)
df_queries['max'] = df_queries.max(axis=1)

# dump results to csv
df_queries.to_csv('test.csv')

Python代码可在运行时提取MySQL查询配置文件

0 个答案: