SQL Server像MySQL一样拥有慢查询日志的操作方法(Slow Query Log慢日志)

CREATE EVENT SESSION [slowquerylog]
ON SERVER
    ADD EVENT sqlserver.blocked_process_report
    (ACTION
     (
         sqlserver.client_app_name,
         sqlserver.client_hostname,
         sqlserver.database_id,
         sqlserver.database_name,
         sqlserver.plan_handle,
         sqlserver.query_hash,
         sqlserver.request_id,
         sqlserver.session_id,
         sqlserver.sql_text
     )
    )
    ADD TARGET package0.event_file
    (SET filename = N'E:\DBExtentEvent\slowquerylog.xel')
WITH
(
    STARTUP_STATE = ON
);
GO

--窗口2
--locked process threshold是SQL Server2005推出的一个选项
--设置阻塞进程阈值
sp_configure 'show advanced options', 1 ;  
GO  
RECONFIGURE ;  
GO  
sp_configure 'blocked process threshold', 10 ;   --10秒
GO  
RECONFIGURE ;  
GO

--窗口3
USE testdb;
GO
BEGIN tran
update Account
set name ='Test'
where ID = 2
--commit

-- 窗口4
USE testdb;
GO
-- 这个查询会被窗口3中的事务阻塞
SELECT * FROM Account
WHERE ID = 2

-- 查询扩展事件 XEL 文件内容
SELECT 
    event_data.value('(event/@name)[1]', 'VARCHAR(50)') AS event_name,
    event_data.value('(event/@timestamp)[1]', 'DATETIME2') AS event_timestamp,
    event_data.value('(event/data[@name="duration"]/value)[1]', 'INT') AS duration,
    event_data.value('(event/action[@name="client_app_name"]/value)[1]', 'VARCHAR(255)') AS client_app_name,
    event_data.value('(event/action[@name="client_hostname"]/value)[1]', 'VARCHAR(255)') AS client_hostname,
    event_data.value('(event/action[@name="database_name"]/value)[1]', 'VARCHAR(255)') AS database_name,
    event_data.value('(event/action[@name="sql_text"]/value)[1]', 'VARCHAR(MAX)') AS sql_text
FROM 
    sys.fn_xe_file_target_read_file('E:\DBExtentEvent\slowquerylog*.xel', NULL, NULL, NULL) AS t
CROSS APPLY 
    t.event_data.nodes('event') AS XEvent(event_data);

import pyodbc
import pandas as pd
# 设置数据库连接
conn = pyodbc.connect(
    'DRIVER={SQL Server};'
    'SERVER=your_server_name;'
    'DATABASE=your_database_name;'
    'UID=your_username;'
    'PWD=your_password'
)
# 查询 XEL 文件内容
query = """
SELECT 
    event_data.value('(event/@name)[1]', 'VARCHAR(50)') AS event_name,
    event_data.value('(event/@timestamp)[1]', 'DATETIME2') AS event_timestamp,
    event_data.value('(event/data[@name="duration"]/value)[1]', 'INT') AS duration,
    event_data.value('(event/action[@name="client_app_name"]/value)[1]', 'VARCHAR(255)') AS client_app_name,
    event_data.value('(event/action[@name="client_hostname"]/value)[1]', 'VARCHAR(255)') AS client_hostname,
    event_data.value('(event/action[@name="database_name"]/value)[1]', 'VARCHAR(255)') AS database_name,
    event_data.value('(event/action[@name="sql_text"]/value)[1]', 'VARCHAR(MAX)') AS sql_text
FROM 
    sys.fn_xe_file_target_read_file('E:\DBExtentEvent\slowquerylog*.xel', NULL, NULL, NULL) AS t
CROSS APPLY 
    t.event_data.nodes('event') AS XEvent(event_data);
"""
# 使用 pandas 读取数据
df = pd.read_sql(query, conn)
# 关闭数据库连接
conn.close()
# 显示数据
print(df)
# 将数据保存为 CSV 文件
df.to_csv('slowquerylog.csv', index=False)

# 加载所需的程序集
Add-Type -Path "C:\Program Files\Microsoft SQL Server\140\SDK\Assemblies\Microsoft.SqlServer.XEvent.Linq.dll"
# 定义XEL文件路径
$xelFilePath = "E:\DBExtentEvent\slowquerylog*.xel"
# 创建XEventData对象
$events = New-Object Microsoft.SqlServer.XEvent.Linq.QueryableXEventData($xelFilePath)
# 初始化一个空数组来存储事件数据
$eventDataList = @()
# 遍历每个事件并提取所需的字段
foreach ($event in $events) {
    $eventData = New-Object PSObject -Property @{
        EventName      = $event.Name
        Timestamp      = $event.Timestamp
        Duration       = $event.Fields["duration"].Value
        ClientAppName  = $event.Actions["client_app_name"].Value
        ClientHostname = $event.Actions["client_hostname"].Value
        DatabaseName   = $event.Actions["database_name"].Value
        SqlText        = $event.Actions["sql_text"].Value
    }
    $eventDataList += $eventData
}
# 将事件数据导出为CSV文件
$eventDataList | Export-Csv -Path "E:\DBExtentEvent\slowquerylog.csv" -NoTypeInformation

import pandas as pd
# 定义CSV文件路径
csv_file_path = "E:\\DBExtentEvent\\slowquerylog.csv"
# 使用pandas读取CSV文件
df = pd.read_csv(csv_file_path)
# 显示数据
print(df)