我编写了一个PowerShell脚本,它遍历大量的IIS W3C日志文件并将值插入到MSSQL数据库中。
Set-Variable -Name "UnprocessedDir" -Value "X:\files" -Description "Folder for unprocessed log files" Scope Script
Set-Variable -Name "InputObject" -Value (New-Object -comObject MSUtil.LogQuery.IISW3CInputFormat) -Description "Log Parser input COM object" -Scope Script
Set-Variable -Name "OutputObject" -Value (New-Object -comObject MSUtil.LogQuery.SQLOutputFormat) -Description "Log Parser output COM object" -Scope Script
$OutputObject.clearTable = $false
$OutputObject.createTable = $false
$OutputObject.database = "Database_Name"
$OutputObject.driver = "SQL Server"
$OutputObject.dsn = "DSN_Name"
$OutputObject.fixColNames = $true
$OutputObject.ignoreIdCols = $true
$OutputObject.ignoreMinWarns = $true
$OutputObject.maxStrFieldLen = 511
$OutputObject.oConnString = $null
$OutputObject.password = $null
$OutputObject.server = "sqlserver.domain.com\INSTANCENAME"
$OutputObject.transactionRowCount = 5000
$OutputObject.username = $null
Set-Variable -Name "IISLogs" -Value @(Get-ChildItem -Path $UnprocessedDir -Recurse -File) -Description "Array of files to be imported into SQL" -Scope Script
Set-Variable -Name "LPComObj" -Value (New-Object -com MSUtil.LogQuery) -Description "COM Object used to import Log Parser records into MSSQL" -Scope Script
Write-Output "$(Get-ISOTimeStamp) Beginning SQL import. $($IISLogs.Count) Files to be imported"
$IISLogs | ForEach-Object { $loop = 0 } {
Set-Variable -Name "SubDir" -Value $(($_.FullName).Split('\')[-2]) -Description "Subdirectory where log file is located" -Scope Script
Set-Variable -Name "LogType" -Value $(($_.FullName).Split('\')[-3]) -Description "Type of log being imported" -Scope Script
Set-Variable -Name "ServerName" -Value $(($_.FullName).Split('\')[-4]) -Description "ServerName of file being imported" -Scope Script
Set-Variable -Name "LPQuery" -Description "Query to use in Log Parser" -Scope Script -Value @"
SELECT
-- FIELDS LogFilename,LogRow,date,time,c-ip,cs-username,s-sitename,s-computername,s-ip,s-port,cs-method,cs-uri-stem,cs-uri-query,sc-status,sc-substatus,sc-win32-status,sc-bytes,cs-bytes,time-taken,cs-version,cs-host,cs(User-Agent),cs(Cookie),cs(Referer),s-event,s-process-type,s-user-time,s-kernel-time,s-page-faults,s-total-procs,s-active-procs,s-stopped-procs
-- STANDARD FIELDS date,time,s-ip,cs-method,cs-uri-stem,cs-uri-query,s-port,cs-username,c-ip,cs(User-Agent),cs(Referer),sc-status,sc-substatus,sc-win32-status,time-taken
'$($ServerName)' as [servername],
'$($_.Name)' as [filename],
LogRow AS [row],
'$($LogType)' as [logtype],
TO_TIMESTAMP(date,time) AS [timestamp],
[s-ip],
[cs-method],
[cs-uri-stem],
[cs-uri-query],
TO_INT([s-port]),
[cs-username],
[c-ip],
[cs(User-Agent)],
[cs(Referer)],
TO_INT([sc-status]),
TO_INT([sc-substatus]),
TO_INT([sc-win32-status]),
TO_INT([time-taken]),
0 AS lock
INTO IIS_W3C
FROM '$($_.FullName)'
"@
Set-Variable -Name "LPResult" -Value ($LPComObj.ExecuteBatch($LPQuery, $InputObject, $OutputObject)) -Description "IIS Log File imported into SQL" -Scope Script
If ($LPResult -eq $false)
{
Write-Output "$(Get-ISOTimeStamp) Data imported from `"$($_.FullName)`""
Set-Variable -Name "loop" -Value ($loop + 1) -Description "Increase loop iteration Count" -Scope Script
}
Else
{
Write-Output "$(Get-ISOTimeStamp) Log Parser returned errors importing `"$($_.FullName)`""
Throw "$(Get-ISOTimeStamp) Log Parser returned errors importing `"$($_.FullName)`""
}
}
我导入的日志数量是数万;上面的代码非常适用于几百个文件,但几个小时后,它就会崩溃。从我所知,看起来ForEach-Object
循环的每次迭代都会创建一个新的SQL TCP连接,该连接在循环结束时不会终止。
我尝试在循环内外创建$LPComObj
。我试过了Remove-Variable
。我尝试了一些通用命令,例如$LPComObj.Close()
,.Remove()
,.Quit()
等。MSUtil.LogQuery
方法本身似乎不包含任何关闭SQL TCP的方法连接,并且在脚本运行时,我可以看到越来越多的TCP连接堆积起来。我尝试使用[System.Runtime.Interopservices.Marshal]::
释放/删除COM对象,但没有一个关闭TCP连接。即使关闭PowerShell会话也不会终止连接。
我能做到这一点的唯一方法就是找到dllhost.exe
"正在使用端口并将其终止的进程。但是从脚本中,没有一种干净的方法来获取有问题的dllhost.exe
进程的PID。 (尝试修改Get-Process | Stop-Process
的某些变体可能会有效,但会在执行脚本时添加很多时间。)
我可以通过哪些其他方式解决此问题?