Tag Archives: SQL Server

SQL Server database information

SQL Server backups to Alternate Data Streams or Colons

So in the midst of a very busy day I performed a manual backup of a database for a developer so he could make some major changes, test, and rollback if needed. I entered the backup file name with a colon on the timestamp…DUH! Of course, this worked, as NTFS supports the use of Colons in the filename. Went back later to restore the backup for the developer and the filename was truncated at the colon and was 0 bytes in size. WTF, the backup worked, there was no warning or error from SQL Server? Then I remembered my old NTFS “friend” – Alternate Data Streams. Basically ADS is a way to put data into different streams of the file. If you’ve ever wondered how Windows knows to warn you when you run an executable downloaded from the internet via IE, this is how. IE places a “zone.identifier” in the ADS to let Windows know this file might not be safe.

There’s a couple of ways to get around this and recover the backup regardless of the truncated filename and the size of 0. The quickest and easiet way is to just restore the database or log from TSQL. So if you backup a database with

BACKUP DATABASE test TO DISK='test_11:30.bak'

That will work fine. In your default backup directory, you’ll see a file “test_11” and it will be 0 bytes in size. If you then try to use the SSMS GUI to restore this, it will fail.

If you instead use TSQL…

RESTORE DATABASE test FROM DISK='test_11:30.bak'

it will work.

The colon tells the OS to create a file with aname of everything before the colon, and all the data into an ADS with an idetifier of everything after the colon… so in our example, test_11 has an ADS in it with an identity of :30.bak. The backup data is all there in that stream.

So now you’re saying “Well, what if I don’t know the stream identity”! If that’s the case there are a number of tools that can tell you all the ADS identities in a file… I use STREAMS from that Sysinternals genius, Mark Russinovich. If will spit out the ADS in the file you give it. I’ve also used notepad and a Windows port of the *nix “cat” command to pull the data out of that ADS and into a new file. That new file would then be able to be restored from with the SSMS GUI. With Notepad, just open a command prompt and type NOTEPAD test_11:30.bak and give it some time, and it will have all that data in Notepad. Save that as test.bak and you can restore anyway you want.

Denali Always On adventures

I’ve built a new AD controller VM and two SQL server VM’s with Windows Server 2008 R2 EE. Joined both SQL servers to the domain. Installed the Failover Cluster feature on each. Installed Denali RC0. Enabled AlwaysOn High Availability in the SQL Server Configuration Manager.

Then I created a FayWorks database, and a new Availability Group. I set up the primary and then a replica / read only / preferred backup. Then I set up a Availability Group listener. I connected with SSMS to the AG via the listener, started a script that inserted 5000 rows into a temp table in RBAR fashion. Initiated a failover. The insert failed at some point, but I was able to restart the insert without reconnecting or changing anything after the failover completed. Slick as a pan covered in bacon grease.

This combines the best of HA Failover Clustering with the best of Mirroring / Log Shipping / Etc. Being able to geographically set up an Availibility group, having the mirror be targeted for backups, reporting, etc, having up to 4 replicas, completing a failover of just an AG, creating a virtual instance name / IP. Oh yeah, Denali is a game changer.

Some links I have used

http://msdn.microsoft.com/en-us/library/hh213080(v=sql.110).aspx
http://msdn.microsoft.com/en-us/library/hh213417(v=sql.110).aspx

How to Set Up SQL Server 2012 Always On Availability Groups

There seem to be a lot of questions and even some misinformation popping up with regards to SQL Server 2012 licensing. Microsoft is moving away from licenses based on the number of processor in a server to figure out per/cpu licensing. They are now licensing hardware on a per/core license.

Right or wrong, agree or disagree, here’s the details as I know them. Based on a number of sources, including a meeting with my employers VAR and an internal MS licensing expert.

Per core licensing is based on “Core-Packs”. Each core-pack covers two cores, and there is a minimum purchase of two core-packs. This will be an expensive premium if one plans on building a single processor dual core machine, and you’re paying to license a minimum of four cores.

I’ve been told that the core-packs cost 50% less than current per/cpu licenses. That makes the magic number a total of 4 cores with 2012 licensing costing the same a single processor license with current license costs.

I’ve read that with 2012 licesning – in the case of virtualization if you license all the physical cores you have unlimited rights to virtual OS’s (vOS). Current day licensing with Enterprise Edition would only allow a total of 4 vOS’s per license.

Also, current customers with an EA will retain their current purchasing plan until the expiration of the EA, regardless of when that is.. if it’s Jan 1 of 2013, then all of 2012, you continue to purchase the licenses as you have. Once the EA is up, you will have to submit the number of cores in your currently licensed environment and MS will “trade” those for the equivalent number of Core-Packs. I’ve heard conflicting reports of a hard limit of 20 cores or 10 Core-Packs per server, and other reports that do not mention that limit.

Powershell scripts to document SQL Servers

Starting at a new company as their first really dedicated SQL Server DBA, I inherited over 20 production SQL Servers. These systems had varying levels of documentation from little to none, so I’ve had a painful time of it just getting up to speed on the current environment. I’ve read a number of articles from the masters, including Brent Ozar. I’ve used his SQL Blitz scripts, and his new sp_blitz procedure to gain knowledge of existing systems. Building on that, since I’m as lazy as any other good DBA, I’ve created – and by created, I really mean that I modified – some existing powershell scripts to gather information about servers or SQL Server and dump that information into Excel. I plan on running these scripts on a monthly basis to keep them current. I’ll add to them, modify them, and create new ones as I see fit. My real thoughts are that I can use this information in the event of a disaster, and rebuild a server from backups without having to memorize information such as logins, data file locations, sql server configuration options, etc.

This is the master post, and here are the links to each of the posts that I have created so far.

The original script which started all my work. Database properties
SQL Logins
SQL Server configuration settings
Database file settings
!!!Updated OS / Hardware level settings!!!
OS / hardware level settings

Powershell to capture OS / hardware level settings

This is part 4 of a multipart posting on some PowerShell scripts that I’ve found and made some modifications to. These scripts I have used to document my environment without having to manually type information in to the spreadsheets. The master which links to all the scripts is here.

— Update, I have updated this powershell script with a much easier to read and spits out a CSV that can be opened right in Excel without any manipulation. —

The new version of this script is now here

This script gathers information about the server that SQL Server is running on. This one is unique from the rest, as it connects not to SQL Server at all, but rather to WMI to run queries to capture things like the number of CPUs and Cores (think SQL Server 2012 licensing!), memory configuration, OS, IP’s and other network info. This script also expects a text file with a list of servers, I have a file called PhysicalServers.txt that has a list of hostnames. I redirect the output of this script to a csv file, and then import that data into a excel spreadsheet. This one has only a single worksheet, one row for each server in the text file.

The base of this script came from Jesse Hamrick and his awesome site of Powershell Knowledge. Update 5/20/2016 – it seems the site I originally referenced is compromised now with malware. I have removed the link to the website.

 #* Connect to computer
 


 $serverList = "PhysicalServers.txt"
 foreach ($server in (Get-Content $serverList)) 
 {   
    $strComputer = $server
 
    #  "SysInfo Information"
    $myOutput = $strComputer + ","
    $colItems = Get-WmiObject Win32_ComputerSystem -Namespace "root\CIMV2" `
     -ComputerName $strComputer
     

     
    foreach($objItem in $colItems) {
      $myOutput += [string]$objItem.Manufacturer + ","
      $myOutput += [string]$objItem.Model + ","
      $myOutput += [string]$objItem.TotalPhysicalMemory + " bytes,"
     }
     
     
    #  "System BIOS Information"
    $colItems = Get-WmiObject Win32_BIOS -Namespace "root\CIMV2" -computername $strComputer
     foreach($objItem in $colItems) {
      $myOutput += [string]$objItem.Description + ","
      $myOutput += [string]$objItem.SMBIOSBIOSVersion + "." + $objItem.SMBIOSMajorVersion + "." + $objItem.SMBIOSMinorVersion + ","
      $myOutput += [string]$objItem.SerialNumber + ","
     }
     
    #  "Operating System Information"
    $colItems = Get-WmiObject Win32_OperatingSystem -Namespace "root\CIMV2"`
     -Computername $strComputer
     
    foreach($objItem in $colItems) {
        $subOSName = $objItem.Name.IndexOf("|") -1
      $myOutput += $objItem.Name.Substring(0,$subOSName) + ","
      $myOutput += [string]$objItem.OSArchitecture + ","
     }
     
     # "Processor Information"
    $colItems = Get-WmiObject Win32_Processor -Namespace "root\CIMV2" `
     -Computername $strComputer
     
     $count=0
     $temp=""
     
    foreach($objItem in $colItems) {
      $count++
      $temp = $objItem.Name + "," + $count + "," + $objItem.NumberOfCores
     }
     
     $myOutput += [string]$temp + ","
    
    # "Memory Information"
    $colItems = Get-WmiObject Win32_PhysicalMemory -Namespace "root\CIMV2" `
     -ComputerName $strComputer 
     $iter=1
    foreach($objItem in $colItems) {
      $iter++
      $myOutput += [string]$objItem.Capacity + ","
      $myOutput += [string]$objItem.BankLabel + ","
     }
     
     # Fill out the remaining memory slots with blanks (in my environment, the biggest server I have has 12 slots, so that's the max)
     while($iter -lt 13)
     {
        $myOutput += [string]",,"
        $iter++
     }
         
    #  "Network Information"
    $colItems = Get-WmiObject Win32_NetworkAdapterConfiguration -Namespace "root\CIMV2" `
     -ComputerName $strComputer | where{$_.IPEnabled -eq "True"}
     
    foreach($objItem in $colItems) {
      $myOutput += [string]$objItem.DHCPEnabled + ","
      $myOutput += [string]$objItem.IPAddress + ","
      $myOutput += [string]$objItem.IPSubnet + ","
      $myOutput += [string]$objItem.DefaultIPGateway + ","
      $myOutput += [string]$objItem.MACAddress
     }
         
$myOutput    
    
}

Powershell to document Database file settings

This is part 3 of a multipart posting on some PowerShell scripts that I’ve found and made some modifications to. These scripts I have used to document my environment without having to manually type information in to the spreadsheets. The master which links to all the scripts is here.

This script gathers the data and log file settings for each database. It also expects a flat file named Servers.txt with a list of SQL Servers to connect to. It will connect via SSPI, so make sure you run it with a Windows account with the appropriate permissions on each SQL Server.

#change these settings to your environment
$Filename='DatabaseFiles'
#these are actually User ODBC DSNs DSNs. Better for Excel.

$serverList = "Servers.txt"

# constants.
$xlCenter=-4108
$xlTop=-4160
$xlOpenXMLWorkbook=[int]51
# and we put the queries in here
$SQL=@"
SELECT DB_NAME(mf.[database_id])AS [Database Name], 
        suser_sname(d.owner_sid) AS [Owner], 
		mf.STATE_DESC AS [Status],
       mf.[file_id] AS [File ID], 
       mf.name AS [File Name], 
       mf.physical_name AS [Physical Location], 
       mf.type_desc AS [File Type],
       CONVERT( bigint, mf.size/128.0) AS [Total Size in MB]
FROM sys.master_files mf
inner join sys.databases d
on d.database_id = mf.database_id
ORDER BY DB_NAME(mf.[database_id]);
"@

# before we do anything else, are we likely to be able to save the file?
# if the directory doesn't exist, then create it
#if (!(Test-Path -path "$DirectoryToSaveTo")) #create it if not existing
#  {
#  New-Item "$DirectoryToSaveTo" -type directory | out-null
#  }
$excel = New-Object -Com Excel.Application #open a new instance of Excel
$excel.Visible = $True #make it visible (for debugging more than anything)
$wb = $Excel.Workbooks.Add() #create a workbook
$currentWorksheet=1 #there are three open worksheets you can fill up
foreach ($server in (Get-Content $serverList)) 
	{ #only create the worksheet if necessary
	if ($currentWorksheet-lt 4) {$ws = $wb.Worksheets.Item($currentWorksheet)}
	else  {$ws = $wb.Worksheets.Add()} #add if it doesn't exist
	$currentWorksheet += 1 #keep a tally
	#if  ($server.version -eq 2000) {$SQL=$SQL2000} #get the right SQL Script
	#if  ($server.version -eq 2005) {$SQL=$SQL2005} 
	#if  ($server.version -eq 2008) {$SQL=$SQL2008}
	$currentName=$server.Replace("\","_") # and name the worksheet
	$ws.name=$currentName # so it appears in the tab 
	# note we create the query so that the user can run it to refresh it
    
    $myconnectionstring = "Provider=SQLOLEDB.1;Integrated Security=SSPI;Initial Catalog=master;Data Source=$server"  
    $adOpenStatic = 3 
    $adLockOptimistic = 3  
    
    $objConnection = New-Object -comobject ADODB.Connection 
    $objRecordset = New-Object -comobject ADODB.Recordset  
    
    $objRecordset.CursorLocation = 3
    
    $objConnection.Open($myconnectionstring) 
    $objRecordset.Open($SQL,$objConnection,$adOpenStatic,$adLockOptimistic)  
    
    $objRecordset.MoveFirst()
    
    $qt = $ws.QueryTables.Add($objRecordset,$ws.Range("A1"))
	# and execute it
	if ($qt.Refresh())#if the routine works OK
		{
	              $ws.Activate()
		$ws.Select()
		$excel.Rows.Item(1).HorizontalAlignment = $xlCenter
		$excel.Rows.Item(1).VerticalAlignment = $xlTop
		$excel.Rows.Item(1).Orientation = -90
		#$excel.Columns.Item("F:AH").NumberFormat = "[Red][=0]û;[Blue][=1]ü"
		#$excel.Columns.Item("F:AH").Font.Name = "Wingdings" 
		#$excel.Columns.Item("F:AH").Font.Size = 12
		$excel.Rows.Item("1:1").Font.Name = "Calibri"
		$excel.Rows.Item("1:1").Font.Size = 9
		$excel.Rows.Item("1:1").Font.Bold = $true
		$Excel.Columns.Item(1).Font.Bold = $true
        $Excel.Columns.Item("A:H").Font.Size = 9
        [void] $Excel.Columns.AutoFit()
		}

	}
$filename=$filename -replace  '[\\\/\:\.]',' ' #remove characters that can cause problems
$filename = "$filename.xlsx" #save it according to its title
if (test-path $filename ) { rm $filename } #delete the file if it already exists
$wb.SaveAs($filename,  $xlOpenXMLWorkbook) #save as an XML Workbook (xslx)
$wb.Saved = $True #flag it as being saved
$objRecordset.Close() 
$objConnection.Close() 
$wb.Close() #close the document
$Excel.Quit() #and the instance of Excel
$wb = $Null #set all variables that point to Excel objects to null
$ws = $Null #makes sure Excel deflates
$Excel=$Null #let the air out
# Hristo Deshev's Excel trick 'Pro Windows Powershell' p380
[GC]::Collect()

Powershell to document SQL Server configuration settings

This is part 2 of a multipart posting on some PowerShell scripts that I’ve found and made some modifications to. These scripts I have used to document my environment without having to manually type information in to the spreadsheets. The master which links to all the scripts is here.

This script gathers the information from sys.configurations. It also expects a flat file named Servers.txt with a list of SQL Servers to connect to. It will connect via SSPI, so make sure you run it with a Windows account with the appropriate permissions on each SQL Server.

#change these settings to your environment
$Filename='SQLServerProperties'
#these are actually User ODBC DSNs DSNs. Better for Excel.

$serverList = "Servers.txt"

# constants.
$xlCenter=-4108
$xlTop=-4160
$xlOpenXMLWorkbook=[int]51
# and we put the queries in here
$SQL=@"
SELECT name [Name], 
    value as [Configured Value], 
    value_in_use AS [Running Value], 
    minimum AS [MIN VALUE], 
    maximum AS [MAX VALUE], 
    CASE WHEN CAST(CAST(Value_in_Use as int) - CAST(value as int) as tinyint) = 0 THEN 1
	ELSE 0 
	END
	AS [RUNNING AS CONFIGURED],
    cast(is_dynamic as tinyint) AS [Dynamic], 
    cast(is_advanced as tinyint) AS [Advanced],
    description AS [Description],
    configuration_id
FROM    sys.configurations 
order by name;
"@


# before we do anything else, are we likely to be able to save the file?
# if the directory doesn't exist, then create it
#if (!(Test-Path -path "$DirectoryToSaveTo")) #create it if not existing
#  {
#  New-Item "$DirectoryToSaveTo" -type directory | out-null
#  }
$excel = New-Object -Com Excel.Application #open a new instance of Excel
$excel.Visible = $True #make it visible (for debugging more than anything)
$wb = $Excel.Workbooks.Add() #create a workbook
$currentWorksheet=1 #there are three open worksheets you can fill up
foreach ($server in (Get-Content $serverList)) 
	{ #only create the worksheet if necessary
	if ($currentWorksheet-lt 4) {$ws = $wb.Worksheets.Item($currentWorksheet)}
	else  {$ws = $wb.Worksheets.Add()} #add if it doesn't exist
	$currentWorksheet += 1 
	$currentName=$server.Replace("\","_") # and name the worksheet
	$ws.name=$currentName # so it appears in the tab 
	# note we create the query so that the user can run it to refresh it
    
    $myconnectionstring = "Provider=SQLOLEDB.1;Integrated Security=SSPI;Initial Catalog=master;Data Source=$server"  
    $adOpenStatic = 3 
    $adLockOptimistic = 3  
    
    $objConnection = New-Object -comobject ADODB.Connection 
    $objRecordset = New-Object -comobject ADODB.Recordset  
    
    $objRecordset.CursorLocation = 3
    
    $objConnection.Open($myconnectionstring) 
    $objRecordset.Open($SQL,$objConnection,$adOpenStatic,$adLockOptimistic)  
    
    $objRecordset.MoveFirst()
    
    $qt = $ws.QueryTables.Add($objRecordset,$ws.Range("A1"))
	# and execute it
	if ($qt.Refresh())#if the routine works OK
		{
	              $ws.Activate()
		$ws.Select()
		$excel.Rows.Item(1).HorizontalAlignment = $xlCenter
		$excel.Rows.Item(1).VerticalAlignment = $xlTop
		$excel.Rows.Item(1).Orientation = -90
		$excel.Columns.Item("F:H").NumberFormat = "[Red][=0]û;[Blue][=1]ü"
		$excel.Columns.Item("F:H").Font.Name = "Wingdings" 
		$excel.Columns.Item("F:H").Font.Size = 12
		$excel.Rows.Item("1:1").Font.Name = "Calibri"
		$excel.Rows.Item("1:1").Font.Size = 9
		$excel.Rows.Item("1:1").Font.Bold = $true
		$excel.Columns.Item(1).Font.Bold = $true
        $Excel.Columns.Item("A:I").Font.Size = 9
        [void] $Excel.Columns.AutoFit()
		}

	}
$filename=$filename -replace  '[\\\/\:\.]',' ' #remove characters that can cause problems
$filename = "$filename.xlsx" #save it according to its title
if (test-path $filename ) { rm $filename } #delete the file if it already exists
$wb.SaveAs($filename,  $xlOpenXMLWorkbook) #save as an XML Workbook (xslx)
$wb.Saved = $True #flag it as being saved
$objRecordset.Close() 
$objConnection.Close() 
$wb.Close() #close the document
$Excel.Quit() #and the instance of Excel
$wb = $Null #set all variables that point to Excel objects to null
$ws = $Null #makes sure Excel deflates
$Excel=$Null #let the air out
# Hristo Deshev's Excel trick 'Pro Windows Powershell' p380
[GC]::Collect()

Powershell to document SQL Logins

This is part 1 of a multipart posting on some PowerShell scripts that I’ve found and made some modifications to. These scripts I have used to document my environment without having to manually type information in to the spreadsheets. The master which links to all the scripts is here.

This script gathers Login information and membership in the built in roles. It also expects a flat file named Servers.txt with a list of SQL Servers to connect to. It will connect via SSPI, so make sure you run it with a Windows account with the appropriate permissions on each SQL Server.

#change these settings to your environment
$Filename='SQLLoginInformation'
#these are actually User ODBC DSNs DSNs. Better for Excel.

$serverList = "Servers.txt"

# constants.
$xlCenter=-4108
$xlTop=-4160
$xlUp=-4162
$xlSolid=1
$xlAutomatic=-4105
$xlOpenXMLWorkbook=[int]51
# and we put the queries in here
$SQL=@"
select 
	name AS [Login],
	dbname AS [Default Database],
	createdate AS [Create Date],
	updatedate  AS [Update Date],
	cast(hasaccess as tinyint) AS [Access],
	cast(denylogin as tinyint) AS [Deny Access],
	cast(isntname as tinyint) AS [Windows Name],
	cast(isntgroup as tinyint) AS [Windows Group],
	cast(isntuser as tinyint) AS [Windows User],
	cast(sysadmin as tinyint) AS [System Admin],
	cast(securityadmin as tinyint) AS [Security Admin],
	cast(serveradmin as tinyint) AS [Server Admin],
	cast(setupadmin as tinyint) AS [Setup Admin],
	cast(processadmin as tinyint) AS [Process Admin],
	cast(diskadmin as tinyint) AS [Disk Admin],
	cast(dbcreator as tinyint) AS [Database Creator],
	cast(bulkadmin as tinyint) AS [Bulk Admin], 
	totcpu AS [CPU Governed],
	totio AS [IO Governed],
	spacelimit AS [Space Governed],
	timelimit AS [Time Governed],
	resultlimit AS [Resultset Governed]
 from sys.syslogins
 order by name
"@


# before we do anything else, are we likely to be able to save the file?
# if the directory doesn't exist, then create it
#if (!(Test-Path -path "$DirectoryToSaveTo")) #create it if not existing
#  {
#  New-Item "$DirectoryToSaveTo" -type directory | out-null
#  }
$excel = New-Object -Com Excel.Application #open a new instance of Excel
$excel.Visible = $True #make it visible (for debugging more than anything)
$wb = $Excel.Workbooks.Add() #create a workbook
$currentWorksheet=1 #there are three open worksheets you can fill up
foreach ($server in (Get-Content $serverList)) 
	{ #only create the worksheet if necessary
	if ($currentWorksheet-lt 4) {$ws = $wb.Worksheets.Item($currentWorksheet)}
	else  {$ws = $wb.Worksheets.Add()} #add if it doesn't exist
	$currentWorksheet += 1 
	$currentName=$server.Replace("\","_") # and name the worksheet
	$ws.name=$currentName # so it appears in the tab 
	# note we create the query so that the user can run it to refresh it
    
    $myconnectionstring = "Provider=SQLOLEDB.1;Integrated Security=SSPI;Initial Catalog=master;Data Source=$server"  
    $adOpenStatic = 3 
    $adLockOptimistic = 3  
    
    $objConnection = New-Object -comobject ADODB.Connection 
    $objRecordset = New-Object -comobject ADODB.Recordset  
    
    $objRecordset.CursorLocation = 3
    
    $objConnection.Open($myconnectionstring) 
    $objRecordset.Open($SQL,$objConnection,$adOpenStatic,$adLockOptimistic)  
    
    $objRecordset.MoveFirst()
    
    $rows = $objRecordset.RecordCount+1
    
    $qt = $ws.QueryTables.Add($objRecordset,$ws.Range("A1"))
	# and execute it
	if ($qt.Refresh())#if the routine works OK
		{
	              $ws.Activate()
		$ws.Select()
		$excel.Rows.Item(1).HorizontalAlignment = $xlCenter
		$excel.Rows.Item(1).VerticalAlignment = $xlTop
		$excel.Rows.Item(1).Orientation = -90
		$excel.Columns.Item("E:I").NumberFormat = "[Red][=0]û;[Blue][=1]ü"
        $excel.Columns.Item("J:M").NumberFormat = "[Green][=0]û;[Red][=1]«"
        $excel.Columns.Item("N:Q").NumberFormat = "[Red][=0]û;[Blue][=1]ü"
        # Color the columns below to highlight the superuser permissions
        $excel.Range("J2","M"+$rows).Interior.Pattern=$xlSolid
        $excel.Range("J2","M"+$rows).Interior.PatternColorIndex=$xlAutomatic
        $excel.Range("J2","M"+$rows).Interior.Color=12320247
        $excel.Range("J2","M"+$rows).Interior.TintAndShade = 0
      
		$excel.Columns.Item("E:Q").Font.Name = "Wingdings" 
		$excel.Columns.Item("E:Q").Font.Size = 12
		$excel.Rows.Item("1:1").Font.Name = "Calibri"
		$excel.Rows.Item("1:1").Font.Size = 9
		$excel.Rows.Item("1:1").Font.Bold = $true
		$excel.Columns.Item(1).Font.Bold = $true
        $Excel.Columns.Item("A:Q").Font.Size = 9
        [void] $Excel.Columns.AutoFit()
		}

	}
$filename=$filename -replace  '[\\\/\:\.]',' ' #remove characters that can cause problems
$filename = "$filename.xlsx" #save it according to its title
if (test-path $filename ) { rm $filename } #delete the file if it already exists
$wb.SaveAs($filename,  $xlOpenXMLWorkbook) #save as an XML Workbook (xslx)
$wb.Saved = $True #flag it as being saved
$objRecordset.Close() 
$objConnection.Close() 
$wb.Close() #close the document
$Excel.Quit() #and the instance of Excel
$wb = $Null #set all variables that point to Excel objects to null
$ws = $Null #makes sure Excel deflates
$Excel=$Null #let the air out
# Hristo Deshev's Excel trick 'Pro Windows Powershell' p380
[GC]::Collect()

Time

How do I manage my time. My employer is not a large enterprise, but they are growing rapidly. The IT department is stretched thin. There’s over 800 “server” classed objects in Active Directory, I personally know of 40 SQL Server servers, 30 of which are “Production”. I probably have over 200 applications that are connected into these SQL Servers. And yet last week and this I found my time torn between trying to get our Monitoring System back operational for the entire infrastructure and pulled into security audits for File Shares and a single application that uses SQL Server to store file locations of engineering CAD drawings. Sure, at my company those files are our IP, our bread-and-butter, our “Family Jewels”, but at the same time we had a major outage for Sharepoint and a system that we sell to our customers. The Sharepoint outage was a SAN volume that filled up from outside into the LUN, and took the LUN offline. The other outage was a LUN running out of space due to a weekly index rebuild filling up a TLOG drive.

Both would have been easily preventable, if I had the time to be proactive and spend two hours on Friday reviewing system health reports. Instead I spent my time fixing SLA reports and internet web status Macros for our Monitoring solution, and then in meetings on how we have 78 logins that are Domain Administrators in our primary domain. Couple that with the report I typed up saying “Yes we can turn on successful and failed login auditing, and set up minimum password requirements for SQL Logins”, and you see why I didn’t have time to ensure the index rebuilds would blow up to take grow the TLOG to 300% of normal size. Just ignore the fact that the IT organization turns deaf, dumb, and mute when someone mentions an SLA, or DR, or BC.

Any suggestions other than “Dice, Monster, Update Resume”?