Tecumseh IT Consulting - Christopher Plunkett

Ever wonder if an "old" database is still being used? Here's how to find out in SQL

2011-10-26T20:15:00.000-04:00

Ever wonder if an "old" database is still being used? Here's how to find out in SQL 2005. This will tell you any active connections. In SharePoint I first spun up all known web apps and then ran this:

DECLARE @retTable TABLE (
SPID int not null
, Status varchar (255) not null
, Login varchar (255) not null
, HostName varchar (255) not null
, BlkBy varchar(10) not null
, DBName varchar (255) null
, Command varchar (255) not null
, CPUTime int not null
, DiskIO int not null
, LastBatch varchar (255) not null
, ProgramName varchar (255) null
, SPID2 int not null
, REQUESTID INT
)
INSERT INTO @retTable EXEC sp_who2
SELECT Status, Login, HostName, DBName, Command,
CPUTime, ProgramName, BlkBy AS [Last CMD Time] -- *
FROM @retTable
--WHERE Login not like 'sa%'
ORDER BY DBName;

Ever Wonder How Many Documents Are In Your SharePoint Document Libraries?

2010-12-17T12:46:00.002-05:00

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using Microsoft.SharePoint;
using Microsoft.SharePoint.Utilities;
using Microsoft.SharePoint.Administration;
using System.Configuration;

/* *********************************************************************************
PURPOSE: to answer the question "How many total documents do I have in the document libraries in my web application?"
Given a SharePoint web application, my code iterates through each site collection of the webapp, then each subweb of the site collection to get a collection of document library objects, and then counts the number of documents in each. It then pushes this information to a custom list on an adminsite, one per web app.

Christopher Plunkett
December 2010
chris.plunkett @ gmail.com
********************************************************************************
*/

namespace HowManyDocuments
{
   class CountDocuments
   {
   static void Main(string[] args)
   {
   // Read variables from app.config. You need to pass in the adminwebsite
   // that will store your custom list of information you want to capture
   // and also the Web Application in question
   string adminWebsite = ConfigurationSettings.AppSettings["adminwebsite"];
   string siteToCheck = ConfigurationSettings.AppSettings["sitetocheck"];

   // First, create the list on the adminsite
   CreateList(adminWebsite, siteToCheck);

   int number = 0;
   List spwebinfos = new List();

   try
   {
   // SPWebApplication.Sites --> SPSiteCollection
   SPWebApplication webApplication = SPWebApplication.Lookup(new Uri(siteToCheck));
   SPSiteCollection siteCollections = webApplication.Sites;

   // For each SPSite in SPSiteCollection...
   foreach (SPSite siteCollection in siteCollections)
   {
   SPWebsiteInfo spwebinfo = new SPWebsiteInfo() { Documents = new List() };
   spwebinfo.Documents = new List();

   // SPSite.AllWebs -->; SPWebCollection
   SPWebCollection collWebsite = siteCollection.AllWebs;

   // For each SPWeb's in SPWebCollection
   foreach (SPWeb subSite in collWebsite)
   {
   spwebinfo.WebsiteName = subSite.Name;

   // SPWeb.GetListsOfType --> SPListCollection
   SPListCollection docliblistcollection = subSite.GetListsOfType(SPBaseType.DocumentLibrary);
   // For each SPList in SPListCollection
   foreach (SPList list in docliblistcollection)
   {
   spwebinfo.Documents.Add(
   new DocLib()
   {
   DocumentCnt = list.ItemCount,
   Title = list.Title
   }
   );

   string adminSite = adminWebsite;
   string theTargetSite = siteToCheck;
   string parentSite = list.ParentWeb.ToString();
   string docLibName = list.Title.ToString();
   int numOfDocs = list.ItemCount;

   // Now go Update the custom list
   UpdateList(adminSite, theTargetSite, parentSite, docLibName, numOfDocs);

   // Writing to the console is redundant and unecessary. I originally
   // did it for debuging of course but decided to leave it
   Console.WriteLine("On site: " + parentSite);
   Console.WriteLine("Doc Lib Name: " + docLibName);
   Console.WriteLine("Number of Docs: " + numOfDocs);
   // Running totals...
   int tempNumber = list.ItemCount;
   number = number + tempNumber;
   Console.WriteLine("Total Docs in " + parentSite + "= " + number);
   }
   subSite.Close();
   }
   siteCollection.Close();
   spwebinfos.Add(spwebinfo);
   }
   Console.WriteLine("YOU CAN NOW LOOK AT YOUR DATA HERE: " + adminWebsite + "/Document Library Count List for " + siteToCheck);
   //Console.Read();
   }
   catch(System.Exception e)
   {
   Console.WriteLine("An Error has occurred! " + e.Message, e.InnerException, e.Source, e.StackTrace.ToString());
   Console.Read();
   }
   }

   private static bool listExists(SPListCollection collection, string title)
   {
   // Check to see if the list exists. If it does we will destroy it soon...
   foreach (SPList list in collection)
   {
   if (list.Title == title)
   {
   return true;
   }
   }
   return false;
   }

   static void CreateList(string adminWebSite, string siteToCheck)
   {
   // This creates the custom list in the given admin website
   SPSite site = new SPSite(adminWebSite);
   SPWeb web = site.OpenWeb();
   SPListCollection lists = web.Lists;

   if (listExists(web.Lists, "Document Library Count List for " + siteToCheck ))
   {
   // If the list exists, delete it
   SPList deleteThisList = web.Lists["Document Library Count List for " + siteToCheck ];
   deleteThisList.Delete();
   }

   {
   // Create new Generic list called "Document Library Count List for NAMEOFSITE"
   lists.Add("Document Library Count List for " + siteToCheck, "", SPListTemplateType.GenericList);

   SPList list = web.Lists["Document Library Count List for " + siteToCheck];

   // Rename default Title column and then create the other columns
   web.AllowUnsafeUpdates = true;
   // Here we rename Title column, which is created by default by SharePoint
   list.Fields["Title"].Title = "Subweb";
   list.Fields["Title"].Description = "From this website";
   list.Fields["Title"].Update();
   // Now add the other two
   list.Fields.Add("DocLib Name", SPFieldType.Text, true);
   list.Fields.Add("Number of Documents", SPFieldType.Integer, true);
   // Next we make new column visible in default view
   SPView view = list.DefaultView;
   view.ViewFields.Add("Number of Documents");
   view.ViewFields.Add("DocLib Name");
   // Finally, group by Title column (original name still applies)
   string viewQuery = "";
   view.Query = viewQuery;
   view.Update();
   web.AllowUnsafeUpdates = false;
   }
   web.Close();
   site.Close();
   }

   static void UpdateList(string adminWebsite, string theTargetSite, string parentSite, string docLibName, int numOfDocs)
   {
   // This function simply updates the admin SharePoint list with the information you are trying to capture
   SPSite site = new SPSite(adminWebsite);
   SPWeb web = site.OpenWeb();

   SPList theList = web.Lists["Document Library Count List for " + theTargetSite];
   SPListItem listItem = theList.Items.Add();

   listItem["Subweb"] = parentSite;
   listItem["Number of Documents"] = numOfDocs;
   listItem["DocLib Name"] = docLibName;
   listItem.Update();
   }
   }

   public class SPWebsiteInfo
   {
   public string WebsiteName { get; set; }
   public List Documents { get; set; }
   public int DocumentCnt
   {
   get
   {
   return this.Documents.Count;
   }
   }
   }

   public class DocLib
   {
   public string Title { get; set; }
   public int DocumentCnt { get; set; }

   public override string ToString()
   {
   return String.Format("{0}\t{1}", this.Title, this.DocumentCnt);
   }
   }
}

Enable Cascading Menus in SharePoint QuickLaunch

2010-04-07T15:03:00.001-04:00

Use SharePoint Designer and navigate to the root.
Drill-down to _catalogs/masterpage (Master Page Gallery)
Open up your .master page you want to edit

   Look for this section:


   id="QuickLaunchMenu"
   DataSourceId="QuickLaunchSiteMap"
   runat="server"
   Orientation="Vertical"
   StaticDisplayLevels="2"
   ItemWrap="true"
   MaximumDynamicDisplayLevels="0"
   StaticSubMenuIndent="0"
   SkipLinkText=""
   >

Change StaticDisplayLevels to "1" from "2"
Change MaximumDynamicDisplayLevels to "1" from "0"

To use this Master Page serverwide, after checking in and approving, copy to the 12 hive\TEMPLATE\FEATURES\PublishingLayouts\MasterPages

Using Scheduled Tasks to run a powershell script with a file parameter

2009-10-26T12:34:00.003-04:00

While having no problem running my scripts from a cmd line, I had a heck of a time getting my PowerShell script to work from Scheduled Tasks. My script takes a config filename as a parameter. Then, I discovered that if I push the current working path on the stack using pushd and pipe that to the command it works. Like so:

(Start in: C:\PowershellScripts)

Then, in the run field in Scheduled Tasks:

PowerShell -command "pushd C:\PowershellScripts | & C:\PowershellScripts\MOSSBackup.ps1 MOSSBackupConfigDEV.xml"

PowerShell script to fully back up a MOSS Farm (Supported)

2009-10-02T15:51:00.004-04:00

In a previous post, I outlined Microsoft's article that restoring a MOSS FARM is not supported unless one used either DPM or a manual method. I don't have access to DPM in my current position so I created the following PowerShell 1.0 script which serves to work-around the problem of being able to restore a MOSS Farm from backup - without DPM - stated in this whitepaper.

Overview:

1. Reads the XML config file to assign variables

2. Calls stsadm to backup top-level items, not including CA items.

3. Stops all MOSS services on all WFEs

4. Calls a SQL script to backup using osql the two CA databases

5. Zips up the 12 hive

6. Starts all MOSS services

7. Backups up all metabases

8. Emails the results

Please Note: you will need the PowerShell Community Extenstions for "write-zip" cmdlet to work correctly: http://www.codeplex.com/Pscx

Save this as *.ps1. I've posted the script first and a sample XML doc below the script-

MOSSBackup.ps1

# ************************************
# Author: Christopher Plunkett

# Email = chris.plunkett @ gmail.com

# ************************************

param([string]$configFileName)

if ( $configFileName.Length -eq 0 )
{
    $ErrorActionPreference = "Stop"
    Write-Host "Please supply the XML configuration file name..."
    Start-Sleep -Seconds 5
    exit
}
else
{
    [xml]$cfgFile = Get-Content .\$configFileName
}

#[xml]$cfgFile = Get-Content .\MOSSBackupConfig.xml

$BackupDestination = $cfgFile.CONFIG.BackupDestination
$CAServer = $cfgFile.CONFIG.CAServer
$DBServer = $cfgFile.CONFIG.DBServer
$MetabasePassword = $cfgFile.CONFIG.MetabasePassword
$smtpServer = $cfgFile.CONFIG.SMTPServer
$EmailFrom = $cfgFile.CONFIG.EMAILFrom
$Remote_OSQL_BackupScriptName = $cfgFile.CONFIG.Remote_OSQL_BackupScriptName

# ******************************************************************************
# First, use stsadm.exe to backup the items in the farm MINUS the CA databases.
# I manually created this list of items using "stsadm -o backup -showtree" and
# entered into the XML doc. The name of the items are obviously going to be
# different in your environment.
# ******************************************************************************

" ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::" >> "$BackupDestination\backuplog.txt"

foreach ( $i in $cfgFile.CONFIG.items.item )
{
    $stsbackupcmd = "stsadm.exe -o backup -directory " + ([char]34) + "$BackupDestination" + ([char]34) + " -backupmethod Full -item " + ([char]34) + $i + ([char]34)
    psexec \\$CAServer cmd /c $stsbackupcmd

    if ( $LASTEXITCODE -eq 0 )
    {
        [DateTime]::Now.ToString() + ": Backup completed of -item $i. Backupmethod was 'Full'" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + ": ERROR! Backup NOT completed of -item $i!" >> "$BackupDestination\backuplog.txt"
    }
}

# **********************************************************************************
# Stop all MOSS Services locally and remotely. This is REQUIRED as a pre-req to
# backing up your CA databases. Grabs all your farm's servers from the config xml.
# **********************************************************************************

foreach ( $s in $cfgFile.CONFIG.Servers.Server )
{
    foreach ( $ServiceName in $cfgFile.CONFIG.Services.Service )
    {
        $expression = "Stop-Service $ServiceName -Force"
        $commandBytes = [System.Text.Encoding]::Unicode.GetBytes( $expression )
        $encodedCommand = [Convert]::ToBase64String( $commandBytes )
        psexec \\$s cmd /c "echo . | powershell -EncodedCommand $encodedCommand"
    }

    if ( $LASTEXITCODE -eq 0 )
    {
        [DateTime]::Now.ToString() + ": Services were stopped successfully on $s" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + " >>>>ERROR! There was a problem stopping services on $s" >> "$BackupDestination\backuplog.txt"
    }
}

Start-Sleep -Seconds 15

# ***********************************************************************************************************************************************
# Next, call the wmi class Win32_Process on the SQL Server, invoking the Create method to
# spin up a remote shell, then pass it instructions for running the .sql backup script locally:
# ***********************************************************************************************************************************************
#( [WMIClass]"\\$DBServer\ROOT\CIMV2:Win32_Process" ).create( "cmd.exe /c " + "osql -S localhost -E -i $Remote_OSQL_BackupScriptName") | Out-Null

# ***********************************************************************************************************************************************
# Scratch that... it worked but I could not figure out how to make it wait for the
# command to finish before the script moved on, so I ended up using psexec.
# You need to have a sql script written on the remote SQL Server for this to work.
# ***********************************************************************************************************************************************

$osqlstatement = "osql -S localhost -E -i $Remote_OSQL_BackupScriptName"
psexec -i \\$DBServer cmd /c "echo . | $osqlstatement"

    if ( $LASTEXITCODE -eq 0 )
    {
        [DateTime]::Now.ToString() + ": Backup completed for CA Databases!" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + " >>>>ERROR! There was a problem backing up the CA Databases on $DBServer" >> "$BackupDestination\backuplog.txt"
    }

Start-Sleep -Seconds 10

# ***************************************************************************
# Backup the 12 hive on all servers. Again, you'll need the
# PowerShell Community Extenstions for "write-zip" cmdlet to work correctly:
# http://www.codeplex.com/Pscx
# ***************************************************************************

$12HivesExistence = Test-Path $BackupDestination\12Hives
if ( $12HivesExistence -match "False" )
{
    mkdir "$BackupDestination\12Hives"
}

foreach ( $s in $cfgFile.CONFIG.Servers.Server )
{
    $12Hive = Get-ChildItem "\\$s\c$\Program Files\Common Files\Microsoft Shared\web server extensions\12" -Exclude "LOGS"
    write-zip -level 9 -IncludeEmptyDirectories $12Hive "$BackupDestination\12Hives\$s.zip"

    if ( $? -match "True" )
    {
        [DateTime]::Now.ToString() + ": 12 Hive backed up successfully on $s" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + " >>>>ERROR!! There was a problem backing up the 12 Hive on $s" >> "$BackupDestination\backuplog.txt"
    }
}

# ****************************************************
# Now, start all services on all farm machines.
# ****************************************************

foreach ($s in $cfgFile.CONFIG.Servers.Server)
{
    foreach ( $ServiceName in $cfgFile.CONFIG.Services.Service )
    {
        $expression2 = "Start-Service $ServiceName"
        $commandBytes2 = [System.Text.Encoding]::Unicode.GetBytes($expression2)
        $encodedCommand2 = [Convert]::ToBase64String($commandBytes2)
        psexec \\$s cmd /c "echo . | powershell -EncodedCommand $encodedCommand2"
    }

    if ( $LASTEXITCODE -eq 0 )
    {
        [DateTime]::Now.ToString() + ": Services successfully started on $s" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + " >>>>ERROR! There was a problem starting services on $s" >> "$BackupDestination\backuplog.txt"
    }
}

Start-Sleep -Seconds 15

# *****************************************************
# Backup the metabases locally and remotely
# and save them to the Backup Destination
# *****************************************************

$MetabaseBackupsExistence = Test-Path "$BackupDestination\MetabaseBackups"
if ( $MetabaseBackupsExistence -match "False" )
{
    mkdir $BackupDestination\MetabaseBackups
}

foreach ($s in $cfgFile.CONFIG.Servers.Server)
{
    $expression = "cscript.exe C:\WINDOWS\system32\iisback.vbs /backup /e " + ([char]34) + $MetabasePassword + ([char]34) + " /b " + ([char]34) + $s + ([char]34) + " /overwrite"
    psexec \\$s cmd /c $expression

    Copy-Item "\\$s\C$\WINDOWS\system32\inetsrv\MetaBack\*.*" -Destination "$BackupDestination\MetabaseBackups\" -Recurse -Force

    if ( $LASTEXITCODE -eq 0 )
    {
        [DateTime]::Now.ToString() + ": Metabase successfully backed up on $s" >> "$BackupDestination\backuplog.txt"
    }
    else
    {
        [DateTime]::Now.ToString() + " >>>>ERROR! There was a problem backing up the metabase on $s" >> "$BackupDestination\backuplog.txt"
    }
}

# Kill metabase backup files older than five days:
Get-Childitem $BackupDestination\MetabaseBackups -recurse | ?{!$_.PSIsContainer -and ($_.LastWriteTime -lt (get-Date).adddays(-5))} | remove-item -force

# *************************************************************************
# Now email the log file!
# *************************************************************************

foreach ( $recipient in $cfgFile.CONFIG.EMAILTo.recipient )
{
    $backuplogfilename = “$BackupDestination\backuplog.txt”
    $msg = new-object Net.Mail.MailMessage
    $att = new-object Net.Mail.Attachment( $backuplogfilename )
    $smtp = new-object Net.Mail.SmtpClient( $smtpServer )
    $msg.From = $EmailFrom
    $msg.To.Add( $recipient )
    $msg.Subject = “Nightly MOSS Backsup Log File - $BackupDestination”
    $msg.Body = “The nightly MOSS Backup log file is attached from $BackupDestination”
    $msg.Attachments.Add( $att )
    $smtp.Send( $msg )
}

And here is a sample XML config file, which you will pass to the PS1 script above as a parameter.
MOSSBackupConfigDEV.xml

OK the formatting on this post doesn't want to take XML. Email me if you want it.

Perfmon Counters Can Not Be Trusted On A Guest VM

2009-09-29T23:29:00.003-04:00

This is a well written article on why perf counters can't be trusted in VMware virtual environments, and here is the same problem explained in Hyper-V.

VMware, being aware of this problem, has added additional perfmon counters for single vCPU (only) Windows guests (no Linux) in an ESX 4 environment. It's a step forward.

Disaster Recovery of a MOSS 2007 Farm

2009-09-09T13:38:00.007-04:00

Backup, Recovery, and Disaster Recovery of a MOSS 2007 Farm

Let's start out with an alarming and important notice from Microsoft:

"Important: Although the configuration database and Central Administration content database can be backed up, restoring backups of the configuration database and Central Administration content database taken from a running farm by using the tools built in to SharePoint Products and Technologies or SQL Server is not supported."

You can recover a farm, including the configuration database and Central Administration content database, in the following two ways:

1. You can use farm-level backups of a running farm taken with System Center Data Protection Manager to recover an entire farm, including the configuration database and Central Administration content database. For more information, see How to Recover a Windows SharePoint Services Farm (http://go.microsoft.com/fwlink/?LinkId=102831). This is not currently an option for me.

2. You can restore a backup of the configuration database and Central Administration content database taken from a fully stopped farm. For more information, see Move all databases to a new server (http://go.microsoft.com/fwlink/?LinkId=118325).

This Wiki serves to outline my experiences with number 2 above.

Backing up the Farm

Clone all virtual machines in the farm (minus the databse) if virtual, P2V if physical.
Make sure you have the following minimum rights:
- You must be a member of the Farm Administrators SharePoint group
- On the computer that is running the SharePoint Central Administration Web site, you must be a member of the Administrators group
- On the database server
  - Member of Adminstrators Group
  - Member of the db_backupoperator fixed database role

First you must use the built-in tools to backup the SSP and it's associated databases. Using SQL Backups are not supported in the SSP because of sync issues.
- stsadm.exe -o backup -showtree
  - This will show you what SSP's can be backed up.
- CA / Application Mgmt / Office SharePoint Server Shared Services / Create or configure this farm's shared services: Take note of SSP's and associated Web Applications
- stsadm.exe -o backup -directory \\\MOSSBackup -backupmethod Full -item portaltestssp
- Stop the following services:
  - Microsoft Single Sign-On service
    Office Document Conversions Launcher service
    Office Document Conversions Load Balancer service
    Office SharePoint Server Search service
    Windows SharePoint Services Administration service
    Windows SharePoint Services Search service
    Windows SharePoint Services Timer service
    Windows SharePoint Services Tracing service
    Windows SharePoint Services VSS Writer service
- iisreset /STOP
- Use SQL Mgmt Studio to then backup all the remaining databases by hand
Once databases are backed up, do the following:
- iisreset /START
- restart the services you previously stopped
- un-quiesce farm

Browsing local website in IIS causes a 401.1 Acces Denied error

2009-09-03T09:50:00.002-04:00

This issue occurs when the Web site uses Integrated Authentication and has a name that is mapped to the local loopback address.

You must either disable the loopback check or add the host names that are mapped to the loopback address; the latter is the preferred method.

Method 1: Specify host names
Note Microsoft recommends that you use this method.

To specify the host names that are mapped to the loopback address and can connect to Web sites on your computer, follow these steps:

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa\MSV1_0
Right-click MSV1_0, point to New, and then click Multi-String Value.
Type BackConnectionHostNames, and then press ENTER.
Right-click BackConnectionHostNames, and then click Modify.
In the Value data box, type the host name or the host names for the sites that are on the local computer, and then click OK.
Quit Regedit and restart the IISAdmin service.

Method 2: Disable the loopback check

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa
Right-click Lsa, point to New, and then click DWORD Value.
Type DisableLoopbackCheck, and then press ENTER.
Right-click DisableLoopbackCheck, and then click Modify.
In the Value data box, type 1, and then click OK.
Quit Regedit and then restart your computer.

MOSS 2007 - One content database per site collection HOW TO

2009-08-28T10:44:00.005-04:00

Did you know that you can create a site collection with one content database per site collection?

In MOSS 2007, it all starts with a web application. The web app is truly an IIS web app and provides the namespace "service" among many other things that aren't important to this discussion. It has an associated WSS_Content database but it's useless without at least one site collection.

So, we create a site collections because by default, no new site collections are created at the time of the the Web application creation.

This first site collection essentially takes the root ( / ) application space and becomes a new empty top-level site collection. The database for this site collection is still the *WSS_Content db we created a moment ago.

Now, in this single default site collection ( defined at either the path: / or /sites if you have not set up managed paths), we can have many subwebs. But if you run the command stsadm.exe -o enumsites -url you'll see Site Count="1" bc Microsoft calls Sites and site collections the same thing (in this case) and should not be confused with sub-sites or subwebs. (It's confusing I know)

For each new site collection (and of course subweb) we create, they all are associated with this single first site collection database.

So, instead of having up to 15,000 top-level sites/site collections per web application and its single associated content db, you can split them up per site collection.

This should ease restores and sort of organize more intuitively but also places administrative burden on the DBA group and the Admin group, not to mention top link navigation will have to be customized. Never-the-less, you can do it and here's how:

Using stsadm cmd line tool to create site collections

In this scenario, I'm creating site collections but creating new databases per collection rather than using the default that comes with the first web app. They are site collections instead and therefore can have their own database (sites/subwebs can't have their own databases while site collections can IF you do this; they normally do not)

stsadm -o createsiteinnewdb -url http://myinfodev -owneremail EMAIL -ownerlogin DOMAIN\cplunkett -ownername CPlunkett -sitetemplate sts#0 -ds DATABASESERVER -dn MOSS_MYINFODEV_ROOT_WSS_CONTENT

Before continuing you must define a managed path. In this example, /departments:
stsadm -o addpath -utl http://myinfodev/departments -type Wildcardinclusion

Now go ahead and create these new site collections under the myinfodev/departments namespace and enjoy as you have now one database per site collection!

stsadm -o createsiteinnewdb -url http://myinfodev/departments/EnterpriseApplications -owneremail -ownerlogin \cplunkett -ownername EMAIL DOMAIN\CPlunkett -ds DATABASESERVER -dn MOSS_MYINFODEV_slash_departments_slash_EntApp_WSS_CONTENT

stsadm -o createsiteinnewdb -url http://myinfodev/departments/EnterpriseArchitecture -owneremail -ownerlogin \cplunkett -ownername EMAIL DOMAIN\CPlunkett -ds DATABASESERVER -dn MOSS_MYINFODEV_slash_departments_slash_EntArch_WSS_CONTENT

- Then you should limit in CA each new db you just created for the site collection to one site (set "Maximum number of sites that can be created in this database" to 1)

How to ReSID the IIS 6.0 Metabase

2009-08-13T10:02:00.000-04:00

How to ReSID the IIS 6.0 Metabase

(Note: This process is not required for an IIS 7.0 server)

When a server is created via Ghost image or Virtual clone, the server must go through a reSID process by running NewSID or sysprep run on the server. If this is an IIS 6.0 server, in order for IIS to function properly, the IIS metabase must also be updated to reflect the new machine SID. This is done by either uninstalling and reinstalling IIS or running a utility called MetabaseACL on the server.

MetabaseACL can be used to identify the orphaned SID in the IIS metabase for the local IIS_WPG group. Then when the new SID for the local IIS_WPG group is retrieved, it can then be used by the MetabaseACL utility to ReSID the metabase.

Below is the process to ReSID the IIS 6.0 metabase.

1. Download MetabaseACL from http://www.bluewire-technologies.com/downloads/blog/MetabaseACL.zip then unzip the files to a local drive.
2. Download PsGetSID from http://download.sysinternals.com/Files/PsTools.zip then unzip the tool to a local drive.
3. Retrieve orphaned SID for local IIS_WPG group from IIS 6.0 metabase.
a. Open a command prompt and run the command metabaseacl “IIS://servername”
b. Look for the orphaned SID for the IIS_WPG group. This will be on the next line below where IIS_WPG appears in the output of the command. This orphaned SID will be referred to as “Old SID.”
4. Retrieve new SID for local IIS_WPG group from the server.
a. Open a command prompt and run the command PsGetSID \\servername IIS_WPG
b. This will show the SID for the local IIS_WPG group from the server. This new SID will be referred to as “New SID.”
5. Set New SID in the IIS 6.0 metabase.
a. Open a command prompt and run the following command metabaseacl IIS://servername fixup
b. Run the following command to verify that the process worked properly metabaseacl IIS://servername
c. In the output, verify that there are no orphaned SID’s.

Setting up a multicast NLB Cluster in a Virtual Environment

2009-08-05T10:03:00.002-04:00

Setting up a multicast NLB Cluster in a Virtual Environment

It should be noted here that this is only done once per cluster creation, not every time a host is added to the cluster.

The first step is to shut down the new vm, either before or after the IIS install, and add a second virtual nic card.

Create a new cluster by adding a new IP address and cluster name. Once this is done, go to the properties of the NLB Cluster and note the mac address. For example, the mac address might be 03-bf-ac-11-09-66

Next we need to log into the datacenter Cisco router to add a static arp entry to the IOS. This is required because multicasting is disabled by default for security. This is how to do it:

1. PuTTY to the router (example 172.17.1.1 is the Peak10 router)
2. login
user: admin
password in KeePass
3. type en
4. enter the enable password (find in KeePass by searching for 'Cisco')
5. type show run and page down to see the static arp entries. You should see the following:
arp ipaddress macaddress (must be in 48 bit octet format as shown below) ARPA
example: arp 172.17.9.x 03bf.ac11.0943 ARPA
6. type config t
7. add the IP of the new cluster with the new mac address exactly as shown:
arp 172.17.9.102 03bf.ac11.0966 ARPA
the last ARPA must be capitalized
8. now you can test your new cluster
9. if it works, type exit to get out of config mode
10. the last step is to save the current config from DRAM to NVRAM:
11. copy running-config startup-config
12. chose the default location

How to examine minidumps

2009-02-06T13:26:00.002-05:00

First, download and install the Windows debugging tools, visit the following Microsoft Web site:http://www.microsoft.com/whdc/devtools/debugging/default.mspx

Create a folder called "symbols" on C:

Go to cmd prompt

cd to C:\Program Files\Debugging Tools For Windows (x86)

windbg -y srv*c:\symbols*http://msdl.microsoft.com/download/symbols -z c:\windows\minidump\minidump.dmp

This command grabs the latest symbols and dumps them to C:\symbols and then opens the debugger where you can run various commands. See below.

From MS website:
Examine the dump fileThere are several commands that you can use to gather information in the dump file, including the following commands:

The !analyze -show command displays the Stop error code and its parameters. The Stop error code is also known as the bug check code.

The !analyze -v command displays verbose output.

The lm N T command lists the specified loaded modules. The output includes the status and the path of the module.Note

The !drivers extension command displays a list of all drivers that are loaded on the destination computer, together with summary information about their memory use. The !drivers extension is obsolete in Windows XP and later.

To display information about loaded drivers and other modules, use the lm command. The lm N T command displays information in a format that is similar to the old !drivers extension.For help with other commands and for complete command syntax, see the debugging tools Help documentation. The debugging tools Help documentation can be found in the following location:
C:\Program Files\Debugging Tools for Windows\Debugger.chmNote If you have symbol-related issues, use the Symchk utility to verify that the correct symbols are loaded correctly. For additional information about using Symchk, click the following article number to view the article in the Microsoft Knowledge Base:
311503 (http://support.microsoft.com/kb/311503/ ) Use the Microsoft Symbol Server to obtain debug symbol files
Simplify the commands by using a batch fileAfter you identify the command that you must have to load memory dumps, you can create a batch file to examine a dump file. For example, create a batch file and name it Dump.bat. Save it in the folder where the debugging tools are installed. Type the following text in the batch file:
cd "c:\program files\debugging tools for windows"kd -y srv*c:\symbols*http://msdl.microsoft.com/download/symbols -i c:\windows\i386 -z %1When you want to examine a dump file, type the following command to pass the dump file path to the batch file:
dump c:\windows\minidump\minidump.dmp

Overview of memory dump file options for Windows Server 2003, Windows XP, and Windows 2000

Linux Open-iSCSI and MPIO to a SAN

2008-11-26T23:48:00.001-05:00

Linux Open-iSCSI

"Open-iSCSI project is a high-performance, transport independent, multi-platform implementation of RFC3720."

Once you install the package, there is a command line tool called iscsiadm that one uses to manage your connections to the SAN. Open-iSCSI persistent configuration is implemented as a DBM database. Technically, this iscsiadm tool manages (updates, deletes, inserts, queries) the persistent database. The utility presents a set of operations that a user can perform on iSCSI nodes, sessions, connections, and discovery records.

The database contains two tables:
- Discovery table (/etc/iscsi/send_targets);
- Node table (/etc/iscsi/nodes).

The usual place for iSCSI database files: /etc/iscsi/nodes

These tables correspond to what they call the "operational mode" of iscsiadm. You will notice each command starts out with one or the other of the operational modes:

--mode specify operational mode op =

From the MAN page:

Open-iscsi does not use the term node as defined by the iSCSI RFC, where a node is a single iSCSI initiator or target. Open-iscsi uses the term node to refer to a portal on a target, so tools like iscsiadm require that --targetname and --portal argument be used when in node mode. For session mode, a session id (sid) is used. The sid of a session can be found by running iscsiadm -m session. The session id is not currently persistent and is partially determined by when the session is setup.

How to:

Install open-iscsi

aptitude install open-iscsi

Change IQN in /etc/initiatorname.iscsi (optional)

vi /etc/iscsi/initiatorname.iscsi

InitiatorName=

Restart open-iscsi daemon

/etc/init.d/open-iscsi restart

Discover Targets

iscsiadm -m discovery -t st -p :3260
iscsiadm -m discovery -t st -p :3260

example:

iscsiadm -m discovery -t st -p 192.168.11.1

iscsiadm -m discovery -t st -p 192.168.11.2

See what has been discovered

iscsiadm -m node

should see something like:

192.168.11.1:3260,1 iqn.2000-08.com.datacore:psan01-2
192.168.11.2:3260,1 iqn.2000-08.com.datacore:psan02-2

Check Configuration of the Connection

iscsiadm --mode node --targetname

example:

iscsiadm --mode node --targetname iqn.2000-08.com.datacore:psan02-2
* note the node.conn[0].startup property is set to manual

Change to Auto Login to Storage

iscsiadm -m node -T -p :3260 -o update -n node.conn[0].startup -v automatic
iscsiadm -m node -T -p :3260 -o update -n node.conn[0].startup -v automatic

example:

iscsiadm -m node -T iqn.2000-08.com.datacore:psan02-2 -p 192.168.11.2:3260 -o update -n node.conn[0].startup -v automatic

* note - if you run the command again from "Check Configuration of the Connection" section (above), you should now see node.conn[0].startup = automatic now

Set All Additional Discoveries to Auto-Login

vi /etc/iscsi/iscsid.conf

node.startup = Automatic

Check if Sessions Started

iscsiadm -m session

should see:

tcp: [1] 192.168.11.2:3260,1 iqn.2000-08.com.datacore:psan02-2
tcp: [2] 192.168.11.1:3260,1 iqn.2000-08.com.datacore:psan01-2

To see a more verbose output, add a -P 1

example:

iscsiadm -m session -P 1

DO NOT Reboot!!

Multi-Pathing ( MPIO )
Multipathing allows redundency so that if one of the paths goes out, we still have access to the SAN.

Aptitude install multipath-tools

Devices using the software initiators do not have a persistent naming scheme, and do not guarantee that a device (i.e. /dev/sdc) will always have the same device node. Persistent Naming describes mechanisms where the system identifies devices without relying on the /dev node, and provides a reference point for it that does not change at reboot.

First you have to comment out the 'Blacklist all devices' section in /etc/multipath.conf file
Note: the first time the file will not be in /etc. You must copy it from the examples directory as follows:
cp /usr/share/doc/multipath-tools/examples/multipath.conf.synthetic /etc/multipath.conf

Now edit /etc/multipath.conf
Make sure the blacklist section is commented out with #
It should look like this:

#blacklist {
# devnode "*"
#}

Restart the multipathd daemon
/etc/init.d/multipath-tools restart

Now check that the dev-mapper has configured the volume:
multipath -v2
multipath -ll
You should see something like the following. Take note of the UUID (shown in Red below):
root@pmon02:/dev/mapper# multipath -ll
DataCoreSAN (360030d90504d4f4e30325f56766f6c31) dm-0 DataCore,SANmelody
[size=150G][features=1 queue_if_no_path][hwhandler=0]
\_ round-robin 0 [prio=2][active]
\_ 3:0:0:0 sdb 8:16 [active][ready]

Now we must again edit the /etc/multipath.conf file:
You will see a section called "multipaths". You will need to change the UUID from the previous step and enter it into the wwid parameter. Enter a "friendly name" for your SAN with the alias paramenter (shown in blue below). Make sure all other parameters are set as below, esp the rr_min_io:

multipaths {
multipath {
wwid 360030d90504d4f4e30325f56766f6c31
alias DataCoreSAN
path_grouping_policy multibus
path_checker directio
path_selector "round-robin 0"
failback manual
rr_weight priorities
no_path_retry 5
rr_min_io 10
}

Save the file, then run:
multipath -v2
multipath -ll

root@pmon02:/dev/mapper# multipath -ll
DataCoreSAN (360030d90504d4f4e30325f56766f6c31) dm-0 DataCore,SANmelody
[size=150G][features=1 queue_if_no_path][hwhandler=0]
\_ round-robin 0 [prio=2][active]
\_ 4:0:0:0 sdc 8:32 [active][ready]
\_ 3:0:0:0 sdb 8:16 [active][ready]

Confirm that you now have persistent names to access the volume:
cd /dev/mapper
ls -l

root@pmon02:/dev/mapper# ls -l
total 0
crw-rw---- 1 root root 10, 63 2008-11-24 15:14 control
brw-rw---- 1 root disk 254, 0 2008-11-24 15:14 DataCoreSAN

Next we shall create a filesystem on the iSCSI volume:
mkfs.ext3 /dev/mapper/DataCoreSAN

Make sure the iSCSI filesystem mounts at Boot time:

In order to mount a filesystem that exists on an iSCSI Volume connected through the open-iSCSI Software initiator, you need to add a line to the /etc/fstab file. The format of this line is the same as any other device and filesystem with the exception being that you need to specify the _netdev mount option, and you want to have the last two numbers set to 0 (first is a dump parameter and the second is the fsck pass).

The _netdev option delays the mounting of the filesystem on the device listed until after the network has been started and also ensures that the filesystem is unmounted before stopping the network subsystem at shutdown.

example:

/dev/mapper/360030d90504d4f4e30325f56766f6c31 /mnt/DataCoreSAN ext3 _netdev,defaults 0 0

Now you must configure the multipath connections:
To create the multiple logins needed for Linux dev-mapper to work you need to create an ‘interface’ file for each GbE interface you wish to use to connect to the array.

First, get the MAC address: ifconfig eth2 grep HW (in my case eth2 is the configured ethernet port that connects to the SAN)
eth2 Link encap:Ethernet HWaddr 00:10:18:40:14:66
This is the MAC address. Next create a blank iface.eth2 file in /etc/iscsi/ifaces
touch /etc/iscsi/ifaces/iface.eth2
add the line with the MAC address you discovered above. Example:
iface.hwaddress = 00:10:18:40:14:66
Do this for each card on the iscsi LAN, if there is more than one.

Rediscover the targets:

iscsiadm –m discovery –t st –p :3260

You should see an entry for each interface you specified. In my case, it is only one.

Next, log into the volumes:

iscsiadm -m node -l -T iqn.2000-08.com.datacore:psan02-2 -p 192.168.11.2:3260
iscsiadm -m node -l -T iqn.2000-08.com.datacore:psan01-2 -p 192.168.11.1:3260
iscsiadm -m session
This last command will show that both adapters have connected to the array

Next, if the interfaces are on different networks (11 vs 12 like we have here) you have to remove the pointers to the interfaces so that it will work:
example:
rm /etc/iscsi/nodes/iqn.2000-08.com.datacore:psan01-2/192.168.11.1,3260,1/iface.eth3
rm /etc/iscsi/nodes/iqn.2000-08.com.datacore:psan02-2/192.168.12.1,3260,1/iface.eth2

Do exactly the same thing for /isci/send_targets

Reboot:

Shutdown -r now

Optional:

Edit /etc/iscsi/iscsi.conf with the following:

node.conn[0].timeo.noop_out_interval = 5
node.conn[0].timeo.noop_out_timeout = 10
note: If there are a lot of IO error messages, then the above values may be too
aggresive and you may need to increase the values for your network conditions
and workload, or you may need to check your network for possible problems.

2008-08-11T14:26:00.002-04:00

System State Backups on Domain Controllers: Behind the Scenes
Here is the process in a nutshell:

1. An AT job runs C:\WINDOWS\system32\sysstatebackup\sysstatebackup.cmd four times per day. This backs up the System State to local C:\SysStateBackups_LocalTemp, appending a date and time string for uniqueness.

SET computername=%computername%FOR /F "TOKENS=1* DELIMS= " %%A IN ('DATE/T') DO SET CDATE=%%BFOR /F "TOKENS=1,2 eol=/ DELIMS=/ " %%A IN ('DATE/T') DO SET mm=%%BFOR /F "TOKENS=1,2 DELIMS=/ eol=/" %%A IN ('echo %CDATE%') DO SET dd=%%BFOR /F "TOKENS=2,3 DELIMS=/ " %%A IN ('echo %CDATE%') DO SET yyyy=%%BSET date=%mm%%dd%%yyyy%SET hourstring=%time:~0,2%SET minutestring=%time:~3,2%SET secondsstring=%time:~6,2%SET thetime=%hourstring%%minutestring%%secondsstring%SET datetime=%date%%thetime%REM ECHO %datetime%C:\WINDOWS\system32\ntbackup.exe backup "@C:\Documents and Settings\ScheduledTasks\Local Settings\Application Data\Microsoft\Windows NT\NTBackup\data\%computername%_SYSTEMSTATEOnly_FULL.bks" /n "%computername%_SYSTEMSTATEOnly_FULL_%datetime%.bkf created %datetime%" /d "Set created %datetime%" /v:yes /r:yes /rs:no /hc:off /m normal /j "%computername%_SYSTEMSTATEOnly_FULL" /l:s /f "C:\SysStateBackups_LocalTemp\%computername%_SYSTEMSTATEOnly_FULL_%datetime%.bkf"

2. The next step in the batch file is to run File Replicator. File replicator will move these .BFK's from C:\SysStateBackups_LocalTemp a remote locations for backup. We used to move them to the NAS units but due to space limitations, the current locations are now \\pbackup01\NTBACKUP and \\pbackup02\NTBACKUP. We have double backups, so all DC SysState backups replicate to BOTH backup servers in the following job:
C:\WINDOWS\system32\FileReplicator\FileReplicator.exe /source "C:\SysStateBackups_LocalTemp" /destination "\\pbackup01\NTBACKUP" /destination "\\pbackup02\NTBACKUP" /archivedays "1"

3. Regarding the issue of clean-up on source and destination: the source is taken care of by Step #2, passing in the parameter /archivedays, while the destination needs to be cleaned up with a different method.

So, the final step is run automatically by at AT job on both Backup servers. This AT job is run once per day and is called DeleteAged. It simply calls Arthur Penn's Python script, which is located in C:\scripts\DeleteAged.PY. The first parameter tells it which folder to look in while the second tells it how many days of backups to keep.

RUN: python C:\scripts\DeleteAged.py F:\NTBACKUP 2
Start In: C:\Python25

*special note* All jobs were created by logging in locally as ScheduledTasks. Therefore any maintenance that needs to be done should be done logged in as this user, not your account

Allow Windows Clients in Active Directory Domain Groups to Access Samba Shares Using Likewise-open:

2008-06-24T14:47:00.008-04:00

Allow Windows Clients in Domain Groups to Access Samba Shares:

Pre-reqs: Likewise has been installed and the system already joined to your Active Directory Domain. See Mr. Penn's great post about this:

Next, we need to install Samba on the Linux system:

su

aptitude install samba winbind

Create a directory named 'idmap' under /usr/lib/samba. Create a symbolic link from /usr/lib/samba/idmap/lwicompat_v2.so to /usr/centeris/lib/idmap/lwicompat_v2.so. Do this for version 4 as well:

su
cd /usr/lib/samba
mkdir idmap
cd idmap
ln -s /usr/centeris/lib/idmap/lwicompat_v2.so
ln -s /usr/centeris/lib/idmap/lwicompat_v4.so

Next, confirm the version of Samba you have installed. I have Ubuntu Hardy Heron 8.04 and my version is Version 3.0.28a
smbd -V
Version 3.0.28a

Now we have to edit the Samba configuration file to set the following parameters. If the parameters are not included, add them.
security = ads
workgroup = enter workgroup from /etc/samba/lwiauthd.conf here
realm = enter realm from /etc/samba/lwiauthd.conf here
# use lwicompat_v2 for Samba 3.0.0 to 3.0.24
# idmap backend = lwicompat_v2
#
# use lwicompat_v4 for Samba 3.0.25 and later
idmap backend = lwicompat_v4
idmap uid = 50-9999999999
idmap gid = 50-9999999999
Create a userPrincipalName value for the machine in it's Active Directory account object using the command lwinet:
lwinet ads setmachineupn -U Administrator

should get:
Added userPrincipalName value of host/SAMBA@YOURDOMAIN.LOCAL

On the likewise site I found this special note:
Special note for Ubuntu and Debian only: Ubuntu and Debian store
secrets.tdb in /var/lib/samba so you will need to create a symlink back
to /etc/samba/secrets.tdb.

$ mv /var/lib/samba/secrets.tdb /var/lib/samba/secrets.tdb.orig
$ ln –s /etc/samba/secrets.tdb /var/lib/samba/secrets.tdb

To verify:
Start the Samba winbind daemon winbindd
/etc/init.d/winbind start

Use the wbinfo tool to verify various pieces of information. Start by
ensuring that winbindd is honoring the machine trust account
settings:
$ wbinfo -t
checking the trust secret via RPC calls succeeded

Next resolve a name to a SID and that SID to a uid
$ wbinfo -n "DOMAIN\user"
S-1-5-21-1866663475-11234234241-1563435624-1103 User (1)
$ wbinfo -S S-1-5-21-1862675975-116998903123-1366135624-1103
200000

This should match the information returned from getent which is
sent through Likewise Authentication daemon:
$ getent passwd "HCHB\cplunkett"
DOMAIN\:*:39854560:33456401::/home/DOMAIN/:/bin/bash

Next, start up Samba and after you have shared out a folder, try to access a share from smbclient or
a Windows client.
/etc/init.d/samba start

Go to a Windows machine:
\\{name of your linux server}\{name of share}
example: \\samba\test

You can verify the connected user’s using the smbstatus command.
$ smbstatus

Finally, see Mr. Penn's post on how to set up shares

Finding what is using all your diskspace in Linux (root partition fills up)

2008-05-27T15:34:00.000-04:00

Finding what is using all your diskspace in Linux (root partition fills up)

To find file of size greater than 100MB (= 1024*1024*100 bytes), use the following command (as root):

find / -xdev -size +104857600c

This will look for files >100MB in size on your root (/) partition without decending into other partitions.

Building 8.04 Hardy Heron from Scratch

2008-03-19T14:56:00.005-04:00

Building Ubuntu Server 8.04 (Hardy Heron) from Scratch on ESX 3.x Host

First, change root password
sudo passwd root

Next, install an improved text editor:

sudo aptitude install vim-gnome

Next, edit sources.list
sudo gvim /etc/apt/sources.list

Update the Ubuntu source list by uncommenting Canonical's partner repositories (four lines)

Update and Upgrade:
sudo aptitude update
sudo aptitude upgrade
sudo aptitude dist-upgrade

Install SSH to be able to Putty remotely:
sudo aptitude install openssh-server

sudo aptitude install ntp ntpdate

At the console (as root), permit root login via SSH if necessary:
vi /etc/ssh/sshd_config
change:
permitrootlogin = yes
:qw
/etc/init.d/sshd restart

NFS Server

sudo aptitude -P install nfs-kernel-server nfs-common portmap
sudo dpkg-reconfigure portmap (Do not bind to loopback!)
sudo /etc/init.d/portmap restart

Link the Installs folder under /var/data to the root:
cd / && ln -s /var/data/Installs Installs

To specify a folder you will need to edit /etc/exports, which can be done like (replace vi for your favorite editor):
sudo vi /etc/exports

For an example I will share, with full read/write and privileges, my temporary directory. This directory will be broadcasted to every local IP from 192.168.1.1 through 192.168.1.255.
/home/folder/tmp 192.168.1.1/24(rw,no_root_squash,async)

To share a read only drive for Installs, I did the following:
/Installs 172.0.0.0/8(ro,async,root_squash)

sudo /etc/init.d/nfs-kernel-server restart

After making changes to /etc/export you can make the shares effect by issuing the:
sudo exportfs -a

VMTOOLS
[Assume running as root so I will not display 'sudo' here]
aptitude install linux-headers-server
aptitude install make gcc
aptitude install build-essential linux-headers-`uname -r`

From VI Client, right click the machine, choose "Install VM Tools". This will mount a virtual CD.
mkdir /mnt/cdrom
mount /dev/cdrom /mnt/cdrom
cd /tmp
tar zxf /mnt/cdrom/VMwareTools-3.0.2-<xxxx>.tar.gz
umount /dev/cdrom
cd vmware-tools-distrib
./vmware-install.pl
exit

Adding new internal SCSI HD in Linux

2008-03-19T14:53:00.000-04:00

Adding new internal SCSI HD

fdisk -lfdisk /dev/sdb

choose 'n' for make new
mkfs.ext3 /dev/sdb1

You now need the UUID:
blkid /dev/sdb1

cp /etc/fstab /etc/fstab.backup
gvim /etc/fstab
example:
UUID=2fc204f-a3d22-23423-88fa-2342632432 /var/data ext3 defaults 0 0

Starting a stuck virtual machine in ESX 3

2008-01-25T14:12:00.000-05:00

Starting a stuck virtual machine

Problem: You have a virtual machine that when you power on in VI Client, it appears to power up with no erros and yet never starts. Typically no errors are reported. Restarting the ESX host will fix the problem- however- that is not always an option on production machines during the day. This problem might also manifest itself as "A file was not found" or "unable to access a file since it is locked"

Solution:

PuTTy to the ESX host as root
Check the VM state by typing the following command:
- vmware-cmd /vmfs/volumes/path-to-vm-directory/server.vmx getstate
Try to start the machine with the following command:
- vmware-cmd /vmfs/volumes/path-to-vm-directory/server.vmx start
- run the getstate command again (above); if getstate() = off, then go to the next step. If if's on, you are done.
There is an undocumented switch in the vmkfstools -D, that destroys (supposedly) any locks on vmdk's. It did not work but vmware support recommened trying it:
- vmkfstools -D PMON01.vmdk
- we also tried it on the flat file, to no avail: vmkfstools -D PMON01-flat.vmdk
We need to confirm that this ESX host has control of this VM if it's on a SAN. So, once this command is run, tail -f /var/log/vmkernel. Look for a long number that is the mac address of vmnic1
confirm this with ifconfig
You can try via Linux to kill the offending process but this also did not work:
- ps -ef | grep -i /vmfs/volumes/path-to-file/PMON01/PMON01.vmdk
- The second column is your pid of the vmkload_app of the Virtual Machine
- /usr/lib/vmware/bin/vmkload_app -k 9
- Check the vm state and try to start again. This didn't work for us. Moving on:
Now let's work through the kernel instead of Red Hat. So, you can check the vmid of the machine to kill by running vm-support -x (small x)
Kill the virtual machine (while generating core dumps and logs) by running: vm-support -X (capital X this time)
- This would have worked for us, however, it ran us out of space on the local ESX partition!
- To fix, we created an NFS share on our NAS unit and our intent was to point it to this instead. To do that we had to connect to the ESX host using the GUI VI Client, and Enable NFS CLIENT in the firewall rules.
- We had to clean up our dumps first to free up space by running: rm -rf vm-support-pesx01-2008-01-25--15.15.4008/
- cd /mnt
- mkdir /nfs-temp
- mount nas01.blah.local:/nfs-temp /mnt/nfs-temp
- cd /mnt/nfs-temp
- vm-support -X (this can run for hours, but for us it was about 15 min)
At this point, you can check to see if the lock is gone. Run the following command again: vmware-cmd /vmfs/volumes/path-to-vm-directory/server.vmx getstate
If your problem machine is NOT listed here, you will be able to successfully start the machine
If the machine is NOT listed, run: vmware-cmd /vmfs/volumes/path-to-vm-directory/server.vmx start
Now, the machine started for us, however, we lost the ability to manage the machine via Virtual Center. We could however, manage it directly using the VI client and conneting directly to the ESX host using root. To fix this problem you need to restart two daemons. The first is responsible for communicating directly to the VI client, the next is the daemon responsible for connecting ESX to Virtual Center.
- To restart the daemon that talks to the VI client directly: service mgmt-vmware restart
- To restart the daemon that talks to Virtual Center: service vmware-vpxa restart

"The Man" blocking your websurfing at work? You want to surf the web using a 100% secure channel? Here's the way around it that will always work:

2008-01-23T12:31:00.000-05:00

This technique is used to by-pass Websense at any onsite engagement or public internet cafe that is looking at and/or filtering outgoing http traffic. With this scenario I outline below, you are immune from prying eyes. You should be an advanced user to even attempt this but it's not hard to do. This is not only handy in a work environment but if you are in an Internet café or a public network and you want to surf all http traffic encrypted.

On my Ubuntu box at home, I set up the SSH daemon to listen on port 443 instead of 22 by editing /etc/ssh/sshd_config. I did this because traffic on port 443 is assumed always to be http SSL web traffic, which is encrypted. This way, a good admin with a sniffer would not be alarmed to see secure TCP conversations on that port but might wonder why port 22 is going outbound if you didn’t change it from the default.

Next, on my Vista laptop, I set up a PuTTY session to tunnel to my home Ubuntu box; you set the port number you want PuTTY to listen on as Dynamic. I used 8080 but you can use anything really. So, it looks like this on the tunnel tab: D8080. You don't have to specify a target IP or host like you normally would. This basically sets up a secure SOCKS proxy tunnel.

Laptop at Home:
if you do a netstat -an, it shows that now PuTTY process has opened a secure TCP channel on port 443 from my work laptop to home. But so what? Nothing is happening yet other than a secure socket has been established.

Separately, it also shows PuTTY is listening on localhost 8080, ready to fwd SOCKS traffic over this SSH tunnel.

So then you config IE to use proxy localhost:8080

IE fires up, and sends all http requests to the PuTTY process on localhost:8080 which promptly forwards it over SSH protocol on the SSL port (so sneaky).

To test all you need to do is go to whatismyip.org or a similar site and it should show your HOME IP as opposed to your work/public network IP now.

Surf a few webpages and now netstat -a looks completely different!