Create your VMware test lab using AutoLab

AutoLab is a way to bring up a VMware environment through scripted installs to create test/lab environments. Read why Lauren Malhoit calls this tool awesome. 

It's my opinion that every person who works with VMware should have a test lab. Although it would be ideal to have the same hardware as your production lab and completely duplicate the setup, that's often not feasible. There are also things that come along with a test lab that make them annoying, for lack of a better word. Either you messed something up during testing and need to rebuild, or you'd like to test a different setup, which means you need to start completely over. This is where AutoLab comes in.

vExpert Alastair Cooke is the man behind this solution, with some help from Damian Karlsonand Veeam, a virtual backup and recovery company. As of this writing, AutoLab version 1.1a is out. In general, AutoLab automates your lab setup. If you have access to VMware Workstation,VMware Player, or ESXi servers/clusters, then you can use AutoLab for free. I have mine running on VMware Workstation, which is running on a laptop with 16 GB of RAM. It is recommended you have at least 8 GB of RAM on whichever hardware you use. You can bring up different versions of vSphere (4.0 and up) as well. This is nice if you want to check out a new version, or if you need to test something on an older version.

A brief synopsis of the setup

There is a deployment guide that tells you exactly how to run this setup, but just to give you an idea, you download an .OVA file that contains several VMs (including ESXi hosts and a vCenter). You deploy the .OVA and that brings the VMs up in Workstation (or Player, ESXi, etc.), and they're all powered down. You then power on the NAS VM, which is just a FreeNAS solution. Once that VM is powered on, there is a build share you need to populate with several VMware and Microsoft ISOs. Once that share is fully populated according to the documentation, you can begin an automated build of the lab. It starts with your domain controller (the DC VM). All you need to do is power it on, and AutoLab takes care of the rest. Once this is done, there is a Validate script that makes sure all scripts and installs ran properly before you move on to the next step (Figure A). You then move on to scripted installs of the vCenter and hosts. Figure A

Click the image to enlarge it.

To be clear, AutoLab doesn't necessarily take less time to build a test lab, but it takes a lot less effort and babysitting. You can literally start the VM and then leave it alone for an hour to work on your production environment or read a book. All of the install information, networking, storage, etc. is taken care of for you.

In the newer versions of AutoLab, there are scripted installs for VMware Horizon View andVMware vCloud Director. This is perfect for testing these things out. Also, when something breaks or you need a new proof-of-concept, you just restart the build. In addition, you don't need to repopulate the build share if you decide to rebuild; so, once that's done the first time, you're all set. Veeam has two VMs in the .OVA file that let you test its Veeam ONE and Veeam Backup & Replication products.

I mentioned using VMware Workstation, Player, or ESXi before to install AutoLab; however, these still require that you have some hardware behind it all. A company called baremetalcloudis offering an AutoLab solution for those who don't have any hardware handy. You pay baremetalcloud monthly on a subscription basis according to the compute resources you need. A really nice thing about what baremetalcloud has done is they use Clonezilla to manage the images; this helps with the time to provision. What takes AutoLab around three hours to provision on your own hardware, takes only about a half hour using baremetalcloud. Mike Laverick has a great write-up about baremetalcloud and AutoLab.

Summary

AutoLab is such an awesome tool; I highly recommend it to anyone looking to build home or test labs. Some VMware knowledge would be helpful if you're looking to use AutoLab. It's not difficult, but I imagine it could be somewhat confusing to someone who hasn't installed ESXi or vCenter before.

The three mistakes I made creating a Hyper-V virtual machine

When creating a Hyper-V virtual machine from scratch, beware of issues with Windows licenses and disk creation. Oh, and don't forget the administrator password! 

Our upcoming ERP upgrade project needed two new servers. We'd embarked on a virtualisation strategy and already had an IBM host machine, so the clear choice was to build them as virtual machines (VMs). These would be created from scratch rather than being physical-to-virtual (P2V) conversions.

I hit two pitfalls I wasn't previously aware of and made a mistake any self-respecting IT pro would be ashamed of. This article isn't so much a step-by-step how-to as a cautionary tale. 

Activation failure

One of the VMs would be a database server, so I was reassured to read of other IT prosendorsing the use of SQL Server in a virtualised environment. What's more, my IBM host server runs Windows Server 2008 R2 Enterprise, which includes Windows licenses for up to four guest VMs, so I knew I wouldn't have to buy any operating system (OS) licenses. With the IBM server, we were given two Windows license keys: one for the host server and a "virtual key" for use with VMs.

In Hyper-V Manager, creating a VM from scratch is as simple as running the New Virtual Machine Wizard. After allocating the required memory, choosing to create the VM not connected to the network, and specifying a location for the virtual hard disk (.VHD) file, I chose to install an OS from a DVD (Figure A).

Figure A

OS installation options

Since my database application needed Windows Server 2008 R2, I retrieved the OS DVD for another server running that OS and installed from that. When it came to the request to Activate Windows, I entered the virtual license key, which was rejected. Research indicated this was because the virtual key only works if the guest VM runs Windows Server 2008 R2 Enterprise. Now that wasn't obvious.

With a sigh I swapped to the IBM Windows DVD and restarted the setup process, this time choosing the Enterprise option. Since there was already a copy of Windows installed (the one that wouldn't accept my license key), I accepted the option of moving the existing files to a folder called Windows.Old, planning to delete them later. Unfortunately, this "in-place" install failed with an error, and I had to start from scratch, choosing to format the disk partition. Finally, the install completed, and the virtual key was accepted by the activation process.

Password failure

Continuing the setup process, I logged in with my chosen administrator password and followed the steps to upgrade the Hyper-V Integration Services. This required a reboot, after which I attempted to log in as administrator again -- and couldn't get the password right. I looked in disbelief at the password I'd written down, tried it several more times, and then tried a few variations. I couldn't believe I'd made a mistake like this. I'd managed to get the password right a couple of times up to now, but clearly that wasn't the password I'd written down.

There is a non-drastic way out of this embarrassing hole -- a password reset disk. However, I hadn't created one. I never do. I'll never need one of those, or so I thought. That only left me with the drastic solution -- reinstall the OS again. When I did, I chose an admin password no less secure but far less prone to mis-typing.

I've still not created a password reset disk, because, of course, I'll never need one.

Another disk, please

My new VMs were created with a single .VHD acting as the C: drive. I needed extra drives on both of them. To do this, the VM must first be shut down. In the Settings dialog for the VM, selecting an IDE Controller provides the option of adding a new Hard Drive. Creating the drive is wizard-driven, including the choice of disk type, size, and location. I specifically wanted a fixed size disk, rather than dynamically expanding, to guarantee the best performance for SQL Server. On clicking Finish, the wizard creates the new .VHD file.

And then you wait a long time -- at least 20 minutes for a 250Gb drive. Not only that, I started receiving complaints that applications on a different virtual server weren't responding. The explanation, it seems, is that when creating a fixed size .VHD, Hyper-V explicitly zeroes out every part of the disk to be used by the new .VHD. This is done for security purposes. (Note:Today when trying this link to an MSDN blog I was presented with a signup/login page I hadn't seen before. To get to the link I first had to cancel that dialog and then try again.)

My guess was that this zeroing process required so much disk I/O that it was affecting the other VM (which was on the same host). Sure enough, when the disk creation finished, the applications sprang back into life.

I apologised to my users and wondered if I could avoid this slow, resource-hungry process for the other disks I wanted to add. The MSDN article linked to a follow-up post describing a Microsoft tool designed to circumvent the slow creation by overwriting the relevant area of the disk without wiping it first. Since my host machine's disk had very little data on, it would have been safe for me to use that tool. In the end, though, I decided to stick to the normal way, but this time give my colleagues some pre-warning of the knock-on effects.

I (and they) waited patiently while the disks were created, and finally my new VMs were ready.

Summary

In creating two new Hyper-V VMs I discovered that the "virtual license key" supplied with Windows Server 2008 R2 Enterprise only works if the guest VMs also run Windows Server 2008 R2 Enterprise. I also found that adding a fixed size hard disk to a VM can be a slow, resource-intensive process that potentially affects other VMs on the same host.

Set up a basic website with Apache

Here are instructions on getting your small business website up and running with Apache on Linux or Windows. 

Apache is the most widely used web server on the planet -- and with good reason. Not only is Apache flexible and powerful, but it's also quite easy to get running and to start serving up content.

The idea of using a web server that relies so heavily on the command line and flat text file configuration might be rather daunting, but it doesn't need to be. I'll walk you through setting up a basic website with the latest iteration of Apache on Linux or Windows (I'm assuming you have a Linux or Windows machine ready for the task). After you follow this tutorial, you'll have either a Windows, Apache, MySQL, PHP (WAMP) or a Linux, Apache, MySQL, PHP (LAMP) server running and a "Hello World" page greeting your users.

WAMP

There are lots of WAMP servers you can install to get your Windows-based Apache server set up, but my favorite is the WampServer package, which installs everything you need for a web development environment. Here's how to install WampServer.

1. Download the WampServer installer.
2. Double-click the installer to start the process.
3. Walk through the user friendly installer wizard (it's a typical Windows application installation).

Once the installation is complete, you should see an entry for WampServer in your Start menu's All Programs section. In that menu, click the start WampServer entry. When the WampServer starts, left-click the icon in the System Tray to see the menu (Figure A); from this menu, you can start or stop the services and gain quick access to the server's configuration files. 

Figure A

The WampServer menu.

To get your website set up and running, the most important information you need is:

• Document Root: This is where your website files will live. For the WampServer, these files will be in C:\wamp\www.
• index.html: This is the most basic page for your website and should reside within the Document Root.

By default, WampServer will use the index.php file for its default page. Within the C:\wamp\www directory, rename the index.php file to OLD_index.php. In that same directory, create the index.html file and put this single line in the document:

Hello world!

Save that file and make sure your WampServer is running. Open your web browser and point it to the address of your WampServer. You should see the text (in bold):

Hello world!

If you do, congrats! Your WampServer is up and running.

Linux

I'll demonstrate how you can have a full-blown LAMP server up and running with just a couple of commands. For simplicity sake, I'll be demonstrating this on a Ubuntu 13.04 server. Here are the installation steps.

1. Open a terminal window.
2. Issue the command sudo apt-get install tasksel.
3. Type your sudo password and hit Enter.
4. Accept the installation.

After tasksel completes the installation, go back to the terminal window and do the following:

1. Enter the command sudo tasksel.
2. In the resulting window, use the down arrow key and move to LAMP Server (Figure B).
3. Press the Tab key to move down to OK and hit Enter.
4. When prompted, enter the passwords for MySQL.

Figure B

When this installation completes, your LAMP server will be running.

What you need to know:

• Document Root: /var/www/
• Index: The LAMP defaults to index.html for its main page. You can edit that page directly for your website.
• To start and stop Apache, issue these commands: sudo apachectl start or sudo apachectl stop.

Congratulations! You now have a web server. I bet you had no idea setting up Apache for simple web pages was that easy.

Keep in mind that Apache isn't limited to basic web pages; it's also just as easy to serve up robust and dynamic sites with this powerhouse server. 

Learn virtualization by creating a vSphere 5.5 home lab

If you're new to virtualization, a good way to familiarize yourself with the topic is to create a vSphere 5.5 home lab. Here's how to do it. 

I've received a couple of emails from readers asking how virtualization beginners can gain experience. My first piece of advice is to create a home or practice lab. This high-level overview outlines what you need to get a VMware vSphere 5.5 home lab up and running; there's a lot of documentation out there to help you with the actual installs. (Note: The intent is not to suggest the best hardware, because you'll most likely be using old servers or desktops that you already have.)

Hardware 

• One server, though two servers are preferable so you can work with cluster features such as Distributed Resource Scheduler (DRS) and High Availability (HA). These servers need at least 4 GB of RAM each, but more RAM is always better. Also, you need to make sure the servers offer virtualization support, which often can be enabled in the BIOS.
• A storage device (e.g., Iomega StorCenter) that you can connect to your servers. This isn't necessary for a vSphere install, but you'll need it if you want to practice cluster features. It's also possible to use the vSphere Storage Appliance (VSA) with your local drives on the physical hosts, though this is a more advanced installation for a first-timer.
• A switch to connect all the hosts, storage, and potentially the laptop or desktop you plan to use to connect.

Once all your hardware is connected to the switch, you can start installing the software. You'll need to download trial versions of vSphere components. As of this writing, vSphere 5.5 is the most recently released version, though the basics apply to all the versions. To download the trial versions of these components, you'll need to create a new login (it's free when you supply your email address).

Software

• ESXi 5.5: ESXi is the VMware operating system you load on to each of your physical hosts (or servers). 
• vCenter 5.5: This is the management software that gives you all the cool features you hear about, including HA, DRS, vMotion, and dozens more. vCenter can be a Virtual Appliance that you download and deploy, or you can download an installer to put on a Windows 2008 R2 or Windows 2012 server. I recommend using the Windows version first; then, after you gain some experience, you can try to deploy the Virtual Appliance.
• Windows Server 2008 R2 and Windows Server 2012: Although this isn't absolutely necessary, you'll need this for vCenter as noted above, and you'll probably want to create a domain controller so you can test domain connectivity. Download the ISO evaluation from Microsoft.com.

How to create your lab

ESXi

1. Create a boot disk with the ESXi image (.ISO) file you downloaded earlier.
2. Load the CD and start the server (make sure you're booting from the CD).
3. Follow the wizard to install ESXi. You'll specify which drive you want to install ESXi on and click F11 and Enter a lot.
4. Once it's installed, press F2 to configure the server with things like IP addresses, gateways, and host names.
5. Once you have it configured, try to ping the new server from a different computer.
6. Repeat these steps for each physical server.

vCenter

Now you can connect to your ESXi server from a remote computer.

1. On your remote computer, open a browser and browse to the IP address you gave the ESXi server.
2. You'll get a Web UI that has a link to download the vSphere Client. Click the link and open the vSphere Client.
3. In the vSphere Client, enter the IP address of your ESXi server, along with username (root) and the password you specified during the ESXi install. Now you're in the vSphere Client, which is where you'll do most of your work. There's also a Web UI that you'll see during the vCenter install.
4. Create a new VM for your domain controller by right-clicking the host.
5. Attach the Windows Server ISO to the VM and power it on. Install Windows as you normally would.
6. After Windows is installed, do a DCPROMO to create the domain controller.
7. Create domain admins to use for your vCenter install.
8. Create a new VM for your vCenter VM by right-clicking the host.
9. Attach the Windows Server ISO to the VM and power it on. Install Windows as you normally would.
10. After Windows is installed, attach the vCenter ISO to the VM and do the Simple Install (SSO, Inventory Service, and vCenter) to install everything you need for vCenter (use your domain admin account).
11. Once vCenter is installed, you can close your client, re-open it, and connect to the IP address of your vCenter server. 
12. Create a data center and a cluster, and then add all your physical hosts.

Two portable rootkit tools no SMB should be without

Keep Bitdefender's Rootkit Remover and Kaspersky's TDSSKiller on a USB drive, and your SMB will be ready when a machine is compromised by a rootkit. 

When a PC is infected with malware or viruses, you can usually scan with the installed antivirus and/or antimalware and move on. Rootkits, on the other hand, are tricky to remove and can reappear if they are not removed completely. For rootkits, you need the right software. The "right" software is subjective, but in the case of a rootkit removal tool, it either works or it doesn't.

Two tools I find to be effective for the removal of rootkits are Bitdefender's Rootkit Remover andKaspersky's TDSSKiller. Both tools are portable, so there's no installation necessary. When a machine won't allow you to install applications, portable apps might be the only way to remove rootkits. I'll walk you through the process of scanning for rootkits with each tool.

Safe Mode

Before you run a scan on a machine, it's always best to reboot the machine in Safe Mode in case something nefarious is running in the background that prevents the rootkit remover from starting up. If you're not sure, booting into Safe Mode is simply a matter of rebooting and tapping the F8 key until the Safe Mode menu appears. When that menu appears, select Safe Mode With Networking.

Bitdefender's Rootkit Remover

Bitdefender's Rootkit Remover only checks against known rootkits. Bitdefender's Rootkit Remover protects against the following:

• Mebroot
• All TDL families (TDL/SST/Pihar)
• Mayachok
• Mybios
• Plite
• Xpaj
• Whistler
• Alipop
• Cpd
• Fengd
• Fips
• Guntior
• MBR Locker
• Mebratix
• Niwa
• Ponreb
• Ramnit
• Stoned
• Yoddos
• Yurn
• Zegost

Bitdefender also cleans infections with Necurs (the last rootkit standing). New rootkit definitions are added as they become known; because of this, you will want to make sure you check the Bitdefender site and download a new version of the tool frequently.

After you download the .exe file, move it to your USB drive, and you're ready to move to the infected machine and scan. Insert the USB drive, open Explorer, and double click the BootkitRemoval_xXX.exe file (XX is either 64 or 86 depending on your architecture).

When the application starts up, you will be greeted with a window that has no settings, no preferences, and nothing to tweak (Figure A). When the Bitdefender window is open, click the Start Scan button. The scan will run and is incredibly fast. If the app finds a rootkit, it will automatically remove it and prompt you to restart the system. Figure A

Kaspersky's TDSSKiller

Kaspersky's take on rootkit removal is very similar to Bitdefender's, at least in the way its tool functions. The biggest difference is that Kaspersky focuses only on the TDSS rootkits (Rootkit.Win32.TDSS, Tidserv, TDSServ, or Alureon), which are some of the nastiest in the wild. These rootkits began to spread in 2008 and are one of the primary causes for the unauthorized Google Redirect issue (users do a Google search, click on a resulting link, and are sent to a random page). Kaspersky's TDSSKiller can also remove the Sinowa, Whistler, Phanta, Trup, and Stoned rootkits.

Another difference is that Kaspersky's also offers settings that can be tweaked. Kaspersky's TDSSKiller will remove or fix the following:

• Hidden service
• Blocked service
• Hidden file
• Blocked file
• Forged file
• Rootkit.Win32.BackBoot.gen

Here's how you use Kaspersky's TDSSKiller:

1. Download the executable file from the download site.
2. Move the .exe file to your USB drive.
3. Move the USB drive to the infected machine.
4. Double click the .exe file on the USB drive.
5. When the Kaspersky's window opens (Figure B), click the Start Scan button.
6. If Kaspersky's locates a rootkit, it will prompt you to take action.

Figure B

To access the options it has in terms of what objects to scan, click the Change Parameters link.

One thing that is very important with Kaspersky's is that if it does come up with results from the scan, make sure you know what you're about to delete isn't a false-positive. If Kaspersky does come back with a known rootkit, you will move the file to quarantine.

Summary

Bitdefender's and Kaspersky's offerings are a solid one-two punch that can be used to knock out a number of different rootkits. Keep these two tools on a USB drive, and you'll be ready when a machine is compromised.

Connect the Thunderbird email client to your Exchange server

Looking for an Outlook replacement? If so, try using Thunderbird with the ExQuilla addon to get an email client connected with an Exchange 2007 or 2010 server. 

Microsoft Exchange is one of the most widely used email servers in the business world. The problem many smaller businesses have with this is that, without Outlook, the only option is Exchange Webmail. This option is fine for some businesses, but for the ones that want an actual email client, it's been a challenge. That all changes now.

The ability to connect the Thunderbird email client to an Exchange server is finally possible. Thanks to Zendesk's ExQuilla Thunderbird addon, you can connect Thunderbird to your Exchange 2007 or 2010 server through Exchange Web Services (EWS). The setup just requires the installation of an addon and some knowledge about your server.

I will walk you through the installation of the addon as well as the setup of your Exchange account. I assume you already have Thunderbird installed and running, you have an Exchange account, and your Exchange server is either 2007 or 2010 and uses EWS.

Installing the plugin

1. Open Thunderbird.
2. Go to Tools | Addons.
3. In the search field, type Exchange.
4. Click the ExQuilla addon.
5. Click Install.
6. Click Restart within Thunderbird.

The addon should be installed. All that is left is to configure a new EWS account within the addon.

If you install the addon through the Thunderbird Extension Manager, you will most likely wind up with an out of date version of ExQuilla that will not work. If this is the case, install it with these steps:

1. Download the latest version of ExQuilla for your platform (Windows, Linux, or Mac).
2. Open Thunderbird.
3. Go to Tools | Addons.
4. Click the Tools drop-down and select Install From File.
5. Navigate to where you saved the file and select it.
6. Click the Install button (Figure A).
7. After the install completes, restart Thunderbird by clicking Restart Now.

Figure A

After the countdown, you should be able to click Install. (Click the image to enlarge.)

Setting up the account

Before you set up the account in Thunderbird, you must find out your EWS address -- it will be something like https://YOUR_DOMAIN/EWS/Exchange.asmx. After you enter that into your browser, you should be prompted for your credentials. The credentials necessary should be in the form of DOMAIN\username and your Active Directory password. Once you successfully authenticate here, you can move on to setting up the account in Thunderbird.

1. Go to Tools | ExQuilla for Microsoft Exchange and then select Add Microsoft Exchange Account.

2. In the first window of the wizard, enter your email address and password, leave Login With Email Address checked (Figure B), and click Next. Figure B

If the email authentication method fails, go back and use the Domain\Username method.

3. In the Exchange address setup dialog (Figure C), if your Exchange server is configured for auto discover, you should be able to click the Do Auto Discover button, and ExQuilla will locate the server address; if your Exchange server isn't set up for auto discover, you'll have to manually enter the address. Click the Manual button, enter the URL to the server in the Microsoft Exchange EWS URL field, and click Next. Figure C

Enter Your Name and keep the two check boxes at the bottom checked.

4. The window you will be presented lists the setup you just created. Click Finish, and the wizard will be complete. ExQuilla will begin to pull down your Exchange email into a newly created account. This new account should include all of your current emails, as well as your archives and subfolders in your Inbox folder hierarchy.

Conclusion

Although it's not a 100% feature-for-feature replacement, for anyone who can't afford Microsoft Office or just wants an alternative to Microsoft Outlook, using Thunderbird with the ExQuilla addon is a fantastic way to get an email client connected with an Exchange 2007 or 2010 server. You'll find Thunderbird works faster and more reliably than Outlook and is not as prone to the PST issues that can often haunt Outlook.

Troubleshoot Outlook connectivity with these quick tips

When Outlook won't connect to the Exchange server, follow these steps before calling IT for help. 

Microsoft Outlook is often rendered useless because it cannot connect to its Exchange server. Sometimes troubleshooting the issue is as simple as closing Outlook and restarting. In other instances, troubleshooting is much more challenging... or so it seems.

The following troubleshooting tips make solving that connectivity loss a snap. These instructions don't require a computer science degree to understand them, so just about anyone should be able to get Outlook re-connected to their Exchange server. We'll start with the simplest tip and increase the difficulty as we go along.

Uncheck offline mode

Oftentimes when a client calls and says, "My email won't work!" I find that Outlook was somehow set to offline mode. If you're using Outlook 2007 or earlier, click the File menu. If there is a check mark next to Work Offline, uncheck it, and you should be good to go.

If you're using Outlook 2010 or higher, follow these steps:

1. Click the Send/Receive tab.
2. Locate the Work Offline button.
3. Click the Offline button.

At the bottom of your Outlook window, you should see Trying To Connect.... If it connects, your problem is resolved; if not, move on to the next solution.

Restart

You should restart Outlook and, if that fails, restart your computer. I cannot tell you how many times I've seen Outlook connectivity issues resolve with a simple restart. The issue could be caused by the computer having connectivity issues. If you open your web browser and cannot reach a website or internal resources, that's most likely the problem. 

If Outlook still cannot connect and you cannot reach any websites or internal resources, contact your IT department because you have a networking issue. Once that is solved, Outlook will be fine.

Rebuild

Outlook can use two types of data files (.pst and .ost), and both are susceptible to errors that can cause connectivity problems. Here's how I handle this:

1. Close Outlook.
2. Open the Control Panel.
3. Locate the Mail icon (depending on how Windows Explorer is set up, you might have to click the Users section to find the Mail icon).
4. In the resulting window, click Data Files.
5. Select your data file from the list and click Open File Location (Figure A).
6. Locate the data file in question (it will probably have the same name as your email address).
7. If the file has the extension .ost, rename the extension to .OLD. If the file has the extension .pst, do nothing at this time.
8. Close these windows and open Outlook.

Figure A

This window will list all data files in use with Outlook.

Note: You need to be able to see file extensions in order to know if your data file is a .pst or .ost. This is handled through Windows Explorer settings.

If your data file is a .pst, follow these steps to run Scanpst on the file:

1. Search for scanpst.exe through Windows Explorer.
2. After you locate the file (e.g., a location could be C:\Program Files\Microsoft Office\Office14\), double click to run the application.
3. From the resulting window, click Browse (Figure B).
4. Locate your .pst file.
5. Click Start.

Figure B

If you've run Scanpst on your data file before, the location will already be in the field.

Scanpst will run eight passes over the data file; depending on the size of your data file, this can take quite awhile. If Scanpst finds errors in the data file, it will prompt you to click the Repair button. You should also check the box for Make Backup Of Scanned File Before Repairing in case something goes awry.

After the repair is complete, close Scanpst and re-open Outlook. If Outlook still cannot connect, move on to the next tip.

Repair install

You can run a repair installation of Microsoft Office; this will solve problems that standard fixes cannot repair. To do this, follow these steps:

1. Open the Control Panel.
2. Click Programs and Features.
3. Locate the entry for your Microsoft Office installation and select it.
4. Click Change.
5. Select Repair from the resulting window.
6. Click Continue.
7. Allow the repair to complete.
8. Reboot your computer.

After your computer has rebooted, start Outlook and hope for the best.

Recreate your profile

When all else fails, you can recreate your Outlook profile. I prefer to create a new profile (without deleting the old one) -- just in case. In order to recreate your profile, you need to know your account setting, so you should have that information before you begin. Here's how to create a new profile:

1. Open the Control Panel.
2. Open Mail.
3. Click Show Profiles.
4. Click Add (Figure C).
5. Give the profile a name.
6. Walk through the Outlook account setup wizard.
7. Once the profile is known to work, you'll either want to set that profile up as the default or delete the old profile.

Figure C

The Outlook profile manager.

If after all of these steps Outlook is still unable to connect, it's time to call the IT department. It could be a DNS issue, an Exchange issue, or a number of other possibilities that are outside the scope of this article.