Advertisements

Category Archives: VMWare

In Place Upgrade of Server 2008R2


So this week I’ve been engaged by a client to consolidate their virtual infrastructure on to Hyper-V 2012 R2 from a mix and match of VMware and Hyper-V 2008 R2. The client had an unused Hyper-V 2008 R2 box that they had planned on migrating all of this stuff to and the box was not in production but was domain joined so I figured this was a good opportunity to try a in place upgrade of that server. So how about I jump right in.

 

Optional Prerequisite: While not necessary, I would recommend only doing this on servers you either have physical access, or have RMM/IP KVM access to. That way you’ll be able to see what’s going on during the reboots.

 

Step 1) prepare the server for the upgrade. Since this machine was not a domain controller I did not have to do any of the AD prep normally associated with an in place upgrade, so the only preparation step I did was make certain that windows updates were current.

Step 2) Next I installed pfmap173 which you can get here. This is a free Windows application that allows users to mount the contents of ZIP, ISO, Compact ISO, Compact File Set and Private Folder files to the file system as virtual folders. Very useful piece of software.

Step 3) Now that we have the ISO mounted, just run setup.exe and start the install.

 

Step 4) Selecting whether to get updates or not. It’s going to recommend you get updates during the upgrade process, while I’ve done this with desktop upgrades from Windows 7 to Windows 8.1 and not had any issues, I really felt that since this was a server and I didn’t want to drive an hour to go physically lay hands on it if something goes wrong, I chose to be as minimalist as possible and selected the option to not install the updates. This was purely intuition on my part, I can’t cite a specific technical reason not to do this so your mileage may vary.

 

Step 5) Next step you will select what edition you are installing. Obviously this is going to be determined by the license key you have and in this case I’m going to do Standard, but I’ll go over some basic facts about editions.

Data Center can be licensed for 2 sockets and gives you unlimited VOSEs with Automated Virtual Machine Activation. This basically means you can do an unlimited number of VMs on the Data Center host and the AVMA means they don’t need to be activated through KMS or online. BUT, these VOSEs are not transferable to another host, so any cluster replication members must be Data Center as well or at least licensed for the total number of VMs that could be run if one host fails. Example: Host 1 has 2 VMs, Host 2 has 2 VMs and Host 1 and Host 2 are in a cluster, so that means Host 1 must be licensed for 4 VMs and Host 2 must be licensed for 4 VMs because if either host fails the subsequent host will be handling those VMs and must be licensed for them even if it’s just for 1 second.

Core editions are pretty straight forward, it’s basically whatever edition you have but without a GUI. This reduces servicing, management, resource consumption, and attack surface, all useful things in some scenarios. Your license keys can be used on both versions (with GUI and without).

Step 6) Select the upgrade option.

The install process will then examine compatibility, there are any issues you’ll be notified here and it also spits out a report onto the desktop if you wish to cancel the process and go look at it. It will also warn you to go check with your software vendors to make sure software is compatible. I’m guessing you’ve already done this because you’re not incompetent but just in case it gives you the option of stopping now and taking care of that.

 

 

Once you hit next, there’s no going back, it’s going to go through the unpacking and installing process and the machine will reboot several times.

If you don’t have direct RMM/IPKVM access to the machine, you’ll lose your RDP session and won’t be able to connect again until the process is totally complete. When you’re all done, you should be greeted by the lovely Server 2012R2 logon.

 

Zach,

your friendly neighborhood IT solutions engineer

Advertisements

PVS Capture Issue


So after many hours of Trouble shooting thinking I was battling a VMWare 5.5 Hardware Version 10 issue.  Became a super easy Partition layout issue.  So here the run down.

I was trying to capture a Golden image, had everything setup and working.  I was giving a new Base Build image to capture and started having problems when I tried to run through the Image Wizard… This would run through without issues until the reboot, it would come up attach the vDisk but wouldn’t capture the image.  I tried every combo I could think of going to an E1000 instead of the VMXNET3, still got the same results.

Now for the good news, I found that it was a simple Partition issue, the Wizard hard codes the small windows 300mb partitions to the first volume, however I found running the P2PV tool under the Provisioning Services folder, selecting This machine to VHD, and manually moving the 300mb partition to the last part of the Drive and moving the Data Volume to the front resolved the issue.  Granted you have to point the saved VHD to the PVS Server, and you will have to manually import the VHD after it is captured.  

Anyways thought I would share this, thanks for reading.

Why is a VDI Assessment a Good Idea?


I am going to start this off with a bit of a disclaimer.  I really am not trying to sell anything!  Just want to get the information out there, I highly recommend talking to your trusted advisor(s) before starting a PoC/Pilot just my two cents worth.

So you have most likely heard of a “VDI Assessment”, and may have even pondered the thought of having one done.  Well here are a few reasons to take the next step and get that Assessment.

Some of the main things I hear:

1. I want to go VDI but don’t have the resources…

2. I want to go VDI but it’s not “good” or doesn’t work well

3. I want to go VDI but it is to much over head NO ROI

I will just start right in…

1. Most likely you are half right about the resources. You most likely have the infrastrucure to support a VDI Environment, and you may even have the resources you just don’t know it, or you can “Exchange” one type of resource for another… I have seen many cases where we were able to get rid of OLD style VDI (RDP to Desktops) & (RDP one to one Servers) and move to a slimmer Thin Provisioned Style VDI Deployment. For example: Say you have 100 Desktop’s you want to Virtualize.  Windows 7 recommends you go no less then 24GB for a disk I recommend 30-50GB Depending on Application(s) needed. 2GB of RAM, and 2 vCPU’s (avg. config for win7). In theory, you would need 30GB x 100 Desktops in disc space alone!  This is not the case. You will only need 4GB x 100 Desktops as you will only need to account for the delta files (Cache Disk). Note: If you are running ESX you will also need to account for a SWAP file which is the same size as your machines RAM. In this example you would need 2GB x 100 extra disk space.

2. This is a common misconception, which in fact, is what most people see in self-deployed PoC’s and Pilots.  There are a number of tweaks your friendly Partner/VAR should know that will help improve performance and usability.  This is because “this is what we do”. We have ALOT of tricks up our sleeves. I have been involved with several Citrix Projects where “Citrix” was considered such a bad word that the Citrix team could hardly walk down the hall without getting mauled. In less then 6 months we had users asking to be moved to XenDesktop. Talk about a culture shift! Remote workers suddenly jumped by a good 15-20% because they could actually function while working remotely.  Everything roamed with them. (as advertised)  No more working in one environment while at work and a completely different environment while working remotely… Another common issue I see is a lot of people think going VDI also means going with some sort of application streaming, or thin-app.  This is not always the best solution, as it doesn’t necessarily save you all that much disk space in the long run. Using the example in Statement 1, the only thing you really need to be concerned about is your Delta (cache disks). So, in some cases, streamed or thin-apps can cause double the overhead.

3. Well this one is a fine line. Really, you can make VDI as simple or as complicated as you want :).  Most of the time I hear “My users won’t buy off on VDI.” Again, this is a false statement burnt into you from the Beginning of Statement 2.  With the proper configuration there is no reason users would complain and in most cases once the users logon they are hooked.  ROI really should be Return On Individual, if we can make our users (which in fact are our customers) happy and more Productive we have succeeded.  I have seen the cost estimation on the Individual’s time and that was staggering… We are talking 100’s of thousands of dollars a year in regained work and help desk calls.  I saw a customer where their computers would take roughly 10 minutes just to login, this was on top of the 10 minute boot time if you abide corporate policy and shutdown your machine which took 10 minutes… as you can see time was adding up in the NOT working area extremely fast. So for pretty much every HelpDesk call workers and the company where out about a half hour of actual work…  Now for the ROI part if you go with thin clients, and high capacity servers you can really start saving money just in power, then looking at the management of a VDI deployment this again saves money as everything is the Same! No guessing if a patch, update, or user screwed up the machine simply have them reboot and back to the Golden Desktop.

With all this being said, NOT every application, or use case is a “Good Fit” for VDI.  And this is one of the key discussion in having a VDI Assessment done, we can tell you whether or not going VDI is a good fit.

Thanks for reading.

XenDesktop 5.5 and Provisioning Services 6.0


If you are planning on deploying XenDesktop 5.5 and Provisioning Services (PVS) 6.0 with a XenServer or ESX infrastructure you may want to check out the following Citrix KB  http://support.citrix.com/article/CTX130851 This is the fix/work around for the famous “Black Screen” issue.

Just in case you are impatient like me you just have to run the VDA installer manually as follows:

XenDesktopVdaSetup.exe /NOCITRIXWDDM
This can be found under the XenDesktop 5.5 Installer DVD

Cloning a Citrix 4.5 box in VMWare


  1. Login to the Citrix server you want to Clone
  2. Verify no one is logged in
  3. Shutdown the server…
  4. Once it is shutdown,
    1. open Virtual Center
    2. select the Citrix Server
    3. select Clone to New Virtual Machine
      1. Enter the Name of the New Virtual Machine
      2. Select the Folder
      3. Select the Host/Cluster
      4. Select the correct Resource Pool
      5. Select the Correct Datestore
      6. Select Customize using the Customization Wizard
        1. You will need to enter a valid server name
        1. Enter the Windows License Key and the correct TS CAL’s
        1. Enter the correct Local Administrator Password
        1. Select the correct Time Zone
        1. Citrix and myself Recommend either Static IP or Reserved DHCP
          1. Click Customize…
            1. You will need a valid IP Address to use DO NOT USE THE ONE IN THIS EXAMPLE
        1. Make sure you DO NOT Select “power on the new virtual machine after creation”
      7. Once the Clone has complete successfully
      8. Power on the OLD Server
      9. Click Edit Settings on the New Virtual Machine
          1. Select Network Adapter
          2. Uncheck the Connect at power on
          3. Click OK
        1. Power on the New Virtual Machine
        2. Open the Console
        3. Time to make Citrix Work!
          1. Login once it has finished its sysprep
            1. You will get an alert that a service couldn’t be started
          2. First go through the Registry and replace the old servername with the new servername
            1. There will be roughly 20-30 places to make this change once you start the search you can press F3 to find the next.
          3. First go through the Registry and replace the old IP Address with the new IP Address
          4. Browse to c:\program files\citrix\Independent Management Architecture\MF20.DSN and change the old servername to the new servername Note: you will have to open the file with Notepad
            1. WSID=new servername
          5. Browse to c:\program files\citrix\System32\CtxSta.config
            1. Change UID=STAxxxxxxxxxxxx
              to some random number that hasn’t been used on another server.
          6. Shutdown the Server…
      10. Click Edit Settings on the New Virtual Machine
          1. Select Network Adapter
          2. Check the Connect at power on
          3. Click OK
        1. Power on the New Virtual Machine
        2. Open the Console
        3. Login and add the server to the domain
        4. Change the Page file to 4092
        5. And make the following changes
        6. Reboot and you have yourself a New Citrix Server in about an 1hr – 2hrs
          1. Just make the Following “house cleaning changes” and you are all set.
      11. Open the Presentation Server Console not the Access Console.
          1. Move the Cloned Server to the Correct ZONE
      12. Within the Access Console
          1. Select the New Server and Click Add, Click OK
        1. Close all windows
        2. Reboot
%d bloggers like this: