AutoLab

80 comments

1. Pingback: vSphere 5 AutoLab
2. Pingback: FullAware » Blog Archive » vSphere Lab Builder Kit

3. AutoLab looks very promising. The Installation Instructions PDF is missing a few things.

   1) On page 10 screenshot there is a reboot.ps1 where does it come from ?
   2) There is a Deploy.cab on the page 10 screenshot, where does it come from ?
   3) Page 10, you state to boot the DC and VC vm with “Windows Server 2008 R2 CDROM” I’ve tried using en_windows_server_2008_r2_with_sp1_x64_dvd_617601 and en_windows_server_2008_datacenter_enterprise_standard_sp2_x64_dvd_342336 but I get an error message during the install. Can you confirm which “Windows Server 2008 R2 CDROM” you are referencing to ?

   thanks
   Erik

   1. 3) I guess it’s the Microsoft Windows Server 2008 R2 180 day trial DVD that is mentionned in the list above my question. Retrieving the iso file and will try it.

      1. 3 ) The DC requires the en_windows_server_2008_r2_x64_dvd_x15-50365.iso (Windows Server 2008 R2 (x64) release on Release Date: 8/31/2009|). I’ve tried various version with integrated SP1 but the unattend.xml doesn’t work with a Build 6.1.7601

   2. 1. I’ll update the screen shot, that was a test script.
      2. Deploy.cab is off your WinInstall.iso file & is the source for Sysprep. I’ll update readme.html since It’s been lost off there.
      3. For the moment only use Windows 2008R2 RTM, I’ll work on getting SP1 support into the next release.
      4. ESXi 5.0 Update 1 doesn’t appear to deploy correctly, also to be added to the next release.

      Thanks for the feedback.

      Al.

      1. Great work.

         Updating autoUnattend.xml on floppy image (DC and VC) from :-

         With

         Will allow VM’s to deploy using Windows 2008 R2 SP1 iso

         Regards,

         Paul

         1. Oops didn’t like the tags. Replace the version number in autoUnattend..xml from version=”6.1.7600.16385″ to version=”6.1.7601.17514″

            1. This is the solution, it doesn’t look like a single .xml file can work with multiple versions of Windows and do the things I need it to do.
               AutoLab build 0.8 will come with floppies for both 2008 R2 RTM and 2008 R2 SP1.

4. Great job, Alastair! Very useful and huge contribution to the community!

5. Just wondering if freely distributed also means we are free to change and redistribute also?

   I work for HP and I can are this being very useful with our VSA, Tipping Point, etc.

   1. Hi Dan,

      I released the AutoLab to help people, you are very welcome to modify and redistribute if that helps. Please do attribute the work that contributed, ours and the Open Source developers.
      If the finished product is suitable for public release then please do release and let us know.

      Al.

6. Pingback: vSphere 5 AutoLab

7. great work, thank you for this, you have made many peoples like much easier, interms of home lab setup
   Thanks once again

8. Great work!
   I just drop my old lab layout in favor of AutoLab, too cool to miss out.

   I found a few things and before they gone lost in my mind….

   1. The router user name is root, there is no admin in /etc/passwd
   2. There is an screenshot of Openfiler but the NAS is FreeNAS
   3. In the IP Addressing section it says “OpenFiler NAS”

   1. Thanks for these, I’ve updated the documents for the next release which I hope to have out this month.

9. Hi,

   Thanks for the wonderfull job !!!

   regarding the install on esxi :
   – the ovf import failed because it must not be done directly on esxi but on a VC as there is multiple VM :-(
   Would you please provide the several ovf to solve this issue (it is worlinkg fine from a VC but as the goal of this autolab is to build the VC…)

   Regarding the install on Fusion :
   – as there is no GUI to manage network even in the beta of the future version , I found a usefull tool to create the vmnet3
   http://nickapedia.com/2012/01/10/breaking-new-ground-an-uber-tool-for-the-mac/ (if you can include it in your next documentation it should save users time)

   Global issue I met :
   – the link you provide for 2008 R2 is with SP1 included so in order to have the autounnatended.xml file to work it need to be edited and to have version=”6.1.7601.17514″ in the line (must be done for DC.flp and VC.flp):

   – In the documentation you say we need to provide the iso of vmtools inside the folder, in fact it is the content of the iso files

   This comment is just there for correction or at least for people looking for a solution
   Again thanks a lot for this wonderfull tool

   1. Hi Benjamin,

      Thanks for your feedback. There’s a new version (0.8) being released very soon to address these issues.
      I’ll pass the info you’ve provided to Alastair to make sure we get each item fixed.

      Thanks again.

      Cheers,
      Nick

10. Pingback: #vBrownBag: vCloud Director & the AutoLab — DAMIAN KARLSON

11. I recieved error ‘No such file or directory /tardisks/scsi-rst.v00’ when running the esxi50 host installs (manual and automatic)

    To resolve I added the following /EXXi50/scsi-rst.v00 under modules in the besx1-5.cfg and besx2-5.cfg files in the tftp-root directory on DC.
    This can also be added to the unattend\DC folder if you have not already deployed the DC

    I was using esxi5 up1 v623860 from the vm site.

    1. Hi Corey,

       The error you have is due to VMware changing the boot files with 5.0 Update 1. The next release of the AutoLab will support Update1, along with Windows 2008R2 SP1.

12. Nick and Alastair,

    I think you two have done a bang up job with this lab. I had a few misfires at first but I was able to figure it out and now I can very quickly deploy a fresh lab when needed. I’m using this to prep for the VCAP5-DCA beta.

    I only suggest updating the readme file a bit. It doesn’t show the Deploy.cab file needed in the build folder. The ESXi* and VIM_* folders ask for the contents of the install DVD or CD… OK, I’m friggin dense and I just copied the damn ISO file. (DOH). maybe you can tell idiots like me to extract the files from any ISOs into these folders…

    Also, for some reason, the extraction of files from deploy.cab never worked for me. I need to try building a new lab to see if I can replicate the issue and give you more details…

    Once I got past my own idiocy, I found that I had a damn good lab. Thank you for saving me a lot of work…

    1. Thanks Rick. We’re working on the wording to make the readme.html prettier and clearer that you extract the ISOs.
       The extraction of sysprep files from deploy.cab has been a bit hit or miss, the next version extracts deploy.cab from your WinInstall.iso file.
       The new version also uses the SQL Client installer from the vCenter install files so you don’t need to download so much.

       1. I’m using Autolab v0.8 today. I have to say that everything is working PERFECTLY… Again, nice work. I’m recommending this to all my guys at work…

13. Hi All,

    Another error I found is within these 2 files: build\unattend\validate.ps1 and build\unattend\vc\vcbuild2.cmd.
    Just do a find for sqlcli.msi and replace all with sqlncli.msi.

    Hope it helps someone.

14. Pingback: Montar Laboratório de VMware ESXI usando VMware Workstation 8 « Blog sobre Infraestrutura de Servidores
15. Pingback: LabGuides.com – vSphere 5.0 AutoLab version 0.8 released

16. Very nice lab automation! You did a Good job with Version 0.8!
    I was able to manage the installation of the DC and the NAS. But the VC allways hangs by the vCenter Repository Installation.
    On my DC the vCenter Database has been built and the script shows no error.
    Do you have an idea what I’m doing wrong?

    Best regards

    1. Thanks for your reply. I have to add that I’m using it on a VMware Workstation installation with 20GB RAM. Proably I got a problem with the floppy-router because he can’t get an IP over DHCP, but I can’t imagine, this reflects the installation of the vCenter….

    2. Glad you like the AutoLab.
       It does sound like the DC build is broken since it is the DHCP server as well as the DB server.
       Check the output of the Validate script on the DC & possibly simply rebuild.

17. @sporca : did you move the VM from the build datastore , if not this could be the issue

    Very good work as usual, I was able to deploy inside esxi without issue

18. Pingback: APAC vBrownBag returning in July | Demitasse
19. Pingback: APAC vBrownBag returning in July

20. script add host apresent problem

    error:
    This script should not be ” Run As Administrator”
    Just double click the shortcut

    tanks

    1. Hi Ronaldo,
       The message occurs when the script detects that it has been run as administrator rather as the current logged on user, VI-Admin.
       The error will probably also display if the script is run on the DC rather than VC, the DC doesn’t have PowerCLI installed so the script won’t work there.
       As the message says, just double click the shortcut on the desktop of the VC.

       1. I was having the same problem with…for people that don’t know much (like myself) this happens if UAC is disabled. Just Enable it and reboot the VC server.

          By the way…This is awesome!!!!

21. Hi there,

    First of all kudos to this gr8 project and its developers. We hope you will continue adding new features to the project.

    I have found the following minor typos with the 0.8 pdf documentation:
    In the Virtual Machines section, for the FreeNAS VM it should be memory 128 MB (512 MB) and the Hard Disk should be 80 GB Thin Provisioned as per the settings of the VM.

    Rdgs,
    User

    1. Thanks, there will definitely be new features coming, once I get some more time!
       I’ve updated my master doc for the typos and updates.

       Al.

       1. Thanks Al for your reply and hard work.

          Can you pls give a brief outline what new features you plan to add? Maybe I can try adding on my own and if it works, I`ll share the configs done.

          Rdgs,
          User

22. Pingback: New Open Source Storage NAS on the block: NAS4Free | Virtual Bits & Bytes

23. Hi Again,

    As you may already know, freenas 7 has moved to a new project called nas4free. So for geeks wishing newer features for freenas we have 2 options: either upgrade to nas4free or move to freenas 8.

    I decided to go with Nas4free because the interface is similar and I didnt have to redo all the settings done in the NAS VM. Pls follow the steps exactly as indicated in the following post and support the developers who are creating great softwares for our labs.

    http://sourceforge.net/apps/phpbb/nas4free/viewtopic.php?f=47&t=36&start=0

    I have successfully upgraded the NAS VM in the autolab to NAS4Free 9.0.0.1.133. I tested this using the NAS4Free 32bit version only and its working fine with all shares enabled. Then I rebuilt my Autolab using this new VM without any issues.

    Hope it will be useful to someone

    Rdgs,
    User

24. hello, everyone
    I’m new to the VM world and i’m having issues setting up the lab. I populated the folders in the build directory, but when trying to setup the DC with auto.xml i get error can’t connect to build folder to build DC. i checked NAS and Router. they are running. don’t know if i missed a step. can someone please help me? thanks

    1. I have the same error, the build share is available and accessible. I am using VMware Workstation 8

       1. The most likely thing is that the DC VM is not connected to the VMNet3 network that the NAS is using. The best test is to ping the NAS (192.168.199.7) from the DC (192.168.199.4) If you are unable ti ping then it is most likely to be the network setup on one of the VMs.

          1. Alastair,
             i’m able to ping 192.168.199.7 but is still receiving error system error 85 has occurred. build sources not found, is the NAS VM running? Make sure Build share is available and populated. I know its connected because it shows up as a network share. thanks for the help. i’m lost..

             1. I was having the same problem, I was able to ping the share and it was mapped in the DC VM as well. I figured out what was wrong. In my case, I missed a step, I did not extract the VM Tools .iso file for VMware Workstation in the VM, I did just that and deleted the DC VM for which the installation had stalled, copied the original DC from the download and powered it on. the installation completed flawlessly.

             2. Looking at the phase2.cmd build script for the DC the only test it does for the Build share is whether the file VMTools\Setup64.exe exists.
                The files inside the ISOs need to be extracted into the folders.

25. Thanks. I tried it and it worked to perfection. time to play and learn. i’m like a kid in a candy store, now..

26. I am getting error when I am trying to install the esx host “Configuration error while parsing esxi50/Besx1-5.cfg
    Fatal error : 15 not found.

    I am not able to proceed further.

    ESX used VMware-VMvisor-Installer-5.0.0.update01-623860.x86_64
    Vcenter VMware-VIMSetup-all-5.0.0-456005

    1. The most likely reason for that error message is that the build script didn’t identify the ESXi v5.0 install folder on your build share when the DC was built. Take a look at this post on ProfessionalVMware http://professionalvmware.com/2012/06/autolab-automation-uncovered-windows-builds/ for more information about how the Windows server builds work.

       1. This is great work Alastair! I appreciate you putting this together and the podcasts.

          I ran in to this too but to detail for the next person:
          I logged in to the Lab_DC
          reconnected to the build drive
          On the build drive, went to: Automate > DC > ESXi5_0_U1 folder and copied Besx1-5.CFG, Besx2-5.CFG, and BOOT.CFG to C:\TFTP-Root and the install would continue.

          1. Hi!
             Don’t mean to hijack your thread, but I am having the same issue, I have tried what you have mentioned but the error still continues. I was trying this with version 5 first, so dowloaded the latest from vmware website (VMware-VMvisor-Installer-5.0.0.update01-623860.x86_64.iso) and copied over the U1 .cfg files to C:\tftp-Root on DC, but the error still comes up.
             Also when I try to boot from the installer I am getting the error /tboot.b00: file not found fatal error:15 (not found)
             Any help will be greatly appreciated.

             Thanks,
             Rehman.

             1. Hi !
                Some progress!!. I edited the phase2 batch to only copy upto the esxi5 files (didn’t wanted to redo the dhcp, dns and database), renamed the TFTP-Root, and executed the phase2 batch file, this copied esxi5 files and config files, it started working after this until it stopped (error below) and also tried a attended install but failed at the below error.
                [Erron 2] No such file or directory : /’tardisk/scsi-rst.v00′

                Thanks,
                Rehman.

27. Hi All,

    I found a small error in the freesco router VM.

    1. Login to router using root account and password
    2. Type setup to enter configuration mode and press c.
    3. Using your keyboard navigate to Configure Servers, press Enter.
    4. Select Port Forwarding and press Enter.
    5. Select y for Worldwide services or ports and then press any key to continue.
    6. Using your down arrow key on keyboard, navigate to the line:
    tcp,122,22,192.168.199.11 # SSH to Host1
    As you will see the port number 122 is wrong as compared to the entry for SSH to Host2.
    7. Edit the entry to the following tcp,222,22,192.168.199.11 # SSH to Host1
    8. Press F10 and then y to save the portfw.cfg file settings.
    9. Press x and then s to save router settings. After this reboot the router with the command reboot.

    Sorry to all experienced users for such detailed procedure. I thought it will be helpful to newbies.
    Hope it helps

    Rdgs,
    User

28. Pingback: THINKING LOUD ON CLOUD » My VCAP DCA Experience

29. hi,all
    host1,host2 esx host “Configuration error while parsing esxi50/Besx1-5.cfg
    Fatal error : 15 not found.

    please help me….

    1. Hi muneer,

       Can you pls tell what is the version of ESXi 5 you are using? ESXi 5 RTM or ESXi 5 Update 1?

       If its RTM, then you can edit Besx1-5.CFG and Besx2-5.CFG in \\192.168.199.7\Build\Automate\DC\ESXi5_0_RTM folder. You will see there is a space before the last line kernelopt=ks=nfs://192.168.199.7/mnt/LABVOL/Build/Automate/Hosts/esx1-5.cfg. This additional space is found in both config files for RTM whereas it is not present in ESXi5_0_U1.

       I can`t confirm this is the problem as I am using ESXi 5 U1 and I have no problems with it. Also I don`t have access to RTM version to verify the bug.

       Pls remove the space mentioned before and rebuild the lab. If it works pls confirm so that others can benefit.
       If it does not work then outline the steps you did for the lab so that I can pinpoint the bug.

       Rdgs,
       User

       1. Hi muneer,

          Pls verify the post of Alastair Cooke and the reply of Mike Munoz a few posts above. They both give a workaround for the problem you are having. Sorry I didn`t notice it before…. :-)

          Rdgs,
          User

30. Pingback: Autolab in VMWARE Workstation 8 on Linux « 24 by 7, 3 6 5
31. Pingback: Autolab (VMware Tool) | Electric Monk
32. Pingback: The MicroServer as a training device continued - Bladewatch.com | Bladewatch.com

33. This is awesome. Just finished building the lab and everything worked flawlessly. Thanks!

34. Hi Guys,

    You have done an excellent job by creating this AutoLab. So well done and keep it up!!
    Now I am getting an error when I am trying to build the first host. As soon as I boot the host I can see the PXE menu, but then it fails with an error:-
    “PXE API call failed, error 0000” and doesnt load the ESXi setup. Any thoughts…

    Thanks in advance
    RM

    1. I get The same failure message … I´m trying to understand it but, I didn´t get it yet

       keep on searching!

35. Anyone building vSphere 4.1 Labs ?
    With the AutoLab 0.8, I was doing that and found that Host2 refuse to boot from the PXE menu, the problem is the entry in default (inside theTFTP-root/pxelinux.cfg/) the entry for Host2 was like this:

    label esx2-4
    kernel /ESX41/vmlinuz
    append initrd=/ESX41/initrd.img debugLogToSerial=1 mem=512M ks=nfs:192.168.199.7:/mnt/LABVOL/Build/Automate/Hosts/esx2-4c.cfg
    menu Label Host2

    I changed it to the same as Host1 with the cfg for Host2.

    label esx2-4
    kernel /ESXi41/mboot.c32
    append /ESXi41/vmkboot.gz ks=nfs://192.168.199.7/mnt/LABVOL/Build/Automate/Hosts/esx2-4.cfg — /ESXi41/vmkernel.gz — /ESXi41/sys.vgz — /ESXi41/cim.vgz — /ESXi41/ienviron.vgz — /ESXi41/install.vgz
    menu Label Host2

36. Further Update!! Looks like there was a problem with the copying of files onto C:\tftp of the labDC. Edited the phase2.bat file to exclude all the DHCP and DNS config and re-ran it. This time waited for it to complete and the installation was 100%!

    1. Hi all,

       Thanks Rehman for pointing out the issue.

       There are times when I rebuilt my lab, I noticed that the ESXi50 folder was not being populated properly. I modified the line as below and it completed successfully. The problem lies within this section of the code in the file phase2.cmd:

       :ESXi50
       If not exist b:\ESXi50\TOOLS.T00 goto ESXiDone
       echo **
       echo * Add ESXi 5.0 to TFTP
       echo * Add ESXi 5.0 to TFTP >> c:\buildlog.txt
       mkdir C:\TFTP-Root\ESXi50 >> c:\buildlog.txt
       xcopy b:\ESXi50\*.* C:\TFTP-Root\ESXi50 /s /c /q >> c:\buildlog.txt

       I modified the line in phase2.cmd to:
       xcopy b:\ESXi50\*.* C:\TFTP-Root\ESXi50\*.* /s /c /q /i /y >> c:\buildlog.txt

       Hope it helps someone.

       Rdgs,
       User

37. HI Alaister, first of all, I wannna thank you for this gift, people who want to get the VCP certification, we must be very grateful…

    in the way to build de lab i have found a little stone, when I build the DC and VC, the VC gives me the next message at the end of the vCenter installation:

    “VC Build Fails to create vCenter Repository”

    but when I rebuild the DC with an “Microsoft SQL Server 2008 R2 RTM – Management Studio Express” the installation seems to have been successful but it throws the following error:

    C:\Users\Administrator>”C:\Program Files\Microsoft SQL Server\100\Tools\Binn\sq
    cmd.exe” -S dc\SQLEXPRESS -i B:\Automate\DC\MakeDB.txt >> c:\buildlog.txt
    Sqlcmd: Error: Microsoft SQL Server Native Client 10.0 : A network-related or i
    stance-specific error has occurred while establishing a connection to SQL Serve
    . Server is not found or not accessible. Check if instance name is correct and
    f SQL Server is configured to allow remote connections. For more information se
    SQL Server Books Online..
    Sqlcmd: Error: Microsoft SQL Server Native Client 10.0 : Login timeout expired.

    not really may be happening, I’ve been running around all weekend, i apreciate some help

    thanks a lot

38. Hi there could someone please help me identify this issue .. I get it when installing both the LAB DC and VC….it occurs after all files have been copied, and just as windows installer is ‘Installing Updates’
    WINDOWS COULD NOT APPLY UNATTEND SETTINGS DURING PASS [offlineServicing]
    After I receive the error the VM just reboots :-(

    1. Problem solved..!!
       in the boot floppies ( .FLP file) extract the contents

       Check the R2 version in autounattend.xml by editting it using notepad or notepad++
       This is for windows 2008 SP1 = version=”6.1.7601.17514″
       This is for windows 2008 R2 (no SP) = version=”6.1.7600.16385″
       save the file ..
       save it back to the .FLP file
       attach the .FLP file back to the vm

       1. Hi Agian
          Sadly it does not work for the LAB VC :;-(
          I am still experieincing the same issue !
          Windows could not apply the unattend settings during pass [offlineServicing].
          There seems to be an issue either with the autounattend.xml or something else :-(
          I wish my posts would be moderated and published so I can get some help.
          I am using the vCenter automation.
          D

          1. Fixed !!
             Windows version is the problem! check reply in second post.
             Gosh these builds take forever !
             isnt it faster just to deply from OVA/OVF
             D

             1. Oh and by the way .. the user VI-admin you have set to autologoon always fails after build with the wrong password, so of course automation halts at login.
                Will post a fix when i get a chance
                D

39. Pingback: AutoLab automation uncovered, PowerShell
40. Pingback: Veeam sponsor AutoLab | Demitasse

41. Download link ?

    1. Just updating the page now, you were too quick :)

    2. Download button doesn’t show up in Chrome. It shows up in IE.

       1. Thanks :-)

42. Pingback: LabGuides.com – AutoLab 1.0 Released

Comments are closed.