I recently wrote a GUI to duplicate MDT database roles based on the excellent Powershell commandlets written by Michael Niehaus:
Although the Powershell code worked fine when executed directly in a Powershell window, I was seeing mixed results when executing the same code in C# through PowershellInstance.invoke(). Specifically, when I compiled the executable I was able to run the powershell code through the utility without issue, whereas it failed for my colleagues for some reason.
It seems that what I needed to do was ensure that the code was compiled specifically for the platform I was running it on as opposed to ‘any CPU’. In my case this was for x64. You can set this under the project properties in Visual Studio:
However, after making this change, I found I was getting compilation errors along the lines of:
An attempt was made to load an assembly with an incorrect format: [path to exe]
To fix this, I needed to change Generate serialization assemply to Off and everything compiled as expected. More on this here.
CM2012 introduces to us the ability to easily script many labour-intensive tasks. I was recently adding numerous boundaries to a ConfigMgr implementation for a client so decided the best approach was to automate this procedure. I fired up the Powershell session from within ConfigMgr and checked the syntax for New-CMBoundary. Now, I had a spreadsheet with a bunch of IP subnets in CIDR notation so it made sense to use this for my script. Running Get-Help New-CMBoundary I got the following:
Example 1: Create a new IP Subnet site boundary
PS C:\>New-CMBoundary -DisplayName "IPSubNetBoundary01" -BoundaryType IPSubNet -Value "172.16.50.0/24" BoundaryFlags: 0 BoundaryID: 6338009 BoundaryType: 0 CreatedBy: Contoso\PFuller CreatedOn 6/10/2012 1:17:42 PM DefaultSiteCode: DisplayName: IPSubNetBoundary01 GroupCount: 0 ModifiedBy: ModifiedOn: SiteSystems: Value: 172.16.50.0/24
Looks straightforward enough but no. This simply doesn’t work as stated. The problem seems to be that it doesn’t understand the ‘/24’ part (in this example). Instead, what you first need to do is to find the subnet ID which is associated with your address. If you’re unsure what this is then I’d recommend you use something like http://www.subnet-calculator.com/cidr.php:
In the example above we have a CIDR range of 10.13.160.0/23. You can see from the calculator that this produces a network ID of 10.13.160.0. Therefore the correct syntax for New-CMBoundary is as follows:
New-CMBoundary -DisplayName “IPSubNetBoundary01” -BoundaryType IPSubNet -Value “10.13.160.0”
All that said, I’d personally avoid IPSubnet boundaries completely. Instead, just go for a range, far simpler to understand. The CIDR calculator is again helpful for this and displays the range at the bottom.
Thankfully, the IPRange syntax is correct so you shouldn’t have any issues.
I recently upgraded to ConfigMgr 2012 R2 SP1 for a client and started seeing the above error appearing in both the WCM.log file and the WSUSCTRL.log files. I could find nothing on the internet about this so eventually I bit the bullet and phoned Microsoft support. Synopsis below:
Failed to create assembly name object for Microsoft.UpdateServices.
In our Environment:
WSUS server is on a separate VM to the CM2012 Site Server and WSUS database (SQL 2012) is on a separate VM from both CM2012 DB and from the WSUS server.
On WSUS Server we periodically see the following message in the D:\SMS\Logs\WSUSCtrl.log:
Failed to create assembly name object for Microsoft.UpdateServices.
This message appears approximately every 30 minutes in the D:\SMS\Logs\WSUSCtrl.log and it also appears every hour in “D:\Program Files\Microsoft Configuration Manager\Logs\WCM.log“.
As per Logs WCM.Log 7/23/2015 3:01:35 PM Checking for supported version of WSUS (min WSUS 3.0 SP2 + KB2720211 + KB2734608) SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Checking runtime v2.0.50727... SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Failed to create assembly name object for Microsoft.UpdateServices.
Administration. Error = 0x80131701. SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Checking runtime v4.0.30319... SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Found supported assembly Microsoft.UpdateServices. Administration version 3.1.6001.1, file version 3.1.7600.256 SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Found supported assembly Microsoft.UpdateServices. BaseApi version 3.1.6001.1, file version 3.1.7600.256 SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138) 7/23/2015 3:01:35 PM Supported WSUS version found SMS_WSUS_CONFIGURATION_MANAGER 4408 (0x1138)
WSUSCTRL.Log 7/23/2015 3:01:36 PM WSUS registry key change notification triggered. SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Checking for supported version of WSUS (min WSUS 3.0 SP2 + KB2720211 + KB2734608) SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Checking runtime v2.0.50727... SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Failed to create assembly name object for Microsoft.UpdateServices.
Administration. Error = 0x80131701. SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Checking runtime v4.0.30319... SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Found supported assembly Microsoft.UpdateServices. Administration version 3.1.6001.1, file version 3.1.7600.256 SMS_WSUS_CONTROL_MANAGER 420 (0x01A4) 7/23/2015 3:01:36 PM Found supported assembly Microsoft.UpdateServices. BaseApi version 3.1.6001.1, file version 3.1.7600.256 SMS_WSUS_CONTROL_MANAGER 420 (0x01A4)
It used to work in CM12 R2 Issue started post upgrading to CM12 R2 SP1.
The bad news is that unfortunately, there isn’t one. The good news is that this error doesn’t (or shouldn’t) actually affect anything and you can safely leave it logging every 30 mins to an hour (depending on the log you look at).
The official line on this is that it is a known bug in the code and will not be fixed until the next release of ConfigManger (2016), so no hotfix planned. I’m afraid we’ll just have to live with it.
I was recently contracted to do a project for a large retailer which involved upgrading all their POS devices to POSReady 2009 (I know this is 2013, don’t get me started on that one!). Anyway everything was going swimmingly until I started on the scale devices. In fact I hit a number of issues here which made things very difficult for deployment via SCCM, some of which may be addressed in further posts but I’ll try to keep things simple and just concentrate on one issue per article.
By and large, creating new task sequences and MDT roles for new devices is relatively formulaic and with a few small updates to the role and the task sequence OS image my new scale build was pretty much ready to rock ‘n’ roll. The MDT role had been selected and SCCM was figuring out what it had to do but instead of the machine progressing and starting the task sequence, it would simply reboot. Now I’m not foolish enough to think things will work like clockwork first time round, but after retrying it this began to irritate me. Looking at the local BDD.log file it said:
“There is no task sequence available for this computer”.
I went back to check the MDT role and make sure I’d put in the correct advertisement ID, task sequence (program) ID, etc and it was all bang on the money so what was wrong?
Well the good news for those of you reading this looking for the answer is that I have two solutions to the problem. The bad news is that you’d probably rather not do any of them. Unfortunately you have no choice.
This happens because SCCM uses the SMBIOS GUID as the primary means to determining device uniqueness, over and above the MAC address. Network cards can after all be swapped around fairly easily. The SMBIOS GUID is determined by the hardware GUID of the device and, unfortunately there are plenty of OEMs out there that don’t care much for their hardware GUID actually being unique (bit of a contradiction in terms, I know). In my case the GUID was basically a bunch of zeros followed by 0807060504030201. Now, if I had no other devices on the network with this GUID, then the task sequence would have run fine. However if you do have other devices out there with that GUID and they are not in the collection you have advertised the task sequence to and SCCM picks them before it picks the device you wish to build, then it quite rightly says there is no task sequence available for the computer. This isn’t a problem with SCCM, this really is an OEM problem. But that doesn’t help you right now.
I’ll start with the runaway recommended fix:
1. Contact the OEM, ask them to provide you with a BIOS update that will update the GUID appropriately and roll it out to all affected devices. I appreciate this is time consuming and costly but it really is the best way of doing things.
2. There is a back-door workaround which involves updating a couple of stored procedures in the SCCM database that will make SCCM look at the MAC address instead of the SMBIOS GUID. Although Microsoft will unofficially give you this solution, they won’t support it if everything goes tits up. I have yet to hear of anyone who has had an issue but depending on your environment, it could be a problem. It’s your call. As such it is best not to test it first in your production environment if possible. And make sure you have logged the requisite change to cover yourself!
That said, the update is fairly simple.
On the SQL server hosting the ConfigMgr database, open SQL Server Management Studio and modify the following stored procedures:
• Change the line
On xref.machineid = aux.itemkey and aux.smbios_guid0 = @smsbios_guid
On xref.machineid = aux.itemkey and aux.smbios_guid0 = @smsbios_guid + ‘.’
• Change the line
Where (m.smbios_guid0 = @vchsmbiosid) and (isnull(m.obsolete0,0) != 1)
Where (m.smbios_guid0 = @vchsmbiosid + ‘.’) and (isnull(m.obsolete0,0) != 1)
Change the line
Where (upper(m.smbios_guid0) = upper(@vchsmbiosid)) and (isnull(m.obsolete0,0) != 1)
Where (upper(m.smbios_guid0) = upper(@vchsmbiosid + ‘.’)) and (isnull(m.obsolete0,0) != 1)
As I said this not officially supported by Microsoft but the solution did come from them, so make of that what you will. Equally I take no responsibility either, try this at your own risk. The recommended fix is no. 1.
A key requistite in running a successful home-lab setup is to get at least a couple of separate networks connected together. This will create a more realistic ‘work-like’ environment to test any server configuation you have.
Personally I have a bit of a mish-mash of routers and switches to do this with and although I knew the theory of how this should work, I did struggle slightly getting the Thomson to forward packets like it should do. I put this down to the version I have (v8) and any documentation I could find was all v.7.
My other router is a MikroTik RouterBoard and I highly recommend investing in one of these if you’re serious about routing. They’re cheap (sub £50) but offer everything you’d expect on a professional router. Everything. In all honesty, most of the stuff this router will do I’ll never need nor indeed, understand. But if you’re looking for another router for this kind of lab setup, I can’t recommend it highly enough.
Anyway, my goal was to set up my network similar to that below:
Having plugged the WAN interface of my routerboard into the Thomson, I was immediately able to access the 192.168.0.x network and hence, the internet. Great. However going back the other way wasn’t so pretty. “hmmm…looks like I need a static route set on the Thomson router” I thought. However there appeared to be no way of configuring this from the router interface. After a bit of digging, I discovered I was able to telnet to the router and set it that way. I found this link which led me in the right direction:
However I was still unable to contact the 88.x network from the 0.x network. Turns out the problem I was having was that the gateway I added was actually incorrect:
ip rtadd dst=192.168.88.0/24 gateway=192.168.88.1 intf=LocalNetwork srcintf=LocalNetwork static=enabled
So what had I done wrong? Well the gateway should actually be the interface on the 0.x network. What was happening is that the Thomson router was providing a dynamic address to the WAN port of the routerboard, and it was this address I should have been providing as the gateway address. In my case this was 192.168.0.64, so the command line should actually have been the following:
ip rtadd dst=192.168.88.0/24 gateway=192.168.0.64 intf=LocalNetwork srcintf=LocalNetwork static=enabled
Obviously you should really make sure this address is static but this address won’t change in my configuration so it’s fine. My routing problem was then solved.
Some resources that helped me solve this problem (and one or two others)
List of telnet commands on Thomson Router
Set up DHCP configuration via telnet on a Thomson Router (not directly relevant, but helped me out elsewhere)
There could of course be plenty of reasons you’re experiencing this behaviour but I’ll zoom in on two which I’ve noticed are particularly prevalent with Windows 7 and which may not be immediately obvious.
Server Message Block Signing
The first time I saw this behaviour, it was after I’d introduced a hardened-build policy (yes the very same Win7-EC-Desktop baseline policy I refer to in an earlier post). The problem is to do with the Server Message Block Signing policy. There are two fixes – either update the servers holding your DFS shares to acknowledge the policy so Windows 7 can talk to them securely or simply disable the policy on the Windows 7 clients. The first is preferable, the second solution is easier, your choice.
Change group policy to reflect the settings in the image and you should be fine. Obviously this is the easier, second option!
Assuming this has been ruled out, the second, possibly less obvious reason for failure to connect to DFS shares is user token size. For those unfamiliar with token size, it basically relates to the size of the security ‘token’ you are issued with when you log onto a computer in a domain. This token is responsible for security assessment throughout your logon session. As more groups are added to your user account the token size increases. If you were to add a group that has another hundred groups nested in it, then your token size will increase accordingly!
Although this was a problem with Windows XP, it was generally more forgiving than in Windows 7/8 and for this reason you are more likely to notice the problem if you have recently upgraded. To confirm the issue, you are likely to see a message in the ‘system’ log in event viewer like the one below:
Again there are two choices you can make here. Firstly, you can increase the token size available for domain accounts in group policy. This, in my opinion, is the sensible way to fix the issue if you aren’t able to do much about the group sizes. The second, easier option is to remove groups you can do without until the token size is back to a respectable level. If this proves to be a slightly easier choice for you then you might require multiple accounts, eg one with admin groups to do privileged operations and another which lets you do day to day stuff. You can always use the run as… feature to save you logging off/on all the time.
OK so I decided to take a look at Windows Server 2012 and obviously the first thing you want to do after installation is to add AD DS. This is a pretty straightforward operation, not a whole lot different to how it’s done in Windows 2008. Simply select roles and features, add AD DS and follow the wizard through, creating the new forest, domain and DNS on the way.
However this is where I started to see some problems. I can’t say for sure whether I am the only one to experience this problem but as soon as the installation finished I was unable to open any AD tools and I was presented with an error similar to the following:
“The Specified Domain Does Not Exist or Could not be Contacted”.
My initial reaction after some minor troubleshooting was simply to remove and re-install it, after all this was a brand new installation and I really shouldn’t be seeing these issues on such a virgin server. I removed and reinstalled twice but still saw the same issue. Now as much as I love a challenge, I’ve been a busy man recently with one thing and another and finally this was an opportunity to use one of my support calls up that came with my Technet subscription. Last year I didn’t use any and it annoyed me I couldn’t carry them over. But I digress. I will therefore blog their analysis which finally fixed the issue. Apologies for the slightly iffy grammar, I’ve tried to fix the worst of it! I am still at a loss why I have experienced this issue, I’d be interested to hear if anyone else has had similar issues with server 2012.
=> On server 2012 made sure we have default shares “Sysvol & Netlogon” available as shared. This wasn’t the case.
=> With server 2012(a New Forest & Domain) the Sysvol replication is done by DFSR, thus checked the events in DFSR for any errors/warnings. We found the following :-
“The DFS Replication service stopped replication on volume C:. This occurs when a DFSR JET database is not shut down cleanly and Auto Recovery is disabled”
=> We ran the following Command in order to recover from the Issue, and resume the replication :-
“wmic /namespace:\\root\microsoftdfs path dfsrVolumeConfig where volumeGuid=”69718B5D-152F-11E2-93E8-806E6F6E6963″ call ResumeReplication”
(Note: enter appropriate GUID)
=> Since server 2012 is the only server part of Forest/Domain, we followed the KB :- 2218556 (http://support.microsoft.com/kb/2218556) and made the server authoritative.
=> Once done we had the “Sysvol & Netlogon” available as Shared, and were able to open the AD Snap-Ins and DNS was up & running fine.
With Server 2012 being the first DC in domain, the Sysvol replication is done using DFSR and detailed analysis of the DFSR logs yesterday revealed the DFSR service itself was in a state of error. We were getting an event id 2213 wherein the DFSR Jet database was not shutdown properly. This happens every time the service is stopped. Moreover by default auto recovery is not enabled for DFSR and without this starting and running correctly, he Sysvol & Netlogon shares weren’t getting shared. When we found the event we first fixed the DFSR database for Sysvol as per the KB and then resumed the replication using WMI command. This made sure the database is set correctly and the replication is underway. Sysvol and Netlogon were then shared.
Unable to Open AD Snap-Ins and Getting The Specified Domain Does Not Exist or Could not be Contacted
=> We promoted a server to a domain controller running 2012, the dcpromo process went fine without any issues, however when the server was restarted we couldn’t open any of the AD Snap-Ins and not DNS
=> The DNS zones failed to load and the snap-ins error out with “The Specified Domain Does not exist or could not be contacted“
=> When checked the serer we found the Default Shares of “Sysvol & Netlogon” missing
=> The server is running 2012 so the Sysvol replication would be on DFSR, thus we need to make the server authoritative for the same
=> We first made sure we have policies and scripts created fine, once done we followed the steps in KB 2218556 (The steps were followed in LDP as ADSIEDIT would not load the partitions)
=> Once we were able to perform the steps we checked the events, were getting the event for DFSR Database issues, also checked the Debug Logs in Detail, and ran the following :-
wmic /namespace:\\root\microsoftdfs path dfsrVolumeConfig where volumeGuid=”69718B5D-152F-11E2-93E8-806E6F6E6963″ call ResumeReplication
=> We got the event id :- 4602 and the sysvol , netlogon got shared fine
RELATED KNOWLEDGSE BASE ARTICLES :-
I was recently asked to come up with a solution to update our entire print server esate (around 600 boxes) with univeral PCL6 and PS drivers over x86 and x64 architectures. Specifically in this case I was asked to update with Lexmark and HP drivers although I see no reason this solution shouldn’t extend to any other manufacturers.
Since we have SCCM, my immediate idea was to create a package and push all packages out to the servers concerned but I found some shortcomings when creating the driver packages. Lexmark provide a tool to create the packages but I found I was unable to install both x86 and x64 drivers on the same machine. In fact, the executable created for the x86 drivers just threw errors. For the HP drivers, I had to throw in a few command line switches but again it simply wouldn’t install both architectures of driver. I should point out that the print server I was using for the test was Windows 2008 x64.
So another method was required. After much testing, I found the most reliable method was to use the Microsoft utility, prndrvr.vbs. This utility (along with several others) can be found under C:\Windows\System32\Printing_Admin_Scripts\en-US on the print server. I then deployed these using a task sequence in SCCM, running this utility with the drivers in question. Everything installs just fine and I don’t get any of the other rubbish which the manufacturer utilities include, such as extra printer objects, pre-configured ports etc. I have outlined what I did below for the HP drivers but remember, this should work for any print drivers.
- Download HP Universal Drivers package from the net. Create a folder called HPUniversalDrivers in C:\Temp on your test workstation. Create another folder inside called HPUPD. Copy xcopy.exe into this folder. Copy the 32bit and the 64bit folders out of the HP Universal Drivers package into C:\Temp\HPUniversalDrivers\HPUPD. Finally copy prndrvr.vbs into C:\Temp\HPUniversalDrivers\HPUPD. Your HPUPD folder should now contain two folders (32bit and 64bit) and the vbs script.
- Create a package in SCCM pointing the source to the location we have just described on your test box, eg \\testbox\C$\temp\HPUniversalDrivers, create a distribution point and update it. You don’t need a program.
- Create a new custom task sequence. Select General > Run Command Line: NAME=Copy HP Drivers Locally, Command Line=xcopy “HPUPD” “C:\Temp\HPUPD\” /E /Y, select the ‘package’ checkbox and browse to your HP driver package you created earlier. OK
- Decide which driver version you require. You may have to dig about in the .inf files to find the one for your particular requirements but for the purposes of this blog I required v5.4 which could be found in hpcu118d.inf:
- Create a new custom task sequence. Select General > Run Command Line: NAME=HP UPD PS v5.4 x86, Command Line=cscript prndrvr.vbs -a -m “HP Universal Printing PS (v5.4)” -e “Windows NT x86” -i “C:\Temp\HPUPD\32bit\PS\hpcu118d.inf” -h “C:\Temp\HPUPD\32bit\PS”, Start in:= C:\Temp\HPUPD
- Create a new custom task sequence. Select General > Run Command Line: NAME=HP UPD PCL6 v5.4 x86, Command Line=cscript prndrvr.vbs -a -m “HP Universal Printing PCL 6 (v5.4)” -e “Windows NT x86” -i “C:\Temp\HPUPD\32bit\PCL6\hpcu118c.inf” -h “C:\Temp\HPUPD\32bit\PCL6”, Start in:= C:\Temp\HPUPD
- Create a new custom task sequence. Select General > Run Command Line: NAME=HP UPD PS v5.4 x64, Command Line=cscript prndrvr.vbs -a -m “HP Universal Printing PS (v5.4)” -e “Windows x64” -i “C:\Temp\HPUPD\64bit\PS\hpcu118v.inf” -h “C:\Temp\HPUPD\64bit\PS”, Start in:= C:\Temp\HPUPD
- Create a new custom task sequence. Select General > Run Command Line: NAME=HP UPD PS v5.4 x64, Command Line=cscript prndrvr.vbs -a -m “HP Universal Printing PCL 6 (v5.4)” -e “Windows x64” -i “C:\Temp\HPUPD\64bit\PCL6\hpcu118u.inf” -h “C:\Temp\HPUPD\64bit\PCL6”, Start in:= C:\Temp\HPUPD
- Create a new custom task sequence. Select General > Run Command Line: NAME=Remove HP Directory, Command Line=cmd.exe /c RD C:\Temp\HPUPD /S /Q
- Your task sequence is now complete. Next, create a collection and add the print server of your choice to the collection. Advertise the task sequence to the collection and you should be good to go. This example will install both x86 and x64 drivers, PS and PCL6.
Disclaimer: I am in no way responsible if you accidentally deploy something nasty to your whole estate. Please test responsibly before deployment!