In light of "Microsoft loves Linux" initiative, you can now deploy your Linux virtual machines by using templates in the System Center 2016: Virtual Machine Manager. As I was searching on how to do this (successfully), there were couple of articles that helped, so I’ve decided to do a short list of all the necessary steps (in one place).
Steps to make your Linux VM template deployments work:
- create a new (Generation 2) virtual machine (as you would normally do)
- install the Linux operating system in that virtual machine (as you would normally do)
- HINT: A list of supported Linux distributions and versions on Hyper-V is available here.
- install the Linux Integration Services (LIS) (as per this post):
- open the "modules" file
|
|
sudo nano /etc/initramfs-tools/modules |
- add the following to the end of this file:
|
|
hv_vmbus hv_storvsc hv_blkvsc hv_netvsc |
-
- save it (Ctrl+X and Y)
- install LIS and reboot the machine by using the following commands:
|
|
sudo apt-get install --install-recommends linux-tools-virtual-lts-xenial linux-cloud-tools-virtual-lts-xenial sudo update-initramfs -u sudo reboot |
- check if the services are running by using the command:
- install the Virtual Machine Manager agent (as per this post):
- share the folder C:\Program Files\Microsoft System Center 2016\Agents\Linux on your VMM machine
- copy the VMM agent files to Linux virtual machine
- as a real Windows admin, I did it through the GUI
- install the agent:
|
|
sudo ./install scvmmguestagent.1.0.2.1075.x64.tar |
- fix the boot for Generation 2 virtual machine (boot information is by default stored in the VM configuration file, not on disk – Ben wrote a great article on this “issue”)
- Ben’s way (didn’t work for me):
- change directory to the boot EFI directory
- copy the ubuntu directory in to a new directory named boot
- change directory to the newly created boot directory
- rename the shimx64.efi file
|
|
sudo mv shimx64.efi bootx64.efi |
- TriJetScud’s way in the comments (worked for me with Ubuntu 16.04 Generation 2 VM):
|
|
sudo grub-install --target=x86_64-efi --efi-directory=/boot/efi --no-nvram --removable |
- shutdown the virtual machine and copy its VHDX to the VMM Library
- HINT: Don't forget to refresh the VMM Library.
- go to the VMM Library, right-click the copied VHDX and select the Create VM template option
- proceed with creating the template as you normally would, to the part Configure Operating System
- HINT: If you are using Secure boot, don’t forget to select the MicrosoftUEFICertificateAuthority template in hardware settings.
- there, under Guest OS profile, you select the option to create a new Linux operating system customization settings
- next you specify your guest OS settings and finish creating the template
- now you can create a new Linux virtual machine from the template you’ve configured!
Hope it helps!
Cheers!
Not so long ago, there was a thread about this issue on TechNet Forums – long story short, in some cases (if you didn’t do a clean installation of System Center 2016 – Operations Manager, for example), the shiny, new feature called Updates and Recommendations didn’t work.
Even better – there was a rather cryptic error saying “An error occurred while displaying the Updates and Recommendations view. This might be because the database query has encountered an error…”.
So… it looks that maybe the database query has indeed “encountered an error”.
What can we do to make sure and resolve this?
As the user Chandra Bose suggested, we can look for duplicates in our imported management packs… and maybe we will be smarter then.
PowerShell command we can use:
|
|
Get-SCManagementPack | sort -Property Name | ft Name,Version -AutoSize |
This will list our imported management packs and their versions, and we can start looking for duplicate(s).
In my case, there were some – some of them were the two management packs called Microsoft.SystemCenter.WebApplicationSolutions.Library.Resources.*.
To get a better look on those two, we can use the following command:
|
|
Get-SCManagementPack | where {$_.Name -like "Microsoft.SystemCenter.WebApplicationSolutions.Library.Resources.*"} |
And the output looks like this:
This shows that we really have two “duplicate” management packs in our SCOM database, one installed in 2013, and another in 2014 (why? and how? don’t really matter 
). We need to remove one, obviously.
For that, we can use the following command (by using the Id property from previous command):
|
|
Get-SCManagementPack -Id "09988ce5-edea-8ed2-868a-0ecc1193338b" | Remove-SCManagementPack |
And, if there are no more duplicates, our Updates and Recommendations view should work now:
Hope this helps.
Cheers!
I’ve recently created a simple lab which gave me some answers around load balancing a SharePoint 2016 farm with SSL offloading.
To start, I’ve created a couple of virtual servers (on top of my “supercool home Windows Server 2016 Hyper-V PC” 
) – a domain controller, a SQL server and two SharePoint servers. I’ve also downloaded a KEMP LoadMaster appliance (there is also a free one here, which would have been just enough for this lab) and prepared my DigiCert wildcard certificate (there is no need for the wildcard option, but this is the one I already have, so I’ve decided to use it).
So… I’ve prepared a domain controller, joined all the other servers to the domain and then installed SQL Server 2016. After that, on SharePoint servers, I’ve ran a preparation wizard and created a new SharePoint farm from the first node… with second node joining to it later. At the end, I’ve done the “Farm configuration” wizard and was all set to do the load balancing part. (And yes – I know that clicking “Next” is lame, but… it works. )
The networking configuration for this lab is pretty simple. I have two VLANs – 111 (backend, where all the servers are residing) and VLAN 101 (frontend, where my LB virtual servers are).
I’ve created a new virtual machine for the load balancer, attached it to the two mentioned networks and also added the virtual disk downloaded from KEMP’s website.
After that, I’ve done the initial configuration wizard of LoadMaster which is actually straight-forward (setting the password, IP addresses, and importing a certificate afterwards).
With this done, we can create our virtual service(s) – there is actually a great guide for configuring the SharePoint load balancing virtual servers with KEMP LoadMaster.
I’ve used the following basic (manual) settings for my virtual service:
HINT: When troubleshooting load balancing – make sure that you have only one node behind the balancer… it makes things so much easier to troubleshoot!
One last thing that wasn’t working with this “Next, Next, Next…” configuration was the Alternate Access Mappings (AAM) part – to be able to access a SharePoint farm over HTTPS and a public name, AAM should “know about it”. There is a great guide about AAM available – make sure you read it.
Default AAM settings for my farm were:
After (a lot) of troubleshooting and research, they were changed to this:
And… that’s it – it works!
My totally awesome SharePoint 2016 site, located behind a load balancer and published with a trusted certificate (with SSL session terminating on my virtual KEMP load balancer), was alive:
To conclude - in all the configuration that was done, setting the AAM right was something that gave me most of the headache (load balancing/redirections not working right, troubleshooting what’s happening, etc.). Pay special attention to it! Once you figure it out, you’re done.
Cheers!
Another Advanced Technology Days are now behind us.
It was a great conference, with so many familiar faces (and so many new ones). I didn’t like this year’s venue as much as the last one (let’s be honest – nothing beats a movie theater! 
), but it was fine – it’s always about the great people, not the venue itself.
This year I was speaking about the new System Center 2016 suite and, judging by the full room (thank you!), this topic is very hot right now. I really like the new features in Virtual Machine Manager and the new “speedy” Operations Manager (with web portal which finally works in “modern” browsers ).
Slides (which are actually customized Microsoft Ignite slides – thank you), are uploaded to my SlideShare.
And now… looking forward to
Advanced Technology Days 13.
Cheers!
P.S. Thanks for the pics – Goran & Igor!
Have you ever reinstalled your Hyper-V hosts?
I know, there is not much need for it (as everything usually works just fine), but still… there is a “Windows Server 2016 re-installation wave” coming and maybe you’ll find the the next pieces interesting and useful.
One of the “messy” tasks with Windows reinstallation is networking… and by “messy” I mean “you have more than one network cable in your Hyper-V hosts” and you need to know which networks are connected where. 
What happens is that Windows somehow always forgets your network device order, all the pretty names you’ve applied and you get stuck with names like “Ethernet”, “Ethernet 2”, etc.
There is a way to fix this (there are many, actually) – we can ask our younger colleague to go to the server room and unplug the cables one by one and then plug them back in, following the rename on our (Windows) side. This way we are certain that all the corporate, DMZ, storage, live migration, etc. cables don’t get “confused” when added to their respective teams and if we labeled cables properly, everything will work with fresh Windows installation also.
But… there is another way. We can use what we already have – our documentation. Here I mean “our current setup” – we have our network adapters and teams already configured in our current Windows installation, why don’t we just export this info and use it after the reinstallation?
We can do this easily by using PowerShell!
The idea is to export network adapter names and MAC addresses of our physical network adapters (excluding the virtual and team adapters), in a CSV file, so that we can use it later, to rename our adapters after the reinstallation:
|
|
Get-NetAdapter | where {$_.InterfaceDescription -notlike "Microsoft*" -and $_.InterfaceDescription -notlike "Hyper-V*"} | Export-Csv C:\Hyper-V.csv -Delimiter ";" |
After reinstallation, we can use the following command to rename our adapters, as per our saved CSV file:
|
|
$networks = Import-Csv C:\Hyper-V.csv -Delimiter ";" foreach($network in $networks){ Get-NetAdapter | where {$_.InterfaceDescription -notlike "Microsoft*" -and $_.InterfaceDescription -notlike "Hyper-V*" -and $_.MacAddress -eq $network.MacAddress} | Rename-NetAdapter -NewName $network.Name } |
And voilà – our networks are named nicely again (and our colleague didn’t need to go to the cold server room… this time).
Cheers!
Don’t be sad! 
In case you’ve missed the Microsoft Ignite conference this year (like me ), all of its content is available online! You can visite the conference homepage (https://ignite.microsoft.com/) and catch video streams and…
… even better – there is a PowerShell script that can help you download the videos and slides of about 570 Ignite sessions!!! How cool is that???
Script was made by Michel de Rooij and Mattias Fors, and can be found here (read the “help” section to figure out what you need and how to use it).
When launched (properly), it looks something like this:
Happy downloading, watching and learning!
Cheers!
P.S. With the script alone, you’ll need the YouTube downloader utility available here (it actually says so in the script’s “help”, but who reads the documentation, right? ).
A couple of days ago, another achievement was unlocked – Microsoft rewarded me with another MVP award (7th in a row, actually 
), again for my contributions to the Cloud and Datacenter Management community.
Thank you, Microsoft! (and also a big “thank you” to all of you who were with me during this wonderful 7 years… thank you for all your support!)
And for the year ahead, I wish myself many “challenging” situations in the cloud and datacenter area… where so many interesting things take place.
Cheers!
As mentioned before, I was at the MobilityDay 2016 conference on Thursday, where I’ve had an honor to deliver a session about Citrix XenMobile & NetScaler solutions (yup, I know… it’s not Hyper-V… go figure… but my demo environment was still runnning on top of Hyper-V, so…).
Session was actually a short introduction to what Citrix XenMobile solution offers and how to set it all up, combined with NetScaler, for a complete management of mobile devices and mobile applications.
Conference was great, as always – so much old (and new) faces, with lots of fun. This time it was actually a 10th birthday of MobilityDay!
(photo by Romeo)
Happy birthday, MobilityDay! (and congrats to the organizers) 
Slides are available at my SlideShare:
Cheers!
It’s that time of the year again – conference time! 
First of the conferences I was at was the KulenDayz 2016 conference. I love KulenDayz! So many nice people, friends, colleagues, MVPs and kids gathering at awesome (and relaxed) sessions, enjoying wonderful nature (and food).
This year I’ve had the honor to speak about Hyper-V best practices – I’ve tried to gather the most of best practices around Hyper-V deployments that I’ve seen/learned/done over the years in an easy 50-minute session. Considering the reactions from the attendees and other speakers, it was a great session! (thank you all!)
As every one of us has hers/his best practices around Hyper-V, I’m expecting your comments, thoughts and corrections… we are all learning from each other!
Session slides have been uploaded to SlideShare:
Big thanks to the organizers for another great KulenDayz experience!!!
Next conference is Mobility Day 2016, which is in about two weeks. I’ll be speaking about something completely different – Citrix XenMobile and NetScaler. Fun times!
Cheers!
There was a barbecue today… on the roof… of the Microsoft office! How cool is that??? 
I’ve had an honor to join my colleagues and friends, and contribute to this great event with a presentation called “Nano Server – budućnost je tu!”, which was actually inspired by Jeffrey Snover’s Ignite speech about Nano Server – the new installation option in Windows Server 2016.
The whole Nano Server idea is very simple – let’s remove all the “unnecessary” roles and features (especially the famous GUI), and leave only the parts that we really need for whatever purpose (i.e. we’re getting the “purpose built servers”). All the purposes/roles that are supported in this preview version are Hyper-V, File Server and Failover Clustering, and they are working great on this “nano-sized” server operating system (even in this early preview).
I’m really excited about Nano Server, and I’m so glad that I’ve had a chance to talk about it at such great event. Because there were so many questions about it during the day/evening, I’ve noticed that Nano Server is a very cool subject for attendees also. Nice!
The presentation is available on SlideShare – check it, install the Nano Server and start rocking your datacenters!
Cheers!
