Automated Fabric Patching for the Private Cloud
When speaking of private cloud, the fabric is arguably the most unique piece of the cloud solution that separates it from a traditional datacenter. We’ve talked about the private cloud fabric in previous articles on the Private Cloud Architecture blog, such as the one done by Yung Chou entitled System Center Virtual Machine Manager (VMM) 2012 as Private Cloud Enabler (2/5): Fabric, Oh, Fabric .
But another key feature of any private cloud solution is intense and pervasive automation. It would make little sense to have a highly automated private cloud deployment that doesn’t include automated fabric patching.
To help us with this, Michel Luescher shares his “Automated Fabric Patching” methodology that leverages System Center Virtual Machine Manager, System Center Operations Manager, System Center Orchestrator and Windows Hyper-V.
I think Michel did a great job with this article and it highlights some key capabilities and provides a framework for approaches to automated fabric patching in the future. If you have any questions on Michel’s approach, you are welcome to write to me at email@example.com and I will forward them to Michel.
Finally, are you interested in private cloud? Do you have some ideas, insights, experiences and triumphs with private cloud? Then why not share them with the Private Cloud Architecture Community and do a guest blog post? All you need to do is send the article to me at firstname.lastname@example.org and I will review the article, do some edits, maybe ask you some questions, then send it back to you for review. If you’re good with it, then I publish it to the blog! I’ll include your picture and any contact information you wish to include. And you’ll be “Internet famous” and you can show your Mom and Dad that you were published on the Private Cloud Architecture blog! It doesn’t get better than that . Thanks!
-- Tom Shinder, The Private Cloud Guy.
In a private cloud session that I presented in May at Microsoft TechDays 11, I presented together with Philipp Witschi a fully automated solution for fabric patching. Many of the attendees have contacted us after the event asking us to tell them more about the automated fabric patching solution. In this article I will share with you more information about our automatic fabric patching approach. There is also a video (in German) if you are you would like to see the original presentation.
The solution for “Automated Fabric Patching” were realized with the following products:
- System Center Virtual Machine Manager 2012 (Beta)
- System Center Operations Manager 2007 R2
- System Center Service Manager 2010
- System Center Opalis 6.3
- Windows Hyper-V
System Center Service Manager
We will use the management console in System Center Service Manager to begin the fabric patching process. To do this, you open a new change request in the Service Manager console and create an activity that is defined by the actions noted in the table below
|VMM Host Cluster||Name of host cluster that will be patched|
|VMM Server||VMM Management Server|
|KBArticle||To install Windows updates|
|VMM baseline||Used to update baseline in VMM|
|Hyper-V Operating System||Operating system used by Hyper-V hosts|
|Hyper-V architecture||Architecture of the Hyper-V hosts|
The completed fabric patching activity change request looks like what you see in the figure below:
Information such as the VMM server, the operating system and the architecture are provided so that this information can be used each patch Tuesday.
As soon as the change request is created, a workflow is started which carries out the following configuration actions:
- New change request/activity is created
- The status of change requests are set to “In Progress”
- The status of the activity is changed to ‘In Progress”
System Center Orchestrator (Opalis)
System Center Orchestrator now enters the picture. Orchestrator is the connector that connects all the applications used in this private cloud fabric patching solution. A dedicated runbook was created for automated fabric patching. This includes child workflows in the current version. Activities are checked and confirmed on a regular basis to confirm if a new patching activity exists. Once an activity has a status of “In Progress”, the following workflow will be started:
The first thing (2.3) that is done is to check that all members of the Hyper-V failover cluster “read out” and enabled in SCOM for maintenance mode of these objects. If necessary, a scheduled restart of the host will allow this without trigger an error or alarm.
PowerShell cmdlets are used by SCVMM for all the operations. The following PowerShell scripts are run from from the SCVMM server remotely to the Orchestrator server:
- Add updates to baseline
- Baseline add to host cluster
- Run a compliance scan
- Start remediation
- Start maintenance mode
- Move from HA VMs using live migration
- Break mode in failover cluster
- Install Windows updates
- Restart the computer
- Run a compliance scan
- Stop maintenance mode
- Start maintenance mode
The remediation process is designed so that all hosts are compliant with the update baseline and will continue until this is accomplished. Once the fabric patching is complete, the virtual machines that host the workloads will be distributed using Hyper-V’s dynamic resource optimization.
Who Designed the “Automated Fabric Patching” Approach?
The Automatic Fabric Patching method described in this article was designed by Michel Luescher (Microsoft Switzerland) , Philipp Witschi (itnetx gmbh) and Marcel Zehner (itnetx) to support a customer they had in Bern. As more System Center 2012 products (such as System Center Service Manager and System Center Operations Manager) become available, we will add new features surfaced by these applications to be used by Automated Fabric Patching method. As the workflow was designed with SCVMM 2012 Beta, we will upgrade this to RTM as well.
Consultant Private Cloud – System Center and Virtualization