June, 2011

This article is cross posted from my original post on PTS Official Blog

See the first two parts of this series here:
Private Cloud Management with VMM 2012 (Part 1): What's new with System Center Virtual Machine Manager 2012
Private Cloud Management with VMM 2012 (Part 2): Creation of a Hyper-V Cluster Using VMM 2012

In my last blog post, we saw how to create a Hyper V Cluster using the VMM 2012 cluster creation wizard as a part of preparing a fabric for the cloud.

Virtual Machine Manager 2012 also provides a feature by which you can manage updates for your virtual machine hosts, library servers, PXE servers, the Windows Server Update Management (WSUS) server, and the VMM server itself in the VMM console.

When you perform update remediation on a host cluster, VMM orchestrates the updates, in turn placing each cluster node in maintenance mode, migrating virtual machines off the host by using intelligent placement, and then installing the updates. If the cluster supports live migration of Windows Server-based virtual machines, live migration is used. If the cluster does not support live migration, VMM saves state for the virtual machines and does not migrate them.

To manage updates in VMM 2012, you need a WSUS server

Here are the prerequisites:

1. You must install the 64-bit version of Windows Server Update Server (WSUS) 3.0 Service Pack 2 (SP2).
2. VMM requires a single, dedicated WSUS root server; downstream servers are not supported.
3. If you install WSUS on a remote server, you must install a WSUS Administrator Console on the VMM management server and then restart the VMM service.
4. Before you install the WSUS server ensure that the server meets all WSUS prerequisites described on the Windows Server Update Services 3.0 SP2 download page

Once you have configured the WSUS server, lets add the WSUS server in VMM 2012.

1. Open the Fabric workspace.

2. On the Home tab, in the Add group, click Add Resources, and then click Update Server.

3. My WSUS server name is sccm.dd181028.com, hence I have put SCCM as Computer name with port number and credentials.

4. Once the WSUS server is added, click on Synchronize to sync the WSUS updates with VMM.

5. You will all the update metadata showing on your VMM

Create Baseline

After you enable update management in VMM, you are ready to prepare for patching by configuring update baselines. An update baseline contains a set of required updates. During a compliance scan, computers that are assigned to a baseline are graded for compliance to their assigned baselines. After a computer is found noncompliant, an administrator brings the computer into compliance through update remediation.

6. Provide a Name and Description for your Baseline

7. Add the updates to your baseline against which your Hyper V hosts will be compared.

8. Assign the Baseline to the host groups.

9. Finish the wizard

10. You’ll see your newly created Baseline

To find out the compliance status for each baseline, you must scan the computer for compliance. When a computer is scanned for compliance, WSUS checks each update in the assigned update baselines to determine whether the update is applicable and, if the update is applicable, whether the update has been installed.

After a compliance scan, each update has a compliance status of Compliant, NonCompliant, Error, or Unknown.

1. In Compliance view of the Fabric workspace, select the computers that you want to scan.

2. On the Home tab, in the Compliance group, click Scan.

While the scan is in progress, the compliance status changes to Unknown. After the compliance scan completes, the computer's compliance status of each update is Compliant, NonCompliant, or Error.

Perform Update Remediations

To perform update remediations, the target computers must be noncompliant. To make a compliant computer noncompliant, you might need to use Add and Remove Programs to temporarily uninstall one or more of the updates listed in Compliance view.

On the Home tab, in the Compliance group, click Remediate. (The Remediate task is only available when the selected objects are noncompliant.)

If you select the host cluster by its cluster name, VMM orchestrates remediations of the hosts in the cluster.

VMM rolls through the host cluster, remediating one cluster node at a time. If a cluster node is compliant, VMM bypasses that node.

Before VMM begins remediating a host, it places the host in maintenance mode and migrates all virtual machines to other hosts in the cluster. If the cluster supports live migration, live migrations are performed. If the cluster does not support live migration, VMM saves state before migrating virtual machines.

If you prefer to restart the computers manually after remediation completes if any updates require a restart, select the Do not restart the servers after remediation check box.

Once the remediation is over, you’ll see your Hyper V cluster hosts as Compliant.

Hope this post was helpful for some of you who are evaluating VMM 2012 its update management feature.

In my next post, we’ll configure the rest of the fabric components like Logical Network, Storage and Load Balancers.

Stay tuned.

Iftekhar

This article is cross posted from my original post on PTS Official Blog

See part one of this series here:                                                                                                                                                                                                                                                                                                     Private Cloud Management with VMM 2012 (Part 1): What's new with System Center Virtual Machine Manager 2012

Before you start deploying cloud and services, we need to first start prepare the fabric.

A cloud can also host HA applications and services and we need to setup a Hyper V cluster to enable that. VMM 2012 has now introduced inbuilt capability of deploying Hyper V Clusters.

Fabric Prerequisites

Make sure that fabric configuration meets the following prerequisites:

• To use shared storage that is under VMM management, storage must already be discovered and classified in the Fabric workspace of the VMM Administrator Console. Additionally, logical units that you want to use as shared storage must be created and assigned to the host group where the Hyper-V hosts are located.
• To use shared storage that is not under VMM management, disks must be available to all nodes in the cluster before you can add them. Therefore, you must provision one or more logical units to all hosts that you want to cluster, and mount and format the storage disks on one of the hosts.
• One or more logical networks that are common across all of the Hyper-V hosts that you want to cluster must be configured in the Fabric workspace of the VMM Administrator Console. Additionally, the logical networks must be associated with physical network adapters on each Hyper-V host. Be aware that you do not have to create external virtual networks on the Hyper-V hosts. External virtual networks are automatically created when you run the Create Cluster Wizard.
• Although highly recommended, having a common logical network is not required to complete the wizard. You can configure virtual network settings for the cluster later. To do this, make sure that the logical networks that you want to use are associated with physical network adapters on each Hyper-V host. Then, in the Fabric workspace, right-click the host cluster, and then click Properties. In the ClusterName Properties dialog box, click the Virtual Networks tab, and then click Create to add the external virtual network. If there is a common logical network, the external virtual network will be automatically created on the cluster nodes.
• Let’s go through step by step deploying Hyper V cluster using VMM 2012.

1. Before you deploy Hyper V cluster, you need to add the hosts in the VMM.

Also if you are not using automated storage unmasking, you need to unmask the shared storage on these hosts.

2. Go to Fabric tab and click on Create Clusters

3. Specify the cluster name and provide your domain credentials which has rights to create clusters.

4. Select the host group where your standalone Hyper V hosts are located.

5. Select the hosts which you would like to be part of the cluster and click ADD

6. Ensure that the Network cards are picked up properly.

7. Provide a cluster IP address to the respective network card.

8. Wizard should automatically detect all the unmasked LUNs on the hosts and it will give you options to format the drives as NTFS or leave it unformatted.

9. You can also enable Cluster Shared Volume from here itself.

10. Wizard will automatically configure the volume with lowest capacity as Witness disk.

11. Now you can configure Virtual Network on all the hosts in a cluster from this wizard itself, In order to do that you need to configure Logical Network on the physical network card.

12. This way you can ensure that all the hosts has an identical Virtual Network to support HA VMs

13. Now on the summary page, click on the FINISH button and your cluster is on its way.

During the cluster creation process, VMM does the following:

• Validates that all hosts meet the prerequisites, such as required operating system and domain membership
• Enables the Failover Clustering feature on each host
• Unmasks the selected storage logical units to each host
• Creates the configured external virtual networks
• Runs the cluster validation process
• Creates the cluster with quorum and enables Clustered Shared Volumes (CSV)
• For each logical unit that is designated as a CSV, assigns the logical unit as a CSV on the cluster.

14. Enabling Cluster Shared Volumes

Now your Hyper V cluster is deployed and you are ready to deploy HA VMs on it.

You can also destroy the cluster using VMM console.

If you choose to destroy cluster using VMM 2012, you need to go to your Active Directory and delete the Cluster object manually.

That’s pretty much for today. In my next post, we’ll talk about update management and how VMM 2012 helps patching clustered nodes in an orchestrated manner.

Stay tuned.

Cheers

Iftekhar

This article is cross posted from my original post on PTS Official Blog

I have been delivering a lot of sessions on the all-new VMM 2012 lately to both external and internal audiences, so I thought of sharing some information with our online community as well.

I am putting together a 3-part series of blog posts to cover all the aspects of VMM 2012 and how it enables Private Cloud Management. If you are new to Private Cloud concepts and its benefits, you may want to read some of my earlier posts on the same topic:

Private Cloud– Part 1

Private Cloud– Part 2

In March at the Microsoft Management Summit 2011 held in Las Vegas, we announced the public availability of VMM 2012 Beta. I know most of you are evaluating VMM 2012 in your environment and may need some guidance—hence this blog series.

When we talk about Private Cloud, we expect following benefits:

1. Reliability and predictability: We would want our application to be highly available and fault-tolerant, the remediation from failure should be simple, and it should be simple to re-start and re-deploy applications.

2. Highly automated: We want our infrastructure to be highly automated to complete tasks like provisioning new systems and configuring storage, networking, and applications on its own without any manual intervention.

3. Agility of speed: We would also expect better responsiveness and turnaround time to meet user needs. We want applications to be deployed faster, and changes to be made much easier and faster—not taking weeks to happen.

4. Focus of Applications and services: Let’s face it, at the end of the day the consumer only cares about applications and services, uptime, and quality of the service as a measure for success of your private cloud. So as a cloud consumer I would want to focus on the application rather than underlying infrastructure.

Today let’s go through a high-level overview of VMM 2012 and what kind of advanced management capabilities it brings to the table to for virtualized infrastructures and the Private Cloud.

Microsoft has made investments in 4 different areas in VMM 2012:

1. Deployment

• HA VMM Server: VMM 2012 brings some enhanced deployment scenarios, where your VMM server itself is not cluster-aware, so now you can host VMM on a cluster and use the cluster name to connect your VMM console to the server.
• Non-HA VMs on HA Host: Now you can create non-HA VMs on clustered hosts.
• Cluster Creation: Now you can create Hyper-V® Host clusters directly from the VMM console itself, without having to go ahead and create clusters using Cluster Manager on the hosts. VMM Cluster creation also allows automated provisioning of storage and creation of CSV and virtual networks. You can even add or remove nodes, clusters, disks, and virtual network from existing clusters using VMM. You can also create clusters in untrusted domains.
  • MM now brings new custom properties not just for hosts, but also for components in VMM which can be really useful from a management perspective.
  • VMM also brings Windows PowerShell® 2.0 with new cmdlets to enable automation.

2. Fabric Management

• Bare-metal provisioning: VMM now allows provisioning of bare-metal Hyper-V hosts. VMM can deploy the OS on the host, enable the Hyper-V role, and add it to the cluster and deploy VMs or Service—so the entire VM host lifecycle is now managed by the VMM itself… Isn't that cool? :)
• Heterogeneous Virtualization Management: In addition to supporting VMware, VMM 2012 now also provides support to manage Citrix XenServer.
• Logical Network: VMM now can help define Logical Networks using VLANs and Subnets by datacenter location, and also provides feature to manage static IPs, load-balancer VIPs, and MAC addresses. You can now automatically provision a load-balancer.
• Storage Management: VMM now also connects and queries your storage boxes with SMI-S provided by storage vendors. It can discover storage arrays and pools, and discover capabilities and capacity of storage devices. It can now automatically configure and assign LUNs to hosts and clusters.
• Updates Management: VMM now offers support for orchestrating the OS update process of clustered hosts. It integrates with WSUS and can now automatically update your cluster nodes without any service disruption to your VMs.
• Dynamic Optimization: If you have worked with VMM 2008 R2, you know the PRO feature where you can move VMs around from one node to another within a cluster if there is any resource crunch on the host. It used to require Ops Manager integration, but VMM 2012 now provides built-in support for load optimization for Dynamic Optimization, and no longer requires Ops Manager. For other PRO capabilities, Ops Manager 2007 R2 and above are required.
• Power Optimization: With Dynamic Optimization, VMM also offers Power Optimization. With this feature you can now consolidate VMs on fewer hosts and shut down the rest of the hosts in a cluster in order to save power—powering them up when the load on these VMs increases and requires spreading out.
• Enhanced Placement: VMM now does over 100 placement checks before placing the VMs on the hosts. It also supports custom placement rules, e.g. when you want to deploy these VMs on the hosts with custom values of Server Room Third Floor. Now you can deploy multiple VMs as a part of a multi-tier service (e g: front-end web server, middle tier, and database tier).

3. Cloud Management: In VMM 2012, a private cloud provides the following benefits:

• Self-service. Administrators can delegate management and usage of the cloud while retaining the opaque usage model. Cloud users do not need to ask the cloud provider for administrative changes beyond increasing capacity and quotas as their needs change.
• Resource pooling. Through the cloud, administrators can collect and present an aggregate set of resources, such as storage and networking resources. Resource usage is limited by the capacity of the cloud and by user role quotas.
• Opacity. Cloud users have no knowledge of the underlying physical resources.
• Elasticity. Administrators can add resources to a cloud to increase the capacity.
• Optimization. Usage of the underlying resources is continually optimized without affecting the overall cloud user experience.

4. Services Management: It’s not VM Management anymore. Now with VMM you can deploy and manage multi-tier applications across multiple servers as a single unit. Scale out the service with minimal effort based on demand.

We'll discuss some of these great details in my new few blog posts and will also see how to deploy some of these. Stay tuned!

Cheers,

Iftekhar