New guide set: Microsoft IaaS Product Line Architecture Guides

 

Microsoft have informed us (MVP mail distribution) and also published two excellent guides about  their Infrastructure as a Service (Iaas) Product Line Architecture (PLA). The first guide is mostly based on Windows Server 2012 storage and virtualization and the second guide is based around System Center 2012. both document will provide you good idea about Microsoft approach. In future they expect to provide more documents like this and also expecting feedback from you.

  • Infrastructure-as-a-Service Product Line Architecture Fabric Architecture Guide (Windows Server 2012)

http://www.microsoft.com/en-us/download/details.aspx?id=38817

  • Infrastructure-as-a-Service Product Line Architecture Fabric Management Architecture Guide (System Center 2012)

http://www.microsoft.com/en-us/download/details.aspx?id=38813

Enjoy the reading during your weekend break Smile

Windows 2012 certification path

Are you new to IT career or someone already having experience trying to next step in certification? Well this visual steps will provide you an clear idea about the Windows 2012 certification plan. As a MCT I found this graphical representation of visual path to Windows 2012 is really useful to anyone who plan to achieve Windows 2012 certification.

2012-certification path

HYPER-V QOS

In this blog article I’m going show how we can enable virtual machine-level Quality of Service (QOS) in Windows 2012 server HYPER-V. QoS is a set of technology that provide you with the ability to cost-effectively manage network traffic and to enhance user experiences in enterprise environments, small offices, and even home networking environments. When we get 10G NICs in the server it is important we manage the given bandwidth to each VM base of their workload. There are various methods you can configure the QoS. In this article I’m taking into consideration Single HYPER-V host which is hosting several VM’s and how we can manage the allocated bandwidth to each VM separately. You can assign QoS to each VM or each virtual switch level as well.

In production environment it is advisable to apply the QoS in switch level rather than individual VM level. Belo diagram show such scenario,

 

But in my demo I’ll focus on the highlighted VM below,

VM-1

When right click and select settings we can see currently bandwidth management has not been setup yet,

VM-2

Now moving back to the PowerShell prompt on the host computer, I’m going to allocate 10 MBPS for this VM,

VM-3

Once that command completed you can copy a file from host computer to the guest VM and monitor the network speed. for the demonstration purpose I’ll copy large file from another guest to QoS enable VM,

VM-4

Due to my test servers performance issues (Sad smile) you’ll see the file copy speed is way below 10 MBPS. never the less if you go back to the PowerShell and disable the QoS, while the file copy process is going on you’ll find the copy process quicken up,

VM-5

VM-6

Now this is a very simple demo but the big picture I want to bring is QoS is important when you planning for server virtualization and when balancing the workload. With 10G cards becoming commodity this is becoming essential feature to enable and use. You can find more information about HYPER-V QoS here.

Hyper-V Offloaded Data Transfer (ODX)Overview

This is one of the unique feature in Windows Sever 2012 HYPER-V I would love to get my hands dirty. The idea behind is this option is to offload the data transfer commands (processing power) away from the HYPER-V HOST computers and instruct the Storage (SAN) to do that task. So basically copying data to memory transfer to the target location free the memory from data bits all these work as in the history. Virtual machine’s directly connected with FC storage or pass through disks can take leverage on this technology as long as the hardware vendor provide that facility in their storage solutions.

Right now VHD or VDHX connected to virtual IDE controllers doesn’t support this capability. The main reason is integrated development environment (IDE) doesn’t support the ODX feature.

You can review the below  video which demo the HP #PAR storage subsystem with Windows 2012 about the ODX feature.

Another video from Dell company

Another blog article released by Dell showcasing the numbers Smile Man…I’d be really happy if I can get my hands dirty on one of these storage systems.

Failover clustering improvements in Windows Server 2012

It’s been a while I post anything to the blog due to the new life challenges I faced Smile Anyway now I’m back and started the practice along with ITPro community session regarding the above subject. You can find the relevant slide deck in Slide share (http://www.slideshare.net/susanthasilva1/windows-server-2012-failover-clustering-improvements)

Looking back at the history and compare the clustering technology starting from Windows NT 4.0 Microsoft has done significant improvements on the technology. Instead of calling high availability now we try to achieve contiounsly availability (Thanks to storage, network improvements) for the mission critical apps through Windows clustering in Windows Server 2012. Apart from that now we have the capability to monitor the clustered services and applications inside the VM’s proactively. In case of a failure of these services we can restart or move to another node (pretty impressive feature where normally SCOM used to do this but now freely integrated with clustering)
Failover clusters also provide Cluster Shared Volume (CSV) (improved)functionality that provides a consistent, distributed namespace that clustered roles can use to access shared storage from all nodes. Apart from that the utilization of scale-out-fileserver and SMB 3.0 are some new add ins to improve the clustering experience and expandability apart from typical SAN environment.

Another new improvements is Cluster Aware Updating (CAU) which describes about the updating the cluster nodes with little intervention from Cluster administration. (Very much welcome feature if you’re handling large clusters)
To have this feature you need to add the particular role to the cluster nodes first. Once it carried out you’ll see the option in the Tools menu in Server Manager.

image 

Once you open the cluster MMC you can select the “Cluster aware Update” option which will open you another menu,

image

This menu will help you to verify the nodes met the required settings to eligible under CAU.

image

Another welcome improvements in the Cluster Shared Volumes (CSV). In 2008 R2 we have the CSV option but limited to use in HYPER-V scenarios only. With Windows 2012 we can extend this to file servers as well. Combine with Scale-out-fileserver scenario we can have HYPER-V cluster in affordable cost as well. In a nutshell CSVs allow multiple nodes in the cluster to simultaneously access the same NTFS file system without imposing hardware, file type, or directory structure restrictions. As shown in the following figure, with CSVs, multiple clustered virtual machines can use the same LUN and still live migrate or quick migrate from node to node independently. Think CSV as a file system which is referred as CSVFS these days Smile

Figure 3  CSV used by virtual machines on 3 nodes

Honestly there are too many information which cannot be written in one article like this. But you can refer to the TechNet article which describe all the new improvements in Windows 2012 Failover clustering (http://technet.microsoft.com/en-us/library/hh831414.aspx#CommunityContent)