#Azure : Map your traditional datacenter compute with cloud VMs

Cloud has completely changed the IT architecture landscape. Since early days of IT till last decade, Architecture was an abstraction that used to play a key role at the time of transformation or new development. Once Architecture developed, it used to continue for many years with very minimal changes. Since cloud came in its inception, architecture has become a key in day to day work life of an IT professional because of its agility. If not daily then most probably weekly, you can observe some changes in the public cloud world and that needs to be taken care seriously.

In this post, I’ll try to simplify the cloud architecture for compute and will compare with traditional compute architecture. Apart from the simplification, I’ll provide you a logical design thinking approach that will make your life easy no matter what role you are playing in IT.

Let start from traditional datacenter.

If you are an experience IT professional, you must have seen or heard about these names at least once in your career.

Traditional type of Servers: Tower, Rack, and Blade servers.

A true traditional server that comes with multiple configuration options such as dual-processor or quad-processor etc.

New type of platforms: Converged and Hyper-Converged.

These new platforms are basically rack based servers that provide inbuilt advanced storage and networking capabilities by leveraging software defined data center technologies.

Virtualization: In last one decade, every organization has leveraged capabilities of virtualization that enable compute to run multiple virtual machines so that you can fully utilize your high-end servers and save cost in multiple aspects.

Now, let me explain complete compute story in public cloud such as Microsoft Azure.

When you look at compute available through cloud, you can easily determine that it is same kind of virtual machines, which we used to have in our virtualized environment. But in the cloud, the only difference is that you don’t worry about the underlying hypervisor and hardware that is being used behind the scene to provide you virtual machines.

In traditional datacenter, we use multiple racks to install different types of hardware and each rack connects with different power supply units through PDUs and these power supply units connect with main power supply unit. In many scenarios each rack deploys top-of-the-rack switches to provide network connectivity to the devices installed in the rack and in some cases one or two of the racks in the same row deploy these TOR switches. To overcome the challenge of entire datacenter failure, we use multiple datacenters in the form of high-availability and site-resiliency. When an administrator performs any maintenance activity in the traditional datacenter, he/she makes sure that the quorum is maintained while performing maintenance activity to avoid any kind of unexpected failures.

In cloud, H/W level high-availability is being provided by fault domain (unexpected) and maintenance level availability is being provided by update domain, and both features fit under one umbrella that i.e. known as availability sets. To provide high availability, Microsoft Azure uses multiple datacenters (at least two-three) in each region, and to support site resiliency Azure provides multiple region options in same geography or across multiple geographies.

I hope, now you will be able to sketch a clear picture in your mind about traditional datacenter vs cloud.

Now, let me help you with the logical design thinking approach. When you think to deploy a VM or set of VMs, follow the following steps in sequential order.

  1. Think about application and its big picture, keep end-users in your mind and their respective locations.
  2. Select the best suitable cloud region.
  3. Consider different tiering of solution.
  4. Consider security, high availability, site resiliency and load balancing requirements.
  5. Illustrate your network requirements.
  6. Illustrate your storage requirements.
  7. Illustrate your compute requirements.

Once documented all the above, create design diagram and find the approach to deploy your solution. For more details specific to Microsoft Azure compute, read the following blogpost.

#Azure : Virtual Machines

#Azure : Virtual Machine Configuration

#Azure : Virtual Machines High Availability

#Azure : Step-by-step Availability Sets

#Azure : Virtual Machines Scale Sets

#Azure : Large Virtual Machines Scale Sets


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s