SimpliVity: Do I Have To Choose To Scale My Infrastructure or Manage IT Infrastructure Lifecycle? No!

I originally wrote this post for but I also wanted to post it here for my normal readers.


One of the problems we have had when we design, implement, and manage our IT infrastructure is how to design for scale but at the same time design for ease of migration and infrastructure lifecycle.  In the past we have always had to design for one or the other because the infrastructure didn’t provide the means to do both.

With traditional converged infrastructure you have to scale multiple systems at a time. For example if you work with current reference architecture based building blocks (pods that flex and blocks that are virtual) and you run out of blade slots in a chassis you have to add more chassis’ and blades.  Depending on the business need you might even have to add storage shelves and drives and possibly some network equipment.  Because each of the systems within the so called “convereged” architecture are still logically and physically separate from each other you have to add them a piece at a time and configure them separately as well.  As you can imagine this can take a lot of time and resources to implement and manage not to mention take a lot of your annual IT budget.

Drilling a little deeper, let’s look at one of the most painful pieces of our infrastructure to add or remove, the storage.  While VMware has made this process much simpler than in previous years, moving tens or hundreds of terabytes of data is time consuming and must be planned extremely carefully if you don’t want to forget any data during the migration process.  We have to carve out the storage that we want to use for our existing virtual infrastructure.  We also need to ensure that backup jobs are now backing up the new datastores from the new array.  Between scheduling data migrations during periods of low I/O, pausing the nightly backup schedule, and trying very hard to not impact the business it can take weeks or months to complete a data migration project.

With the SimpliVity OmniCube and its OmniStack technology we have designed the solution, from the beginning, to solve these major IT problems.  First we allow you to grow as your IT infrastructure grows.  If you need more Compute or Storage simply add another OmniCube.  The way that is accomplished is almost like adding another host to vCenter.  First we add the OmniCube to vCenter the same way you would add any other host.  Next you click on the SimpliVity plugin tab and deploy the OmniCube.  The OmniCube joins the federation automatically and assimilates into all other aspects of the federation.  The new physical storage of the OmniCube becomes part of the physical space of the existing datastores.  You have more CPU, Memory, and Storage for your virtual infrastructure.  It’s as simple as that.  We never left vCenter to deploy the system or add it to the existing virtual infrastructure.  This can be done globally from anywhere in the world without logging into different systems constantly.  If you have a need to grow just compute or you would like to use a specific add on card (say for VDI) then you can just add those servers into the virtual infrastructure and we can present the OmniCube datastores to those new hosts.  All the great OmniCube features are made available to all VMs hosted on an OmniCube datastore.

Next let’s take a look at the example of removing some of those Compute and/or Storage resources from your infrastructure, specifically storage infrastructure.  As mentioned earlier these projects are very time consuming and costly.  With the OmniCube we’ve made it simpler for you by moving your data with a single click.  Once you’ve migrated your virtual machines to the surviving OmniCubes using a standard VMware vMotion, you right click on the OmniCube Virtual Controller you would like to remove and select “SimpliVity – Remove from Federation”.  After that, you go home and get some sleep.  The OmniCube will migrate all of your data to the surviving OmniCubes within the federation automatically without any impact to your production workloads.  All the backups for those virtual machines will move and since the backup policies are global we don’t have to worry about our virtual machines being unprotected after they have moved.

As you can see we’ve put the simple back into your IT infrastructure as well as managing your infrastructure lifecycle.  We allow you to design for scale while at the same time allowing you to plan your infrastructure lifecycle without impacting your business operations.  This will save you money in both your capex and opex budgets and allow you to utilize those assets where you choose.