Unlike OpEx, No Small Workarounds for the Depreciating Value of CapEx
How many times have we sat in CAPEX meetings astonished by the amount of money we spend on things such as hardware and perpetual licensing (read also: Balancing Budgets Part 1: How to Avoid Runaway Software Licensing), it seems to rise every year no matter what! For years the hardware resellers have been making a killing off of the IT industry, as have the manufacturers. Unfortunately CAPEX spends like hardware doesn’t have the small workaround that you can use with OPEX like open source software or things like that. How do you reduce what you spend on things such as physical hardware, the software that goes on that hardware, or the datacenter that the hardware sits in?
The worst part about CAPEX expenditures is that they are a depreciating investment. Just like buying a brand new vehicle, as soon as you drive off the lot the vehicle loses 20% of its value. The same thing happens with hardware and perpetual licenses. You have to map out years in advance to make sure you get your full investment out of your hardware before it becomes obsolete, and then start the cycle all over again. This is such a waste of funds that could be appropriated somewhere else.
What Would You Do with an Extra 30% in Savings?
Better planning is always the first line of defense to this type of over spending. But with spreadsheets and crystal balls it is really hard to accurately plan for future capacity and spending. There are a few ways VMTurbo can help with this issue and keep your spending down, or eliminate it almost completely.
Most customers of VMTurbo see a 30% increase in efficiency, now what does this mean to you? This means that we give you 30% MORE workloads on your already existing infrastructure. This can translate into not having to buy hardware when you thought, or being able to repurpose the hardware you currently have to “spares” so when you do grow you do not need to buy hardware. The big save here isn’t just the physical tin, it’s also the cost to license that server, power that server, and cool that server. When you add that up the cost of a single server is well beyond that of the “physical price.”
How Does VMTurbo Do It?
So how do we do this? Through the use of VMTurbo’s control and automation, we are able to safely drive these densities up across your environment. And all of this is based off of intelligent workload placements; abstracting virtual and cloud environments as an economic market is key. It enables data center entities to follow the principles of supply, demand, and price, figuring out placement, sizing, and start/stop actions among themselves. These entities are driven to get the best price for resources (CPU, Memory, vCPU, IOPS, etc.), whether buying or selling. Buyers get the resources they need and only the resources they need. Each entity is selfishly driven to get the resources they need to perform, for the best price. The collective actions simultaneously drive performance and efficiency. It allows you to safely fit more workloads in your environment, staving of hardware costs by allowing you to grow without purchasing.
Plan for Business Growth, Not Hardware Growth
In my experience as a Systems Engineer, when we first ran optimization plans with VMTurbo in my company I was able to turn off 52 of my roughly 250 hosts. Obviously we didn’t turn them off because they where already there so it would have been a bit of a waste, but what this let us do was stave off that hardware purchase that we thought was looming around the corner. It also gave me the ability to move blades around for things like “test” clusters or swinging clusters that I was using to temporarily hold workloads for onboarding new companies.
The other side of this coin is accurately planning for growth and projects by using VMTurbo’s planner. By being able to accurately plan for future growth and big projects by using your own live data, you can cut hardware and licensing costs. With the ability to do this you no longer have to over by your hardware. As of right now the industry standard is to over spend on hardware “just in case.” With VMTurbo we can work to change the way we approach this budgeting and costs control.