The Data Center Management Guide for Data Center Assets is meant for giving data center managers the upper hand so they can get the best performance and the most money out of their equipment. This guide will cover everything from asset procurement at the beginning of the IT lifecycle to IT Asset Disposition at the end.
Data Center Management: Overseeing Complex Processes
Data Center Management seems about as complex as overseeing a rocket launch.
“It is remarkable how many things you can explode. I’m lucky I have all my fingers,” -Elon Musk on launching rockets.
A certain tenacity is required for both a successful rocket launch and successful data center management.
For data center managers, this means a ridiculous number of systems to constantly maintain, frequent upgrades, inevitably expensive breaks, potential data security oversights (and subsequent leaks), and unexpected downtime that sends end users into a frantic uproar. Overall, a veritable disaster waiting to happen.
How Amazon’s Servers Went Down
IT disasters do happen, even to the best of companies. All the time. Even though data center managers are some of the most qualified and intelligent specialists in any field, costly down time may still be inevitable.
Amazon had a massive outage because of ONE guy! He was authorized and operating within his capacity, and accidentally issued a command to shut down more servers than intended. After hours of downtime and millions in loss, Amazon said it is “making changes” and improving recovery time of its systems and having their data center managers implement new safeguards.
The Best Disaster Prevention: Planning Ahead
That reactive, after-the-fact band-aid fix is a big problem in data center management, even with companies as large as Amazon. Data center providers are under constant pressure to provide a growing volume of service; this rushed dynamic is the fuel for accidents.
Despite frantic schedules, proactive planning and solid strategy can mitigate the majority of the potential mishaps, accidents, screw-ups, pitfalls, kerfuffles(that’s a great word isn’t it?), and errors that would be inevitable without any planning.
If one section interests you most feel free to jump to that section, but otherwise they are meant to be read in order to give an overview of the whole process.
The Processes and Assets a Data Center Manager Oversees
Primary asset management tasks:
- Asset procurement
- Asset maintenance
- Data center decommission
- Data center migration
- DIY Remarketing- When & Why
- IT asset disposition(ITAD)
Section 1: Overview of the Data Center Management Cycle
The processes listed above make or break an enterprise’s total cost of ownership and return on investment. Optimizing that list will result in the best performance, longest life, highest returns, and lowest expenses in their data center operations. Those who do not will see stifled results.
When looking at these areas, begin with asset procurement, since it is the start of the equipment lifecycle. Ideally, equipment procurement is where the planning for maintenance and the end of life also begins.
As needs grow, and technology changes, a data center migration may be necessary. This may be a migration to the cloud or a physical relocation; these events are both fraught with risks and common pitfalls for data center managers to avoid. Eventually, the assets will reach the end of their life, and the data center may need to be decommissioned, or perhaps only some of the assets need to be taken care of. Regardless, data erasure is a significant concern before data center management decides whether to hire an IT asset disposition company (ITAD) or remarket the assets themselves.
Data Center Management Step 1: Asset Procurement
Asset procurement is the first step in the lifecycle of our equipment. Hardware-specific advice on purchasing would be outdated within a week. Secondly, every type of enterprise has different needs for their data center. For example, a hosting company’s data center manager is going to be more concerned with reliable, robust performance from his equipment, but won’t prioritize speed over dependable equipment. Nobody in their right mind would even consider hiring a hosting company if they can’t even host reliably.
On the other hand, a high-frequency trader would give anything to run faster computations, even just by milliseconds. Many of them will buy ANYTHING if it promises speed, even if it’s completely untested. Thus, procurement of actual parts and systems themselves will vary heavily among enterprises.
Plan for Eventual Asset Decommissioning Before Buying New IT Equipment
One thing data center managers all have in common is that they should plan ahead for the eventual retirement of IT assets before they are purchased. If you’re buying hard disk drives, look at what older drives are selling for and how long they appear to sit on the market. If they’re going to be worthless and impossible to sell by the time you would be ready to upgrade, then there’s no point planning for an ideal asset retirement window. If you’re not sure or don’t want to do the research yourself, you can always call Exit Technologies. They have the market experience to provide an idea of how much it may be worth in the future, as well as what the decommission process will look like.
Begin with the end in mind, and you will generally get where you want to go. Now let’s move on to the next phase in the life cycle of our assets: maintenance.
Data Center Management Step 2: Data Center Maintenance Services
In data center operations, maintenance costs for equipment are a sizable part of the budget. The lifetime cost of something like a gun is mostly just what you paid for it to begin with, as maintenance costs are very minimal. On the other hand, costs to maintain enterprise data center equipment will easily exceed the purchase cost by the end of its life. Most data center managers use the original equipment manufacturer’s (OEM) services like Cisco Smart Net Total Care, Juniper JCare, Brocade Essential support, etc., to service their equipment.
Switch to Third Party Maintenance (TPM) After the OEM Warranty
Since OEMs don’t offer multi-vendor support, i.e. Cisco offering support for Brocade equipment, the end result is multiple service level agreements with different OEMs for different equipment. This greatly complicates the process of maintaining equipment. Additionally, it’s not uncommon to save 40% or more when switching from OEM maintenance to Third party maintenance.
As a result, progressive and financially savvy enterprises are realizing that third party maintenance (TPM) services provide incredible value as an alternative to OEM services. You’re pretty much stuck with the OEM for brand new equipment. When an asset is out of warranty, the cost savings and flexibility offered by TPM can be very tempting.
OEMs SELL equipment as well as service it, and their priorities reflect the wide net of operations compared to TPM providers.
Why TPM Companies put You First (not the Equipment)
The only priority for TPM companies is their service. Most would prefer to find a physical therapist who is trying to fix the problem, and not just sell their newest rehab equipment.
Both industry leaders Gartner and Forrester have recommended TPM to be included in any holistic data center management strategy. Less downtime, faster response from technicians/engineers, lower costs, and more flexible service level agreements (SLAs) are typical for any enterprise that switches their equipment from the OEM to a TPM provider.
OEM Costs Keep Growing as Your Equipment Loses Value
Companies can expect the bill from the OEM for maintenance to keep growing every year. A paradox ensues when the book value continues to fall while the cost to maintain it continues to rise. The alternative of locked in rates offered by TPM providers are more financially appealing.
Other Advantages of TPM Providers
Data center managers can expect a nearly 50% quicker first response, and less time spent waiting for the problem to move up the chain of engineers to actually solve it. Ever had an issue with your TV, and the first two Comcast technicians come out and have no idea what they’re doing? This won’t be a problem with a reputable TPM company. A single point of contact is typical for most reputable TPM providers as well, which simplifies the process for data center management.
Another appealing aspect of TPM for data center management is their priority on extending the life of the equipment as opposed to replacing it with new equipment. OEMs get a good portion of their profits from new hardware releases. Obviously, this means that it’s in their best interest if you upgrade your equipment as frequently as possible. If you’re buying a new car every year because your roommate is a pushy car salesman, your bottom line will be suffering.
TPM Providers Tend to be More Qualified
Many believe that the engineers within TPM vendors are less skilled, but this is a myth. In fact, many of the engineers that TPM vendors recruit from OEMs score poorly on the evaluations TPM providers put their new engineers through. It’s not uncommon for a TPM engineer to be considered a specialist in multiple vendors’ equipment, while an OEM engineer typically only specializes in one.
With brand new equipment, it’s still going to be common and a borderline necessity to use the OEM’s service agreement. However, as soon as the equipment is out of warranty, a prudent data center manager should take a look at third party maintenance. It never hurt anybody to have a quick discovery call!
Conclusion and Eventual Next Steps: Data Center Migration or Decommission
That covers the longest phase in the life of your assets: maintaining equipment for unhindered data center operations. If data center management can squeeze the most use out of equipment and keep it in the best condition, returns upon resale will be high and the total cost of ownership will be reduced. Win/win.
Unfortunately, your data center equipment won’t be relevant forever. What happens when it is time to get rid of all that equipment you were able to maintain so well? That’s where the next section about data center decommissioning and data center migration comes in.
Have something to add? Let us know your thoughts in the comments below!