top of page

The Elephant in the Room Regarding Data Center Reliability

When you examine the Uptime Institute Tier Standard: Topology, you quickly realize that unplanned outages should never occur; every key infrastructure system has built-in redundancy, so any single failure should never cascade into a full-on outage of a site built to either Tier-III or Tier-IV data centers.  Looking at Tier-III sites (the vast majority of data centers are built to this standard), the mathematical probability of a Tier-III facility suffering an outage in any given year is 0.018%.  Translating this into odds, that’s 1 in 5,556. 


And yet according to the 2023 Uptime Institute Outage Analysis, 60% of data center operators had suffered an outage in the past three years.  This roughly translates to a 1-in-5 chance of an outage occurring in any particular site (assuming all Tier-III) facilities.

  

Let me put this bluntly: actual performance is three orders of magnitude below where it is supposed to be.


WHY?


The Uptime Institute tier standards are THE standards for the entire industry, because this standard is quite good at establishing expectations not based upon any particular technology, but upon overall system design to achieve performance expectations.  


Like the commercial aviation industry, airplane design standards are vendor-neutral, relying instead upon overall performance.  The FAA sets expectations for performance, similar to what Uptime Institute does, with excellent results; and superb designs that are safe, efficient and reliable.


So we know it’s NOT the design of the facilities.  


Unlike the aviation industry, which of course is heavily regulated due to the nature of business, the data center industry isn’t regulated in the slightest.  It’s more like the Wild West, where anything goes.  But the Uptime Tier Topology Standard sets enough of a rubric that the allowable variations of design still establish a widely accepted standard.


But that’s as far as it goes.


MAINTENANCE IS THE KEY


The aviation industry is heavily regulated by the FAA for design and construction, as well as maintenance after the sale is complete.


There is no controlling authority for the data center industry with regard to maintenance, so enterprise IT organizations have no way of knowing if their equipment is being properly protected.  


In response, one group has sprung up to produce a veneer of respectability and diligence in this regard, namely the American Institute of Certified Public Accountants (AICPA).  


What, you may ask, does that have to do with data centers, and the IT world in general?  


The answer is that the AICPA came up with an audit program to “audit” data centers for things like cyber-security and proper equipment maintenance, under programs like the Service Organization Control Type 2, AKA “SOC-2.”  The stated primary purpose of SOC 2 is to ensure that third-party service providers store and process client data in a secure manner.


The only qualification you need to be SOC-2 auditor, is that you must be a Certified Public Accountant, a CPA.


That’s IT.


Now I ask you, what does a CPA really know about cybersecurity?  Probably about as much as they know about UPS systems, generators, VRLA batteries, switchgear, chillers and DX CRAC units.   Which is to say, “nothing.”


If you were about to take a flight across the ocean on a Boeing 737, and a CPA came up to you and said, “I audited the company, this airplane is certified by me,” would you bet your life on that?


The answer would of course by “heck NO!”


It’s the same thing with the data centers; the so-called auditors don’t know anything about engineering systems and are therefore grossly unqualified to test whether the controls the data center company has put into place actually deliver the expected results.  


THIS test is where the rubber really meets the road, and they’re technically incompetent on the subject matter.




[I’ve been told this by a prior SOC-2 auditor who quit (one of the “Big 4” accounting firms) in disgust.  I’ve had data center managers laughing at the whole situation.  These are people with centuries of cumulative experience- all told me the same thing: no SOC-2 auditor they met actually understood the systems being audited.]


So you have no way of knowing whether the safety-net for your equipment in a leased space (or owned, but where the labor is outsourced) is being properly maintained, or your at risk of an unplanned outage, the next time utility power becomes unstable.


RESULTS


Any high-performance system, if you expect it to function over a long period of time, requires a superb design with equally superb maintenance.  They are two sides of the same coin.  

Data center tier topology helps us to create excellent system designs.


But when data center reliability literally a thousand times worse than expected, exposing enterprise IT organizations to risks of unplanned outages, you suddenly see the elephant in the room.

You can’t un-see it.


IS IT HOPELESS?  WHAT’S THE SOLUTION?


Instead of talking about what is “possible,” I’ll just tell you what I delivered.


I developed a maintenance oversight program, which delivered 100% perfect uptime for 6 ½ years for over 60 sites around the world, which included many Tier-2 sites, and colocation facilities!  Stated another way, I built a maintenance oversight program that delivered a perfect operational record for >6 years, which also incorporated audits of colocation facilities across the globe.  It beat even the 1-in-5,556 odds that Uptime sets as their performance standard, literally improving colo reliability performance by more than 1000x.


I can deliver the same to your company.  


VALUE PROPOSITION


The AMERRUS value-proposition is quite simple: at the minimum, I will assure your facilities meet the performance expectations established by the Uptime Institute standards.  That’s the worst-case scenario.  


But our goal will be the same as we’ve already been delivering: a perfect operational record.

You won’t find anyone else, anywhere, that will make such claims, because their isn’t anyone else out there, that has the team, the tools, the knowledge and the system to deliver the goods.  It isn’t even close.


Contact us at www.amerruss.com We can help your IT organization be more reliable, AND cost-effective, without stressing your budget or your nerves!






bottom of page