Guest Post from Richard Donaldson
I’d been holding out a bit on writing this as it really is a synthesis of ideas (aren’t they all) with special mention of dialogue with Jeffrey Papen of Peak Hosting (www.peakwebhosting.com)…
I’ve been collaborating and speaking extensively with Jeffrey on the next phase of “hosting” since we are now moving beyond the hype cycle of “Cloud Computing” (see previous post on “The end of the Cloud Era”). The community at large (and people in general) love the idea of simple, bite sized “solutions” with pithy and “sexy” naming conventions (think <30sec sound bites) and that was the promise/expectation around “the cloud” as it was popularized – a magic all in one solution whereby you just add applications and the “cloud” will do the rest. Yet, the promise never quite met expectations as the “cloud” really ended up being an open standards evolution of “virtualization” – nothing wrong with that, just not the “all in one” solution that people really wanted the cloud to be (ps – all in one refers to the aforementioned of applications just being pushed thru APIs to the “cloud” and the “cloud” manages all underlying resources).
So, as the Cloud Hype dissipates (love the metaphor), we are sorta back to the same basic elements that make up Infrastructure – Datacetners, Compute (IT), Communications (switches/routers), Software that manages it all (virtualization, cloud, etc), all accessible thru the to be built APIs. Put another way, we are coming full circle and back to centralized, on-demand computing that needs one more element to make it all work – Subject Matter Experts (SMEs).
I was inspired to write this today when I saw this post from Hitachi: http://www.computerworld.com/s/article/9226920/Hitachi_launches_all_in_one_data_center_service - “Japanese conglomerate Hitachi on Monday launched a new data center business that includes everything from planning to construction to IT support. Hitachi said its new “GNEXT Facility & IT Management Service” will cover consulting on environmental and security issues, procurement and installation of power, cooling and security systems, and ongoing hardware maintenance. It will expand to include outsourcing services for software engineers and support for clearing regulatory hurdles and certifications.” This is the comprehensive “build to suit” solutions the market has been seeking since the cloud – it includes everything to get your infrastructure building blocks right and is provided as a service – but what do we call this service????
How about “Operations-as-a-Service“!!
OaaS pulls together the elements in IaaS + PaaS + SMEs. It outsources the “plumbing” to those that can make it far more cost effective thru economies of scale. Sure, there are a select few companies who will do this all in house: Google, eBay, Microsoft, Amazon, Apple (trying), and of course, Zynga. Yet, these companies are at such massive scale that it makes sense – and yet, they even have excess (at least they should) capacity which is why AWS was born in the first place and we are now seeing Zynga open up to allow gamers to use their platform (see: http://www.pocketgamer.biz/r/PG.Biz/Zynga+news/news.asp?c=38455). Yet these are the exceptions and not the rule.
The rest of the world should and is seeking comprehensive, end-to-end Operations as a Service provided by single vendors. It doesn’t preclude the market place from buying discreet parts of OaaS individually, however, the dominant companies that will begin to emerge in this next decade will seek to add more and more of the OaaS solutions set to their product list thereby catalyzing a lot (I mean a lot) of consolidation.
I will be following up this blog with a more detailed look at how this concept is playing out, yet in the mean time would very much like to hear the feed back on this topic – is the world looking for OaaS?
William Dougherty is VP of Information Technology for RagingWire Data Centers, with over 15 years designing, building, securing and operating high-availability computing systems. He tweets regularly on industry issues @bdognet.
Tour any three data centers and you’ll be left scratching your head trying to differentiate between them. As a result, price and proximity become the primary decision points in an otherwise seemingly level playing field. However, there are crucial differences between data center providers that can drive up both your costs and your IT downtime risk. Ask the following questions of your potential data center provider and you will keep your costs and your downtime risk as low as possible.
Which components of the data center facility are both concurrently maintainable and fault tolerant?
Many data centers claim to be N+1 or N+2 redundant. Sounds great! But data center facilities have to maintain their equipment, too. The question you need to ask is, what impact does concurrent maintenance have on the fault tolerance of the data center? Pose the following scenario on your facility tour: generator #1, UPS #2, and CRAH unit #3 are unavailable due to maintenance; the facility loses utility power and generator #4, UPS #5 and CRAH unit #6 all fail to operate. What happens to their customers?
Your critical IT infrastructure operates in a world where utility outages or equipment failures happen. In the above case, N+1 redundancy won’t protect you. Your IT infrastructure needs isolation from multiple simultaneous events (N+2 or 2N redundancy at every level).
What are the average and maximum power densities of the facility on a watts per square foot AND watts per cabinet basis?
About 10 years ago, the first data centers were designed for much lower rack power densities than are required today. As a work-around, some facilities space cabinets farther apart (while charging you more) to accommodate higher-density clients. Most data centers are built to support an average of 100W – 175W per square foot with more modern data centers supporting an average of 225W per square foot and scalable on an individual basis up to 400W per square foot.
Cabinet power densities are equally important. Ten years ago, a 2kW cabinet was sufficient to power a full 42U of x86 servers. With today’s multi-core, high density blade servers, 8kW – 10kW is the new rack power minimum. Expect your required power density to climb and make sure your data center has the infrastructure to grow with you.
How often does the data center load test its generators?
Fuel consumption and expensive test equipment makes load testing generators a costly maintenance item. Sometimes, in lieu of regular load testing, data centers will use unexpected utility power outages as a way to “load test” their generators on client IT loads. If they aren’t regularly load testing, then they are likely to only identify generator problems when utility power fails, which is precisely the wrong time to find an issue. Ask your data center provider if they put every generator on an extended load test – not just a start-up – at least quarterly.
What are the highest risk natural disasters for the area, and what has the data center done to mitigate their impact?
Every data center is subject to natural disasters, but some are more vulnerable than others. The Uptime Institute has published an excellent study of composite risk from earthquakes, hurricanes, tornadoes, snow and other disasters. There are relatively few regions in the U.S. that fit in the low-risk zones. Another scenario: the data center survives a massive earthquake, but utility power is still out with no estimated repair timeline. How long can the facility’s generators run on normal on-hand fuel supply and how many suppliers are contracted for refuel services? It is critically important that you a) understand what disaster scenarios are likely for the facility; and b) work with your provider to make contingency plans based on likely risks.
What are the minimum skill sets of the remote hands and eyes staff?
It is an absolute certainty that at some point your equipment will need to be physically serviced. You can either drive to the data center or use their remote hands and eyes services. Some data centers cut corners by using security guards to provide remote services. You want to make sure the remote hands staff provided by the data center consists of true IT professionals. Ask for minimum job requirements and speak with the service manager so that you know who is answering the phone at 2 AM. Quality remote hands staff can reduce the relative importance of proximity in your decision making process.
What certifications has the data center earned, and do they undergo annual audits to maintain them?
SSAE 16, PCI DSS, LEED Gold, Energy Star, FISMA, HIPAA, SCIF, Tier IV. Each standard is a useful tool in differentiating your data center choices. If you process credit cards, you want your upstream providers to support your PCI compliance by maintaining their own PCI ROC. If you are a financial organization, you need your data center to be SSAE 16. If your company is environmentally conscious, especially if you purchase carbon credits to offset your power consumption, you want a data center with EPA ENERGY STAR and LEED Gold certifications. Ask your data center for proof of their certifications. This information is invaluable because it represents an independent analysis of the facility’s quality, reliability and security.
Bonus question #7 – Does the physical security include sharks with frickin’ laser beams attached to their heads?
Why? Because it would be cool if it did. Look under the raised floor sometime. Maybe there’s something interesting down there…
Gulf Bank inaugurates IT data center facility in Hawally http://bit.ly/j54wdb
Telx Closes $200 Million Amendment and Incremental Senior Secured Credit Facility http://mwne.ws/dGCS4M
Hurricane Electric Establishes PoP in VegasNAP Las Vegas Colocation Facility http://bit.ly/ih9glI