So I have had the opportunity over the last few weeks to talk to a number of folks about server virtualization--likes, dislikes, where its going. Depending on your choice of market research or anecdote, it seems that the number of virtualized production x86 servers in the the neighborhood of 10%. Again, depending on your favorite flavor of Kool-Aid, the expectation is for this to jump to 40-60% in the next 2-3 years. Even the most conservative scenario shows a significant increase in server virtualization.The question I have--and I’d love to hear from folks who have deployed some form of server virtualization--is what needs to happen in your data center to make the jump from 10% to the aforementioned 40-60% range. Do you see any inhibitors to having more of your servers virtualized?One of the areas we see some challenges is around ensuring network and storage services follow virtual machines as the undergo live migration. More than one customer has noted the challenges of using VMotion or DRS in a VMware environment. There are certainly workarounds, but they are often operationally burdensome, and what may work when you have 10% of your production servers virtualized may not be tenable when half your servers are virtualized. Because, by its nature, there is a degree of abstraction in server virtualization, I have also heard a number of concerns around areas like troubleshooting and regulatory compliance.As always, the industry is evolving to meet these new challenges. For example, one of the advantages of unified fabric is the ability to deliver a consistent set of network and storage services to all the attached servers in the data center, which simplifies live migration to some degree. However, there still seem to be gaps. Some gaps might be technological, while others might be more along the lines evolving the org structure to deal more effectively with shared, virtualized infrastructure.Anyway, what do you think--what needs to happen to drive a higher rate of virtualization for production servers?
An interesting write-up from Greg Ferro at ‘Ethereal Mind’ on how a lot of what we do as a company pulls together to create cloud computing style infrastuctures for our customers. What thoughts do you all have on this -- make sense, or way off base?
You want to read as well Building the Cloud with Cisco Unified Computing System , EMC and VMware
Was reading this really interesting article on Data Center Knowledge about Wall Street Grids. This is one of those applications that I am not sure will move to a cloud architecture. My reasons are as follows:1) Latency is of paramount importance in this space. The faster a transaction models, and transmits, ehte faster the trade may execute, and trading is a race. First come, first served. So these architectures tend to be hyper-optimized on transaction latency reductions.2) Corporations tend to not outsource areas of strategic advantage. I was meeting with the CTO of a large investment bank who commented- “IT contributed $7B to the bottom-line of our firm this year, talk to me about how we can grow that next year.” In Financial Services IT can provide tremendous strategic and sustainable competitive advantage. If a resoure is that critical to your business you usually don’t delegate or entrust it to anyone else.3) Growth rates in this area are larger and faster than many people realize. I know of an almost arms race between some firms where the number of servers is north of 10,000 for Monte Carlo simulations to build risk analysis infrastructure. What do you all think? Is this a space that lends itself to a cloud architecture, and if so- what properties and capabilities does the cloud need to offer in order to offer a credible service, and if the environment is that competitive how could the cloud provider sustain more than one customer?dg
Just read the following article published by Mario Apicella from Infoworld. Mario’s article espouses the value of FCoE in providing storage connectivity ubiquitously to a host. When coupled with an application such as VMotion that demands perfect state synchronization not having to move a huge runtime image speeds the move. Having all of the servers connected to a common infrastructures overcomes the addressing and segmented connectivity challenges imposed by a SAN island architecture.Mario- appreciate your testing, and validation that what we have here is a technology that is not only easy to use and real, but is also adding value to virtualization and VMotion.dg
Whether you call it eating your own dog food or drinking your own Champagne, we at Cisco are lucky to have very dynamic IT and facilities teams that trial our solutions as we develop them. Going Green is no exception. We recently kicked of our project within Cisco data centers and labs to benchmark our operative efficiency in order to make practical, targeted reductions in support of a public commitment to cut our Green House Gas (GHG) emissions by 25% from 2007 levels through 2012. By any measure this is aggressive in that it is an absolute reduction with no fine print…That means owned and leased properties as well as employee travel. Our rough breakdown of emissions is as follows: 1) Labs and Data Centers = ~60% 2) Employee Travel = 27% 3) Other; including refrigerants, propane, diesel for gen-sets, etc.You will notice we have an ~ sign for labs and data centers. This is because without specific instrumentation (sub-metering, branch circuit monitoring and IP-enabled power rails) there is only so much you can tell about where your power is going. You know even less about how efficiently you are using said power. The first level of analysis we can do using The Green Gird’s Power Usage Effectiveness (PUE) metric.So how are we framing this project that allows IT and Facilities to take better control of power consumption, capacity and growth? It is more simple than you might think. If you’ve worked with a spanning tree configuration, this is a breeze! We’ve basically broken this out into “digestible” phases that build upon each other. We’re lucky to have partnerships in some key areas to help us through it and a programmatic methodology through our Efficiency Assurance Program (EAP). Basically here is how we are approaching it today: 1) Establish electrical efficiency benchmarks for a pilot site including IT asset and system utilization (i.e. what is our server CPU utilization, some figures suggest the US average is ~20%) 2) Assign a balanced scoring to the pilot site that balances efficiency with risk and cost 3) Asses the IT and facilities architecture(s) to identify and prioritize improvements to net capacity, efficiency expressed as a percentage and growth projections 4) Develop new Standard Operating Environment (SOE) guidelines at the room, row and rack level 5) Develop new Standard Operating Procedures (SOP’s) focusing on operative efficiency like regular IT asset utilization audits 6) Document it all in one place with the ability for users to easily make changes on the fly but governed by committee (Change Advisory Board as an example) 7) Assess the viability to scale the pilot across Cisco’s 52 data centers world wide (most of those are development data centers, lab/DC hybrids)What is clear up front is that there is no silver bullet for Green. Rather we are focused on taking a blended approach towards efficiency with all leading not bleeding edge technologies on the table. Virtualization is of course a front runner but we recognize that it has a big impact on our facilities as well. Improved collaboration experience to reduce employee travel like Telepresence is a great technology but if we didn’t address human travel habits through our booking systems we might be traveling just as much and using Telepresence on top of that. The list goes on but suffice to say, this is one of those areas where the often overused term of “holistic” really does apply. I talk a bit more about this on our upcoming Techwise TV event, airing August 21st. If interested you can register at www.cisco.com/go/efficiencySo how did we even get to this point? Over the last 18 months we’ve developed cross functional teams made up of what were in many cases people who had a personal interest in making their company a deeper shade of Green (one of those people happens to be our CEO). Lucky for us these folks also have strong business acumen and we’ve been able to sell the business case for doing this internally. This is where our Green Data Center Model Calculator was born from also found in the EAP.We will be making sure to share our progress within Cisco IT on how our Green journey is progressing. Drop us a note with any insight you might have to how this is ramping up at your company (even the political organizational stuff which we all know happens).Thanks for reading.