They will have flash storage “architected into the system from the beginning”, directly in the server, Dell calling it tier zero. He talked of Dell’s Fluid Data Architecture extending across EqualLogic and Compellent, with tiering extended from HDD and flash in the array into the server, via technology from the RNA Networks acquisition.
He said these flash-based servers will run database queries sixty times faster, that’s right, 60X faster, than today.
He talked of data centre racks, single racks with 1,024 cores linked by 10GbitE to a Force 10 switch, and the switches talking 40GbitE to a core switch. The rack would contain 40TB of DRAM and 40TB of flash and run “more workloads per server than ever before”.
We went from provisioning servers in days, to getting base virtual machines up and running in under 8 seconds. Want Service and Application images (for established products)? Add another 8 seconds or so. Want to roll it into production globally (changing global DNS/Load balancing/Security changes)? Lets call that another minute to roll out. We used Open Source products and added our own development glue into our own systems to make all this happen. I am incredibly proud of my Cloud teams here at AOL, because what they have been able to do in such a relatively short period of time is to roll out a world class cloud and service provisioning system that can be applied to new efforts and platforms or our older products. Better yet, the provisioning systems were built to be universal so that if required we can do the same thing with stand-alone physical boxes or virtual machines. No difference. Same system. This technology platform was recently recognized by the Uptime Institute at its last Symposium in California.
Another incredible aspect of this new data center facility and the technology deployed is our ability to Quick Launch Compute Capacity. The total time it took to go from idea inception (no data center) to delivering active capacity to our internal users was 90 days. In my mind this made even more incredible by the fact that this was the first time that all these work-streams came together including the unified operations deployment model and included all of the physical aspects of just getting iron to the floor. This time frame was made possible by a standardized / modular way to build out our compute capacity in logical segments based upon the the infrastructure cloud type being deployed (low tier, mid-tier, etc.). This approach has given us a predictability to speed of deployment and cost which in my opinion is unparalleled.
Yahoo Unveils New Customer Care and Data Center in Lockport, NY
By Paul Bonaro, Senior Director, Data Center Operations & Sarah
Taylor, Manager, Customer Experience Operations
celebrated a huge milestone for Yahoo at the grand opening of our new customer
care and data center in Lockport, NY!
Our latest expansion
at our Western New York facility provides both additional capacity to power our
products and world-class customer support for Yahoo users. We’re proud to have
expanded our Lockport team by creating an additional 115 jobs.
Since we first
broke ground here in 2009, Yahoo has invested more than half a billion dollars
into our operations in Lockport. We’ve also given back to the community in a
number of ways, including through our commitment to donate $3.5 million to the
Community Foundation of Greater Buffalo.
Our new data
center in Lockport utilizes our Yahoo Compute Coop (YCC) architecture, evidence
of our commitment to energy efficient designs. The YCC is a world-class,
resource-efficient, green data center design that uses significantly less energy
and water than conventional data centers. The unique shape of the building and
the use of outside air-cooling eliminates the need for expensive and
energy-intensive chillers to keep the servers cool.
We were joined by
an incredible crew to help us mark the occasion. Members of our executive
leadership team – including David Filo, Co-Founder & Chief Yahoo; Mike
Kail, Chief Information Officer; and Catherine Card, VP for Customer Experience
— stressed how our Lockport expansion is another stepping stone in what has
been a rich partnership between Yahoo and the Western New York communities
We were grateful
to host some of our key partners from New York State, including; Lieutenant
Governor Kathy Hochul; New York Power Authority (NYPA) Chairman John R.
Koelmel; Town of Lockport Supervisor Marc Smith; Town of Lockport Industrial
Development Agency (IDA) Chairman Tom Sy; NY State Senator Robert Ortt; NY
State Assemblyman John Ceretto; and NY State Assemblyman Raymond Walter.
The success of
Yahoo’s operations in Lockport are a result of our incredibly talented
workforce, the business-friendly environment of Western New York, and access to
hydropower and temperate weather. Western New York continues to be a great home
and partner to Yahoo.
A quick word about different planes. In any system there are essentially three planes - the Data Plane, the Management Plane and the Control Plane. The Data Plane is where the traffic flows (whether it is a storage cloud or a transportation system). The Management Plane monitors and reports activities. The Control Plane is the set of knobs, the signals, the throttling, the metadata and configuration and so forth.
The current cloud architecture is scale-out, or add more non-redundant smaller machines to increase capacity. But this architecture does not relieve the architect of the responsibility of reliability and availability. What is happening is that the redundancy is moving from the data plane to the control plane and moving away from MTBF to MTTR. The responsibility to accommodate failures is the job of the designers and not the ops. In short, we know the components will fail, in fact fail in droves and so we expect the systems to be up and running in spite of the failures!
Einblicke in Googles Rechenzentrum. Google gewährt erstmals Einblicke in seine Rechenzentren. In insgesamt 8 Städten auf der ganzen Welt stehen die Serverfarmen von Google. Und die Bilder davon sind wirklich beeindruckend.
Located in Stockholm, the ISP Bahnhof gained notoriety when it hosted wikileaks’ sensitive materials. Built into the Pionen mountains, the center took two years to retrofit the caves with enough space for their server racks and backup generators. In total, it is roughly 1,200 square meters (12,900 square feet) and will hold the future of blogging after the apocalypse.
As an introduction, I thought
might be fun to do a popularity check by comparing Twitter followers
(yes, every player on the roster has an account):
for the sake of comparison I threw in a few USMNT players, and a few
other high level female athletes. The USWNT comes out quite visible
considering both Carli Lloyd and Shannon Boxx have more followers than
the last two WNBA MVP’s; over half the team is ahead of star players
from other countries – like Christine Sinclair from Canada and Nadine
Angerer from Germany; and the top stars even pip big name US men’s
players like Landon Donovan and Tim Howard. For better and for worse,
they are the popular kids.
They are also kids who had lots of opportunities at young ages to develop as players, which for me always raises questions about ‘relative-age effects’: do players born at certain times of year gain advantages because they are more athletically mature at younger ages when grouped using somewhat arbitrary age cut-offs? The data for this iteration of the USWNT is probably not statistically significant, but it does look potentially meaningful to me:
Why would more players be born mid-year? Not sure. This is actually the opposite of what you might expect given school-year age cut-offs; most American females of this age would have been grouped into grades based on a birthday cut-off of either September or October 1st. So kids born in June, July, and August (the largest group of USWNT players) would have been competing with kids almost a full year older at younger ages. So the above pattern may just be statistical noise (note some research has found relative age effects in US youth soccer to be less pronounced for girls compared with boys). But it also might be something more interesting–I actually wonder if some of the women’s players benefit from not over-training and over-competing at early ages. But, as the social scientists say, further research required!
As another step on the ladder, and as a hat tip to the cultural oddities of American soccer, every single player on the USWNT has been to college – and had an identifiable major therein (I can’t be 100% sure everyone actually graduated, but it seems possible – which would have to be an extreme rarity for any team in a full World Cup besides the US). But the fields of study are not entirely representative of the broader US higher ed patterns (estimates of national averages from data here):
Of course I’m excited to learn there are so many social science fans on the USWNT - almost half the team majored in psychology (3), sociology (2), communications (5), political science/economy (2), and/or African-American Studies (some double majors)! Sarcasm aside, I suspect this is more to do with social sciences being a reasonable major without a lot of requirements that would get in the way of devoting most of one’s time to soccer. Pre-med and engineering just don’t end up being very manageable when much of the day and week is blocked off for practice and game travel. The over-representation of kinesiology makes sense; the slight under-representation of business is a bit surprising.
Finally, and more seriously, the most sociologically interesting demographic of the team would seem to be the extremely limited racial and ethnic diversity. Though it is hard to get clear data here, with a bit of searching my best guess is that only a few of the twenty-three players are racial minorities – media reports note Shannon Boxx and Sydney Leroux both have African-American fathers but were raised by white mothers, Christen Press also seems more speculatively (based more on discussion boards rather than clear media reporting) to have mixed racial heritage, and Amy Rodriguez’s father had Cuban immigrant parents (though that does not necessarily mean anything about racial or ethnic status). At the very least, that means almost 85% of the team is what the census folks call ‘non-Hispanic white’ – compared to about 60% of the US population. Though an N of 23 might not make this difference statistically significant, it does seem clear to me that women’s soccer in the US has a diversity problem – there is not enough access or opportunity for non-white players and for low-SES players (for a more thorough discussion, see here). The reasons for that are complicated and multi-factorial, but I suspect they are also a real problem for the future success of the team (by virtue of missing so much of the potential talent pool). And it also just seems un-American…
“Systems management software and support is a huge market, but few, if any, legacy vendors have products and knowledge that easily translate into webscale environments. Facebook could stand to make a lot of money by consulting with customers on how to build their data centers and architect their applications, and then selling them the software tools to keep those apps up and running. I’m not saying it’s going to happen, but it wouldn’t surprise me in the least if it did.”
The net of things so far is that Apple’s iCloud service offering will be running on both Microsoft Azure and Amazon AWS infrastructure when it launches this Fall.
Although neither Amazon or Microsoft are willing to comment, Apple Insider has suggested this will allow Apple “…to focus on it’s strength of building a consumer facing service rather than becoming a cloud provider.”
These articles are all good reading and it would be even more interesting to know how they fit in with the new mega data center Apple’s building in Maiden, NC.