January 7th, 2008
What do Xbox Live problems say about Microsoft's back-end infrastructure?
Whether you think the recently proposed class action lawsuit over Xbox Live service holiday outages is crazy or justified, there are undeniable problems with Microsoft’s back-end Live infrastructure that Microsoft needs to address.
Microsoft recently offered its 10 million Xbox Live subscribers a free arcade game as payback for holiday 2007 service outages (or, as Microsoft officials preferred to call them “intermittent Xbox Live issues”), which the company attributed to an insufficient number of back-end servers.
But as a posting this past weekend by Windows Connected contributor Matt Freestone, the Xbox Live outages seemingly still aren’t over:
“Even now as I write this post at 5:30pm Sunday afternoon (1/06/08) Xbox Live is still essentially useless. Last night I played Halo 3 for about 5 hours while getting about 2 hours worth of gameplay. Today, when attempting to play Call of Duty 4, I spent about 2 hours to play about 3 rounds. Getting constantly dropped from the lobby, constantly dropped from games, and sometimes a complete inability to find a game was the norm, and not just for me. I spoke with people in every game lobby (when I was lucky enough to get into one) and they all reported the exact same issues, since Christmas.”
Xbox Live is one of a growing number of Live services that Microsoft is hosting across the network of datacenters that it is building. The company prides itself on its common back-end infrastructure that supports its diverse set of services.
At the Microsoft Strategic Architect Forum last November, Microsoft’s Corporate Vice President of Global Foundation Services, Debra Chrapaty discussed Microsoft’s back-end infrastructure planning and philosophy. Chrapaty is in charge of “strategy and delivery of the foundational platform for Microsoft Online Services including globalization, operational infrastructure, networking and datacenters,” according to her Microsoft bio.
Some excerpts from her presentation, entitled “The Reality of the Cloud and the Future of the Data center” (which are in shorthand note form — all I could find as a transcript):
“We’re building a lot of data centers. Now we plan around kilowatts, or in our world, megawatts. How much to accommodate server and space? Drives our planning. Need to nail close. We are an internal service provider for Microsoft. Data center floor space to networks to security to tier one support. A business inside the business model.
“We’re looking at planning for delivery. I never want to be a bottleneck for Microsoft. Products enormously successful. From planning perspective, conservative and need much more than they planned for. Planning, mindful of agility is on my radar. Don’t ever want to not have a place to scale for power. Capacity from power perspective, 20MW houses the data center about $400 million and more. Big decisions.”
While Chrapaty’s clients are the various product divisions at Microsoft, it’s the end user customers of Microsoft’s services that are the real customers about which the company needs to worry. As Windows Connected’s Freestone noted: “We pay for the stability of the network, and if that isn’t there, there isn’t much worth paying for.” Obviously, the same rule applies for customers of Microsoft’s other consumer and business services….
Any of Microsoft’s Live customers out there encountered recent outages worth writing home about?
Mary Jo has covered the tech industry for more than 20 years. Don't miss a single post. Subscribe via Email or
RSS. You can also follow Mary Jo on Twitter.
Got a tip? Send Mary Jo your rants, rumors, tips and tattles. For disclosure on Mary Jo's industry affiliations, click here or to see Mary Jo's full profile click here.
