Level 7 Technology Group

order from chaos

Infrastructure

Insight Global contract at Microsoft

Early in 2011, I was approached by Insight Global, a nation-wide provider of Information Technology talent to companies world-wide for a contract position with Microsoft, in Redmond, as a Service Engineer in Product Quality Online (PQO). I worked supporting Windows Live Solution Center (WLSC). Windows Live Solution Center is a community-based support application for Windows Live properties such as Hotmail, SkyDrive, Messenger, and the sites in the product family. WLSC provides problem submission by ordinary users and resolution by contract moderators and MVPs within a 24 hour period, defined by a Service Level Agreement (SLA).

The application consists of 6 Mirrored SQL 2008 Servers, 31 virtualized W2K8 R2 Web Servers on 11 W2K8R2 Hyper-V host servers using W2K8 R2, IIS7.5, SQL2008, and SCOM for monitoring and alerting.

I worked with a team consisting of developers, business unit managers, the Microsoft Global Support Desk, the PQO and GFS Infrastructure teams. I teamed up with a FTE (Full-Time-Employee .aka. "Blue Badge") and we worked with the PQO and GFS infrastructure groups to transfer the application to PQO. In addition to resolving problems submitted by our customers, I work with the developers to deploy application upgrades and QFEs .aka. "patches". Our database back-end consisted of two databases: forums and dashboard. Each of them were deployed on a pair of mirrored SQL2008 servers. Database administration consisted of insuring backups and SQL Agent jobs completed successfully and in a timely manner. Since we inherited no documentation, I was tasked with creating the Operations Guide, KB articles and supporting documentation. PQO used One Note 2010 and stored the content on their SharePoint 2010 site.

It's interesting to note the names of the databases reflected the two user groups: the Windows Live, Hotmail, Messenger, Sky Drive folks and the MVPs who resolved the questions in the Dashboard application. The Support tickets we worked were submitted by our "customers", the contract MVPs. Their biggest gripe was what they called "latency", measured by the interval between pressing the button and getting a response. We found latency occurred by poorly written SQL queries which timed out the web servers, created run away SQL processes that maxed out the SQL servers. Our SQL v-team worked with the developers to restructure queries and add indexes as appropriate to improve response times.

At the end of the fiscal year in June 2011, my position was eliminated because the project funding was cut back to a "maintenance" status. Equipment on order was cancelled, support reduced as the entire application was being rolled into Microsoft Answers (www.answers.microsoft.com). This was a great group to work with.