May 14, 2014 (Wednesday), 7pm
Bldg 35/Kalaloch, Microsoft campus, Redmond
Overview of the SQL Server Infrastructure for Windows Error Reporting
Guest Speaker: LeRoy Tuttle, Principal Software Development Engineer, Microsoft OS Services Group
Windows Error Reporting (WER, aka “Watson”) is a flexible event-based feedback infrastructure designed to gather information about the hardware and software problems that Windows can detect, report the information to Microsoft, and provide users with any available solutions. The error reporting feature enables users to notify Microsoft of application faults, kernel faults, unresponsive applications, and other application specific problems. Microsoft can use the error reporting feature to provide customers with troubleshooting information, solutions, or updates for their specific problems. Developers can use this infrastructure to receive information that can be used to improve their applications.
The backend infrastructure that supports this service is a mission critical system built upon Microsoft SQL Server which leverages a wide spectrum of the features offered in the relational engine, in a highly available topology. The system is used to collect & process data received from nearly 3 billion front-end server transactions per day, and it publishes both raw and aggregated data for end-user queries, automation workloads, and data feeds.
This presentation will attempt to provide some sense of the scale of the system, and will offer an overview of the underlying SQL Server topology, the design goals and operational constraints, as well as some patterns, tips, & lessons learnt during implementation of these SQL Server feature areas:
• Horizontal & Vertical partitioning – isolating data volatility and query workloads
• SQL Server Replication – managing 50+ billion rows across multiple publishers & subscribers
• SQL CLR – optimizing ROI of user defined data types and functions
• ColumnStore indexes – managing data and queries for 250+ million new rows of data every day
• Author, Build, Deploy – leveraging code library patterns within SQL Server Data Tools (SSDT)
• Data Feeds – providing coherent data pulls across partitioned & distributed tables
That’s a LOT to cover…!! So, in the interest of time, we’ll focus on a few of the above areas that have the most interest for the audience in attendance; i.e. we’ll make a data-driven decision.
LeRoy Tuttle is a SQL geek; he has worked with every version of the SQL Server platform since v4.2. He has been working in various groups at Microsoft for the last 14 years; developing training & certification material, authoring content, developing code & architecting platforms, and living the database application lifecycle.