PASS Summit 2018 – Keynote Day 1 Live Blog

SeattleToday is the first full day of the PASS Summit in Seattle, Washington. This is the 20th year of the Summit, which brings together several thousand data professionals for a week of learning and networking.

This morning, I will be live-blogging the keynote. This post will be updated periodically through the course of the keynote. If you are not in attendance at the PASS Summit, you can watch PASSTV for a selection of content (including the morning keynotes) throughout the week.


This is the twentieth PASS Summit, and much of the branding around this year’s event is focused on the stories that have developed because of PASS and the Summit.

Since we are at the end of the year, it is PASS board election times. PASS president Grant Fritchey encourages everyone to vote, although this year’s election is unique in that there are three candidates for three open board positions.

The PASSion award is given this year to Michael Johnson, a data consultant and user group leader. Congrats to Michael!

Microsoft Announcements

Rohan Kumar takes the stage next to share more about the goings-on in the Microsoft data stack. He reminds the crowd that we are not only celebrating 20 years of the PASS Summit, but also the 25th birthday of SQL Server.

Kumar shared a bit about the hybrid story. In years past, Microsoft preached an all-in cloud strategy, but clearly many customers weren’t (and still aren’t) ready for that huge step. Similar to what was shared last year, Microsoft is embracing the hybrid environment, containing a mix of on-prem and cloud data sources on varied platforms.

Microsoft recently announced an early public preview of SQL Server 2019. That community technology preview (CTP) can be downloaded here. Among the highlights of SQL Server 2019 is support for data lakes, support for R and Python deployment to big data clusters.

Next up, Bob Ward and Conor Cunningham take the stage, in what Rohan Kumar calls “The Bob and Conor Show”. Together they demonstrate some of the performance improvements around latch hints in SQL Server 2019 by moving temp tables into Hekaton-based structures.

Kumar announced a small but very popular improvement. The vague “string or binary data would be truncated” message will be modified started in SQL Server 2019 to show the actual column name. This new behavior will make troubleshooting truncation or data type issues much easier. Thanks for that fix, Microsoft!

We also hear a bit about Azure Data Studio (formerly SQL Operations Studio), which is designed to give developers and data professionals a single unified view of the organizations data across different architectures. You can download Azure Data Studio here.

Also discussed is Azure Database Migration Service, intended to allow easy migration of data and metadata to cloud structures.

Rohan Kumar announces the general availability of Azure SQL Database Managed Instance on December 1st of this year, which will include the Business Critical SKU. Also, Lindsay Allen shares a demo on Accelerated Database Recovery.

Azure Cosmos DB’s new multi-master allows for near-real-time replication of data, and they guarantee 5 nines of availability. The new buzzword around this is “planet scale”.

Next up, Kumar moves on to discussing the modern data warehouse, showing the pipeline from source data through ADF through Azure SQL Data Warehouse. Among the announcements for Azure SQL Data Warehouse:

  • Row-level security
  • Custom maintenance windows
  • Workload importance (public preview)
  • Accelerated database recovery (ppublic preview)
  • CI/CD with SSDT

We now see a very brief demo which quickly selects six rows out of a petabyte-sized source data set, using around 4000 nodes. Definitely impressive for those cases where exceptionally large data sets require such fast selectivity. Also briefly shown is the workload importance feature.

Rohan Kumar now hands it off to a Microsoft partner (Taboola) who shows the use of Azure Data Explorer, which is now available as a public preview.

Next we learn about the public preview of data flows, which Kumar describes as eliminating the need for some complex ETL. He also announces that SSRS reports will now be available in Power BI. Patrick Leblanc takes the stage to show more detail about paginated reports deployed to the Power BI service, including an enhanced print experience.

Last up, we see a demo by a Microsoft partner from Exxon in which she demonstrates using Microsoft AI to recognize when someone is smoking near a gasoline pump. This is the kind of system that can literally save lives.

That’s a wrap

That’s it for the day 1 keynote. Lots of interesting and exciting announcements, with several things I want to follow up on after this week.

I’ll be live-blogging and tweeting tomorrow’s keynote – same time, same channel.

About the Author

Tim Mitchell
Tim Mitchell is a data architect and consultant who specializes in getting rid of data pain points. Need help with data warehousing, ETL, reporting, or training? If so, contact Tim for a no-obligation 30-minute chat.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.