So I made it to PASS Summit 2018. After a flight from an airport with one gate- yes, you heard me right, one gate. No Wi-Fi, no connectivity and four employees at the airport. It was a new level of disconnect.
After a number of parties last night, I’m at the bloggers table for the first keynote this morning. The first keynote started with an energized talk from PASS president and friend, Grant Fritchey. It was a gracious discussion about the dedication of those in the community and power of those involved.
The theme is #V20, the newest version of PASS Summit, the 20th anniversary of the Summit conference and the same for SQL Server! As someone who’s an old-timer of the conference circuit, I love the maturity of the event here. Its commitment to the community, diversity and inclusion and the technology. It is an integral part of the event, never a second thought.
Rohan Kumar was the first day keynote. Rohan, (@RohanKData) is in charge of the engineering strategy behind Azure. He discussed in his talk what I’ve been telling folks for awhile now- 80% of companies are on a hybrid/multi-cloud environment. Microsoft seems more aware of this than other cloud providers as they have Office 365 along with other Azure products. They know how many are using Azure, even when they only see their data or big data platforms as their ONLY cloud. Microsoft isn’t threatened by hybrid and instead embrace it. There will always be data that MUST be held on-premises. It may be due to policies, archaic systems, but it will be so. As a cloud provider, Microsoft seems to understand, train their models to expect this.
All The DATA
He spoke at length to the importance of java working alongside with TSQL in the database, ability to have R and Python for analytics and more. After he finished filling our brains with information about the newest SQL2019 release, Bob Ward came on Conor Cunningham demonstrated how not to receive waits in tempDB in SQL 2019 in the midst of the keynote.
I have so many customers that have source data from Oracle, (of course I do, I worked in Oracle for how long? :)) unstructured and structured data that I can’t even keep track of it all. They’re tested with challenges of how to pull it into one source and build value from this data. Right now, they’re using less than 10% of their data, but what I’m seeing with SQL 2019, shipping with Spark, using the Azure Data Studio, that I could take this data and query across multiple data sources, no matter if CSV files, Oracle, SQL Server, big data and build joins, use python and actually get the value from data they were never able to do before.
Rohan then went into Managed Instance, (one of the most awesome recent features to come out in Azure, IMHO) which offers much of the PaaS benefits with the option to easily migrate from on-premises environments, (yeah, those 2008 environments have until July 9th, 2019. If you have trouble remembering, my birthday is July 8th. I won’t forget the date… :)) If you had concerns about mission critical workloads, fear not, December 1st the new mission critical workload version of managed instance will be available. That gives you the time to start your migration project now.
Recover Without the Wait
I was incredibly impressed with the Accelerated Database Recovery, (ADR) demo. I am aware of the power of snapshots, built into the SQL 2019 product in Azure to eliminate wait time on recovery. The question for me if this will be built into a cloning mechanism for environment deployment to secondary development and test environments? It will be interesting to monitor this feature as it matures.
Power BI- Reporting Server
Paginated reports will always have a place in people’s hearts. I have so many customers that have been waiting for the integration of RDL, (reporting server paginated reports) into Power BI. Power BI is going to rule the world, (I can say that, I’m in that group… :)) but the ability to create the report you need, no matter if its an interactive report with visuals and dashboards OR paginated is incredibly powerful for the customer. Its now here and I can finally make my customers happier by telling them they can do it all from Power BI.… Oh yeah, that Dataflows feature came out, too… 🙂
There was a cool demo of how machine learning with Spark that used Shell data and a sample camera that could detect a customer with a (pretend) cigarette and alerting the danger of the situation. It was a realization that this isn’t science fiction. We’re there now and I plan to be there in the future to see how machine learning is building more with our data and our technology.
Also published on Medium.