PASS Summit 2017 is now well behind us and there has been good time to reflect on this years conference. First of all, I have to say that Seattle as a venue is a good choice, even though for some of us that is a long way to travel. To me, it is about 20 hours from SEA-TAC to my home with the flights and driving, doable but not necessarily pleasant. As for the conference itself, I feel that it keeps getting better every year. It has definitely changed a quite a bit from my first Summit back in 2013. But so is the world where data platform professionals live, and the products we work with.
Here are some highlights from the PASS Summit 2017!
Keynotes
I go to the keynote sessions every year, it is always nice to get some background info on how PASS is doing and what they have planned for us in the future. Seeing how much they invest in providing the training for us, it is easy to get behind the “by the community, for the community” idea. In past few years we have had the pleasure to see some excellent keynote presentations by David DeWitt and Rimma Nehme, unfortunately Dr. DeWitt was not presenting this year but we did get some good “momsplaining” on Cosmos DB details by Dr. Nehme.
I am also constantly (and pleasantly) surprised by the professionalism of Microsoft presenters on these keynotes. They clearly make the effort to give the audience a good and entertaining show, with usually quite a few jaw-dropping demos. I think we are also seeing lot of new Microsoft there, with their ability to quickly adapt and to change, f.ex. the narrative going from On-premise or Cloud to On-premise and Cloud. I was also quite surprised to see AWS and Google sponsoring Microsoft Data Platform conference, times are definitely changing!
Cosmos DB is quite the something! Globally distributed, scalable, low latency and support for multiple NoSQL datamodels (key-value, graph, column, document) with SLAs on availability, throughput, latency and consistency. Oh yes, it also has 5 different consistency models also. And there is also now a tool for Windows, Mac and Linux to operate your databases with called SQL Operations Studio.
DevOps
This is really something that caught my eye, there was a quite a good number of sessions around the DevOps this year. Also Microsoft had something to show around the topic, with the SQL Server now supporting Docker (and Linux) you can rather easily setup testing and development environments from your version control by adopting the Infrastructure as a Code approach. This is something we have looked at also, so it was nice to see that Microsoft has similar stories to tell.
Azure
Microsoft has done some impressive work with Azure, the are 36 Generally Available regions and 6 more under works. There is also a very special region in Germany, the “Black Forest“, which is logically and physically separated from the rest of the Azure. This site is handled by European Data-Trustee (Deutsche Telekom) making it very interesting option for companies that might be worried of their data ending up outside EU borders. Prices are bit higher, about 10-15 percent, but it might be worth it.
Microsoft has also made a huge effort together with Facebook to build a globally spanning network to connect all these regions and data centers together. They are also working on some very interesting projects to simplify modern data centers, like running the servers on natural gas! It must be nice to have resources 🙂
And so much more!
I did manage to attend plenty of sessions, topics ranging from Azure, DevOps, Monitoring, Master Data and whatnot. I did also have time to visit the vendor booths to check out what the exhibitors had to offer.