On October 11 MapR and data Artisans held their first mutual Apache Flink Meetup in the Bay Area with more than 40 attendees. The Flink community enjoyed a packed agenda with detailed talks and time for socialization and networking. Those who weren’t able to attend the most recent Flink Forward conference in Berlin, were able to participate in variants of three of the best talks from the event over pizza and beer in the intimate setting of the Hacker Dojo.
Kicking off the night, Jeff Bean shared his impressions of Apache Flink since starting at data Artisans back in May 2018, from the perspective of a technical practitioner who’s new to stream processing.
We're at the @ApacheFlink Meetup in Santa Clara with @jwfbean @dataArtisans taking about basics about why you'd use Flink for steam processing. pic.twitter.com/aOymTKSEvE
— (((Ellen Friedman))) (@Ellen_Friedman) October 12, 2018
Ellen Friedman and Ted Dunning from MapR delivered great talks on the benefits of flexibly conceptualizing data (i.e. as data streams) and platforms that enable developers to do so.
Sonali and Shriya from Netflix spoke about using Flink to drive timely recommendations at scale. They discussed the issues they ran into, the attempted workarounds, and what ultimately worked for real-time joins.
You can see their Flink Forward slides here.
Thomas Weise from Lyft shared their experimentation with using Python to power Apache Flink with Apache Beam. You can find Thomas’ Flink Forward slides here.
Steven Wu from Netflix talked about using Apache Flink in the broader context of fault tolerance, including providing their users different approaches for recovery, such as Hive backfill (with a Flink source) or Flink rewind. See his slides here.
We deserve Steven an apology because we didn’t allocate enough time for his talk and the Meetup ran late because of it. Nonetheless, Steven’s talk was so good that most people chose to stay longer to ask him questions.
Here’s a social media snapshot of our Bay Area Apache Flink Meetup:
Hanging @hackerdojo with some pizza. Gonna learn how stream processing is done right. pic.twitter.com/1kC4ijivNh
— Jeff Bean (@jwfbean) October 12, 2018
at @ApacheFlink Meetup: Sonali Sharma & Shriya Arora from @netflix talked about how they use Flink now plus plans to 1.6. @ted_dunning @mapr spoke briefly about appropriate data persistence depending on your application scale pic.twitter.com/SWbScNpQbe
— (((Ellen Friedman))) (@Ellen_Friedman) October 12, 2018
Thanks to MapR and Hacker Dojo for enthusiastically working with us on this event. Our goal is to schedule Meetups and Training in key regions at a 4-6 week interval. If you live in the Bay Area and missed this one, we will be scheduling another Meetup soon. Don’t forget to join the Bay Area Flink Meetup Group to stay updated on upcoming Meetups in your area!
Don’t forget to sign up to the Apache Flink Training scheduled in San Francisco (November 6, 2018) and San Jose (December 4, 2018).
If you’d like to host a Meetup or Training in the future, please reach out. If you attended the Meetup, we hope you had a great time and thank you for coming around!