Top Session Picks for Percona Live

The Percona Live 2017 conference starts next week, running from April 25 to April 27 in Santa Clara. As a vendor in the broad MySQL ecosystem, Percona does a great job bringing together lots of users and vendors that continue to push the envelope of the MySQL project. This will be my second year representing Clustrix at Percona Live, and I know it will be a whirlwind. I am in the midst of organizing my schedule of sessions and booth times, and thought I would share my top picks for sessions to attend and why.

There are hundreds of sessions to choose from, so my choices are of course biased. At Clustrix, we focus on organizations who love MySQL, but have run out of capacity or are looking to avoid extreme measures of complexity. So, topics of extreme scale, or ideas that reduce complexity are a natural filter to prioritize my session list.

One System to Fit Them All: Shared MySQL Hosting at Facebook
Tuesday, 25 April – 11:30 AM – 12:20 PM @ Ballroom F

Facebook is a legendary user of MySQL. It is hard to imagine if any other company has put MySQL to use at a larger scale. Of course, their success comes from a cadre of ingenuity and hard work. Hearing their war stories is a can’t-miss.

Securing your MySQL/MariaDB data
Tuesday, 25 April – 1:20 PM – 2:10 PM @ Ballroom E

As data is moving to the cloud, security for data has become one of the hottest, fastest moving topics. If the MySQL and MariaDB community have anything new to say on the subject, I want to know about it.

Demystifying Benchmarks: How to Use Them To Better Evaluate Databases
Tuesday 25 April – 2:20 PM – 3:10 PM @ Room 210

Full disclosure: This presentation is by Peter Friedenbach, Performance Architect at Clustrix. He is not talking specifically about ClustrixDB, though. He’s a classic, methodical researcher that sets up benchmarks and tests (like this one on interrupt handling on AWS) to empirically derive quantitative results. Benchmarks are an evolving art, and Peter is a master. He wrote this presentation just for this audience, so I can’t wait to hear his guidance.

Exhibit Hall Happy Hour
Tuesday 25 April – 3:00 PM – 4:00 PM in the Clustrix Lounge

After this session there is a break, and Clustrix will be hosting a happy hour in the Exhibit Hall. So I’ll be headed there for a drink and to chat, as will Peter. Find us at Booth #414 right next to the Clustrix Lounge.

MyRocks Engineering: deploying a new MySQL storage engine to production at Facebook
Tuesday 25 April – 5:15 PM – 5:40 PM @ Ballroom B

Facebook has made a lot of progress with their RocksDB KVS, and now integrating it with MySQL via the storage engine framework. I’m keen to see how much RDBMS functionality they’re able to maintain when deploying it in production.

Birds of a Feather
Tuesday 25 April – 6:00 PM – 7:00 PM

With four short lightning talks from the likes of Percona, Oracle/MySQL, Dropbox, Facebook and Github, the names alone will bring me to this room. The topics are pretty varied too, and with the format of a BoF, I want to be part of the conversation.

Amazon Aurora: A deep-dive into what’s new
Wednesday 26 April – 1:00 PM – 1:50 PM @ Ballroom F

I’ll admit I am a skeptic on Aurora. It is still constrained to a single write-server, no matter how robust its AWS storage layer is. With write scale-out not on the current roadmap,  I think it will remain a bandaid to capacity limits that MySQL is constrained to as well. Still, if there’s something new, I will pay attention.

Rolling out a Database-as-a-Service layer at Ghent University
Wednesday 26 April – 2:00 PM – 2:50 PM @ Ballroom D

Admittedly, I am not familiar with Ghent University’s project. From the title though, I am interested in how this can improve on AWS RDS MySQL, and why they chose to do it on their own.

Operationalizing 19 years of infrastructure via MySQL-as-a-Service
Wednesday 26 April – 4:30 PM – 4:55 PM @ Ballroom A

GoDaddy undoubtedly possesses a wealth of war stories for deploying MySQL. I am keen to hear how they automated their way out of the “tremendous time and energy to operate an aging, heterogeneous MySQL infrastructure that struggles to keep pace with the business.”

Lessons learned while automating MySQL deployments in the AWS Cloud
Thursday 27 April – 11:00 AM – 11:50 AM @ Ballroom D

This session sounds like a showcase of complexity and out-of-the-box thinking to bend MySQL scalability limits in the cloud. I’m always curious to hear the realities about how far compensating takes real users.

How Avoids and Deals with MySQL/MariaDB Replication Lag
Thursday 27 April – 12:50 PM – 1:40 PM @ Ballroom E

Clustrix had to rewrite and parallelize our replication parsing because MySQL wasn’t fast enough. It will be interesting to see if they really made it fast enough for our standards.

Moving from sharded SQL to Cloud Spanner: Opportunities and Challenges
Thursday 27 April – 1:50 PM – 2:40 PM @ Room 210

While this title implies moving from a sharded MySQL instance to Spanner is a clear path, there are actually quite a few challenges to migrating a MySQL workload to Spanner, and sharding just adds another layer of complexity.  It should be very interesting to see Google’s recommendations.

Of course, there are a lot more sessions, but I also have to balance being around the Exhibit Hall to discuss ClustrixDB with interested attendees. For more information on sessions, you can look to build your own list by looking at the Day 1, Day 2 and Day 3 published schedules.


If you haven’t signed up yet and want to go, you can also use discount code “CLUSTRIX15” for 15% off registration on the registration page. Hope to see you there!