MariaDB & MySQL community event, Apr 2014

Hilton Santa Clara, Santa Clara, CA
Thursday, 3 April 2014



This community event was hosted by the MariaDB Foundation with support from AccelerationDB. This was a free community event to complement the Percona Live conference. The community event will be a full day focusing on many things that will not be covered at Percona Live.

The themes for this community event are MariaDB 10.0 GA, High availability, and Performance. The talks are seen in the schedule below.

There was also a dinner that night at the Taste restaurant.

Schedule, Videos, and Slides

Here is the schedule of the talks with links to videos and slides of the talks.

09:00amWelcome to the MariaDB & MySQL community event
Michael 'Monty' Widenius
09:10amWhat's is new in MariaDB 10.0
Michael 'Monty' Widenius
Includes features comparison with MySQL 5.5 and 5.6
Slides: MariaDB-10.0-foundation-2014.pdf
Video: http://youtu.be/4Th7JeRxPlo
09:40amSpider in MariaDB
Kentoku Shiba
Slides: Spider_in_MariaDB_20140403.pdf
Video: http://youtu.be/k4O6jJ7ULng
09:55amMariaDB 10.0 optimizer
Sergey Petrunia and Sergei Golubchik, MariaDB developers
Slides: mariadb-10_0-optimizer-apr2014.pdf
Video: http://youtu.be/O-wWYLGCPrA
10:45amMariaDB and ScaleDB showcase
Moshe Shadmon, ScaleDB
Show case of how to insert continuously 1 Million rows/second while doing concurrent reads with MariaDB and ScaleDB
Slides: Scaledb-2014-v4.pptx
Video: http://youtu.be/XQsN6M-ZrWI
11:30amMySql at Tumblr
Bob Patterson Jr, tumblr.
Tumblr manages more than 250+ database servers with 55+ TB of data. We accomplish this through our architecture, tools and process. During this talk I will discuss how the databases at Tumblr are laid out and go into detail about what tables and data lives on each type of machine. Also I will go through how we back our dashboard(news feed) in our application and database layer. Finally I will cover how we manage and deploy our database configuration so our application knows where to retrieve and write data. I will discuss briefly our memcached stats and how we use memcached with over 3 million gets per second and 100 thousands sets per second to support our databases.
12:15pmLunch break
1:00pmMySQL MHA and Continuent Tungsten shootout 2.0
Alex Alexander, SpringbokSQL
1:45pmVitess, CAP, and Hitting the sweet spot @YouTube
Sugu Sougoumarane, Google
Slides: Vitess_and_CAP.pdf
Video: http://youtu.be/rfU-SwI7LWg
2:30pmBreak
2:45pmBig data with NEO (distributed database powered by MariaDB) and Scikit-learn
Jean-Paul Smets, Nexedi
Slides: NXD-Wendelin.Presentation.MariaDB-005-en.pdf
Video: http://youtu.be/gdKhm2p45UI
3:30pmInfiniDB overview
Jim Tommaney, InfiniDB
Learn about InfiniDB to deliver real time reporting and analytics at scale. InfiniDB is an open source columnar analytics engine available via GPL v2 with no restrictions on syntax, scale, or performance. The InfiniDB custom distribution of work and pure columnar storage deliver is well-suited for specific workloads (analytics, reporting) but does not replace MySQL for transactional and other workloads.
Slides: InfiniDB_Overview_Q2_2014.pptx
Video: http://youtu.be/_qfOnaUc_So
4:15pmmysqlv8udfs - Running JavaScript inside MySQL
Roland Bouman
MySQLv8Udfs provides a number of MySQL UDFs that integrate Google's v8 JavaScript engine into MySQL. With mysqlv8udfs you can run JavaScript directly in the database. This allows developers to do all kinds of useful things that are hard or inconvenient to achieve with either SQL, stored routines or UDFs. Examples include JSON document generation and manipulation, powerful regular expression support, and data transformation and transport across MySQL instances. For particular types of tasks, especially CPU-intensive ones, mysqlv8udfs also provides substantial performance benefits as compared to stored routines, and even plain SQL expressions. MySQLv8UDFs is free and open source. Code is available here: https://github.com/rpbouman/mysqlv8udfs
Slides: MySQLv8UDFs-UC2014-samples-hidden.pdf
Video: http://youtu.be/Wj0PDWVauqI
5:00pmSpeeding up MariaDB on Fusion-IO
Jan Lindström, MariaDB developer
This talk will share compression improvements in MariaDB optimized for flash using native interfaces. Get the capacity benefits of compression, without incurring the performance overhead and complexity of packing and re-packing the compressed blocks into one-size-fits-all blocks. Instead, rely on underlying thin provisioning to allocate only what is used and provide efficient garbage collection of unused blocks. We will present some alternatives to address challenges presented and preset performance results from TPC-C and LinkBench.
Slides: Speeding_up-MariaDB-on-Fusion-IO2.pdf
Video: http://youtu.be/QFJbNN3CZIk

This community event was partly inspired by the Baron Schwartz blog post announcing the Percona Performance conference in 2009. We believe that there should be more free conferences about MariaDB and MySQL that will allow anyone to attend and participate.

You can use the 'foundation 'at' mariadb (dot) org' email address if you want to help sponsor a community event. All sponsorships go to pay for the event venue, hotel, and travel for speakers (who could not otherwise afford to attend). In the event there's anything left over the Foundation will use it for further community activities.