See you at O'Reilly MySQL conference starting Monday

I've been to (only?) 2 previous MySQL conferences. I never spoken before, I actually tried to avoid it plus I didn't feel I had any hardcore topics worthy of presenting. Last year all MySQL Sales Engineers were told to submit a talk, so I submitted one that I knew would not be accepted ("no vaporware rule"), and it wasn't. Hey, I obeyed our boss more than my collagues who didn't submit anything! I did feel that for me it was a more useful way to spend time with my customers who were at the conference, and just network.

This year I am doing a talk with Igor. I didn't submit anything this year either, but since Monty Program is a sponsor, we get an extra slot to talk about our "Products and Services", so we will present some of the engineering work we've done for customers who wanted a feature or optimization for their MySQL or MariaDB server.

Here's what I'm looking forward to next week:

MariaDB booth

As the second biggest sponsor, we'll have a killer sized booth in the expo. Part of the booth will be a MariaDB community area: chairs, tables and electricity and you are free to sit down, with or without laptop, and discuss anything MariaDB related. If you feel you know about MariaDB and are part of it, you are free to tell other passers by about it.

Monty Program and MariaDB talks

9:15am Wednesday, Keynote
State of MariaDB, Michael Widenius

10:50am Tuesday
MariaDB: What? Who? How? Why? Where?, Kurt von Finck

3:05pm Tuesday
Valuable MariaDB Features That Somebody Paid For, Henrik Ingo (Monty Program), Igor Babaev (Monty Program)

5:15pm Tuesday
MySQL Plugin API: New Features, Sergei Golubchik (MariaDB)

11:55am Wednesday
MariaDB Release 5.1; What Is It and What to Expect From It, Michael Widenius

8:30pm Wednesday BOF
The MariaDB Knowledgebase, Bryan Alsdorf, Daniel Bartholomew

2:00pm Wednesday
The MariaDB/MySQL Query Executor In-depth, Timour Katchaounov (Monty Program AB)

11:55am Thursday
New Query Engine Features in MariaDB, Sergey Petrunya (Monty Program Ab)

Also MariaDB related talks

2:00pm Tuesday
A Practical Guide to the PBXT Storage Engine, Paul McCullagh (PrimeBase Technologies GmbH)

MariaDB includes PBXT, but until now we have not made a lot of noise or benchmarks about this interesting new engine. It will be great to hear the news from Paul himself.

5:15pm Tuesday
OQGRAPH engine: hierarchies/graphs inside the database made easy, Arjen Lentz (Open Query), Antony Curtis (Blizzard Entertainment)

Still early alpha stage, but a graph engine is an obvious idea for a storage engine. Demand and popularity is of course driven by the rise of social networking. There are other standalone graph databases, but this one will be in MariaDB 5.2. (experimental only, but shipped as a plugin)

Talks I'm personally interested in (assuming I can clone myself to go everywhere)

8:30am Tuesday, Keynote
State of the Dolphin, Edward Screven

Obvious one, everyone is looking forward to this. I've seen Edward speak once before, but the circumstances of that event constrained both him and the audience listening, so I've decided not to pass judgement yet, but will think of this as his first public MySQL appearance. I really look forward to hear what will be announced and what he will say in general. As far as I know Mikael Ronström has not blogged anything since December, suggest that we'll see a replay of last year: Excellent benchmark results in something that has been worked on secretly (meaning, no public launchpad trees) for months and is released only now to generate buzz that a MySQL version is slightly better than anything else out there. I didn't understand it last year, don't understand it now, since the week after the conference we can just share the improvements between us anyway. Anyway, I look forward to MySQL finally actually releasing those improvements to all of us, they have had a big and talented performance team working hard for a year now. It will be nice for the engineers to finally get public credit for their work.

Other rumours: I'm betting MySQL 5.5 will be announced as stable. Engineers I talk to are doubtful. There's also a rumor about something related to InnoDB+ will be announced, such as a closed source database server based on InnoDB. We'll see what it's about - maybe.

9:15am Thursday
The Engines Of Community,
Jono Bacon (Canonical Ltd)

Ubuntu certainly knows how to do Community and Jono is a world thought leader on the topic. As vibrant as the MySQL community is, I always felt we are not really tapping into all the energy that is there. Compared to something like Linux, MySQL is... well, let's just say things could be much better still. Drizzle was able to tap into some of the pent up energy, but for MySQL, MariaDB... we still have much to learn. I really welcome this angle of the Thursday keynotes. (I know Kaj and Sheeri will speak about this topic too, but especially look forward to Jono perhaps bringing in some fresh insights we can learn from.)

2:00pm Wednesday
Using BlackRay As a Storage Engine, Felix Schupp (SoftMethod GmbH)

Just because I've met Felix twice before, even danced with him once at a SourceForge party :-)

3:05pm Wednesday
MySQL Cluster and Pushdown-joins (In Pursuit of the Holy Grail), Jonas Oreland (Sun Microsystems)

Last year the MySQL Cluster team released the multi-threaded data node architecture. I worked with a customer who insisted on Sparc/Solaris servers, and we saw really great performance with that. We were essentially able to provide the same level of crazy performance as "Incumbent Database with shared disk clustering", except that there was a big hardware savings gain when you could remove a big SAN and Texas Ramsan, since MySQL Cluster would do the same with just 6 normal disks.

This year Jonas and the team have finally finished the other feature we were always drooling for, namely Pushdown Joins. This means you can now do some JOINs that were inefficient before, and MySQL Cluster is able to send the entire operation for processing inside a data node, rather than doing several network roundtrips for each part of the query. Having seen Jonas' teasers, it looks like the benchmarks are as good as we believed they would be (or better?) This feature would have made the above mentioned benchmarks even better, and would have allowed the customer to add more JOINs if needed.

2:00pm Thursday
SSD Deployment Strategies for MySQL, Yoshinori Matsunobu (Sun Microsystems)

Have not worked a lot with SSD myself, so the best way to catch up is to listen to others.

Some of Facebook's talks

A funny thing

Monty Program crew will be living in the Hotel called Avatar. I hear it's really amazing: everything in the hotel is in 3D!

A happy thing

That the conference is happening and turning out greater than ever! That we can meet face to face and discuss where things should go the next year.

Previous years the conference was mostly a learning and socializing opportunity, but this year I've been involved here and there, so it is a great feeling to see it all come together finally!

See you there!

How appropriate. Right when I published this post, opensource.com published Community-building tip: surprise is the opposite of engagement by Chris Grams:

"Surprise is the opposite of engagement."

This may be one of the most simple, brilliant things I have ever heard someone say when it comes to creating engaged, active communities.

When we talk about building communities the open source way, we often mention transparency and openness as critical elements of any community strategy. But when I saw this quote, it reminded me why transparency and openness are so important.

When we are open with people, we avoid surprising them. We keep them in the loop.

Nothing kills someone's desire to be an active contributor in a community more than when they feel like they've been blindsided.

This is exactly what I felt was weird last year with the 5.4 announcement. I couldn't quite express it in words, but this is it.

(Since I worked at Sun last year, I knew about the 5.4 work. So I wasn't one of the ones to be surprised, I just felt it was a weird way to develop and release open source software, that's all.)

I am in the same hotel. I think Domas is too. That means one thing, room service courtesy of MPAB. I wonder if they have good scotch.

Maybe there was a no vaporware rule, but there certainly hasn't been a we-hope-to-release-it-this-year rule where features were presented on successive years as the release was delayed.

About the bookAbout this siteAcademicAccordAmazonAppleBeginnersBooksBuildBotBusiness modelsbzrCassandraCloudcloud computingclsCommunitycommunityleadershipsummitConsistencycoodiaryCopyrightCreative CommonscssDatabasesdataminingDatastaxDevOpsDistributed ConsensusDrizzleDrupalEconomyelectronEthicsEurovisionFacebookFrosconFunnyGaleraGISgithubGnomeGovernanceHandlerSocketHigh AvailabilityimpressionistimpressjsInkscapeInternetJavaScriptjsonKDEKubuntuLicensingLinuxMaidanMaker cultureMariaDBmarkdownMEAN stackMepSQLMicrosoftMobileMongoDBMontyProgramMusicMySQLMySQL ClusterNerdsNodeNoSQLNyrkiöodbaOpen ContentOpen SourceOpenSQLCampOracleOSConPAMPParkinsonPatentsPerconaperformancePersonalPhilosophyPHPPiratesPlanetDrupalPoliticsPostgreSQLPresalespresentationsPress releasesProgrammingRed HatReplicationSeveralninesSillySkySQLSolonStartupsSunSybaseSymbiansysbenchtalksTechnicalTechnologyThe making ofTransactionsTungstenTwitterUbuntuvolcanoWeb2.0WikipediaWork from HomexmlYouTube

Search

Recent blog posts

Recent comments