Notes from MySQL Conference 2012 - Part 1, the soft part
I have finally recovered from my trip to Santa Clara enough that I can scribble down some notes from this year's MySQL Conference. Writing a travel report is part of the deal where my employer covers the travel expense, so even if many people have written about the conference, I need to do it too. And judging from the many posts for instance from Pythian's direction, Nokia is perhaps not the only company with such a policy?
There has usually always been something that can be called a "soft keynote". Pirate Party founder Rick Falckvinge speaking at a database conference is a memorable example (I still keep in touch with him, having met him at the Hyatt Santa Clara). This year there was one less day, and therefore less keynotes. The soft keynote was therefore taken care of by Baron using some time out of Peter's opening keynote. Baron's talk was an ode to the conference itself, underscoring the meaning of the conference beyond just learning about technology. Sharing his own journey from a numb ASP.NET coder ("a good day at the office was when I changed a table based layout to pure CSS ...but nobody else seemed to care.") to his role today, he challenged people to network, make new friends and new revolutionary ideas. To me, it was a great opening keynote (and quite obviously would have made less sense on the last day of the conference). The talk, including Peter's part, is available on Percona.TV.
Inspired by Baron's approach, I will also begin my travel report with a post containing softer, meta-level, observations.
The Italian family united again (almost).
The above image was taken on the last day of the conference, which was an addon day of satellite events. To me it represents everything about this year's conference. I worked together with all four of these Italian gentlemen at MySQL/Sun. Now they work at different companies, yet are united by the MySQL community and more concretely by this conference. From left: Ivan Zoratti, former Director of Sales Engireenging at MySQL, now CTO at SkySQL, lives in London; Marco Tusa, former manager of South Europe MySQL consulting, now Head of MySQL practice at Pythian for which he lives in Ottawa; Massimo Brignoli, former telco-team and OEM sales engineer at MySQL, moved to SkySQL just a month ago (citing his desire to travel to the conference as one reason to resign Oracle - I think I know at least one person that has done so every year since 2010); and Giuseppe Maxia, former lead of MySQL community team, now Director of QA at Continuent. The last two still live in Italy. I suppose I should mention as missing from the picture Luca Olivari who still works at Oracle and was therefore not able to attend - you were in our thoughts nevertheless.
Massimo was my mentor when I joined the telecom team at MySQL, so it was truly great to see him again. I also met JJ and Anthony Flynn from SkySQL. You have to feel for Anthony. He originally came to the open source database scene selling BerkeleyDB, and was acquired by Oracle. Eventually he left Oracle to sell MySQL, but was acquired again by Oracle only few months later! Third time's a charm? If you also count Johan and Vinay of Severalnines, and Hana who is at SkySQL though didn't attend the conference, this means that almost everyone of my day-to-day co-workers from MySQL-Sun times have now left Oracle but are still active in the MySQL community. Out of people I worked with on a daily basis, only Matti and Prem are left at Oracle.
But whether you are still at Oracle or have left, it seems the bonds keeping us within the MySQL community are still strong, as very few have left the community altogether. I know some people, even in this picture, planned to "move on", but what can you do - this is just a great community to be with, so it's just not easy to find a better industry to work in! Even people like Sarah Sproehnle and Kostja Osipov who have gone to work with their respective NoSQL systems came back to be at the conference.
The release announcements
MySQL would traditionally throw out some press releases for the morning of every conference. In recent years they've also started accompanying those releases by bombarding Planet MySQL with 15 simultaneously published blog posts. Many people, including myself, have tried to make the point that releasing all blog posts at the same time is self defeating, and in any case the morning of the first conference day is the worst time to do so (because we are not at home reading them).
Considering that Oracle was boycotting the conference, it was somewhat of a shock to wake up on Tuesday, pick up my phone and see the MySQL 5.6.5 release announcement at the top of my Twitter feed. For a moment I thought I had woken up in the wrong year. (And remember, I have weird memories about waking up to MySQL conferences from past years...) Did they do it accidentally out of habit, or was it an attempt to divert attention away from Percona, or was it an underground greeting from people who's hearts and minds were with us at the conference, even if they couldn't attend in person? I chose to interpret it as the latter. In any case, made for good speaking material in mine and many other people's talks.
But seriously, in terms of the code base it seems we are doing better than ever. Do you remember that at MySQL AB times it could take 3 years between releases? We still did a conference every year, and talked about Falcon and Maria engines and all kinds of other vaporware, some of which was never released... This year, all 4 major forks made significant releases: MySQL a labs release (which we in retrospect might be calling an RC, if they do a GA release later this year), Percona their XtraDB Cluster, MariaDB finally did their 5.5 release and Drizzle did their 7.1 release.
In addition some 3rd party software like Tungsten have finally taken a quite strong position (it seems to me the world has changed in a favorable way for Tungsten, now everyone has 5 data centers on their credit card...). All in all, no wonder the conference was so good, because we have so much brilliant technology to cover.
Things back to normal now?
Coming home from the conference I decided that in my memoirs (which I'm pretty sure will not be published!) I will count this conference as a major milestone. Like many others, I have been quite engaged in the quest to save our MySQL community from collapsing - not so much due to Oracle itself, but due to the inevitable ripple effects set off by the acquisition in 2009. Above I have covered the three main aspects of that quest: The MySQL conference is still happening, we are still there and MySQL software continues to live on.
The existence of the conference have been under doubt since the Oracle acquisition. But it seems that phase is now behind us. This year the conference was very good, and attendance and number of exhibitors was up from the past 2 years. Most importantly, Percona has already announced the dates for the conference to happen in 2013. This is the first time since 2009 that we already know that next year's conference will definitively happen, and it will be a great one.
So, I've decided that at least for my personal quest, I will hoist my Mission Accomplished banner. What does it mean? Probably not much visible. I'll continue to blog, code, talk at conferences and enjoy life as I always do. But probably with even more focus on the "enjoy life" part!
Mårten's was surprisingly filled, and to the point, with MySQL substance. But let's face it, regardless of what he said, it was just cool to have him there doing a keynote again. When Sun acquired MySQL AB Mårten used to tell us: plus ça change, plus c'est la même chose (the more things change, the more they stay the same). It is the French way to say that change is constant and inevitable. Mårten himself was of course always quick to embrace change, and at least to me seeing Mårten speak at a conference now hosted by Percona speaks those words stronger than anything else I could think of.
There are a few foods and candy that we typically bring with us when meeting Finns living in exile. One is a Xylitol sweetened chewing gum called Jenkki (the xylitol is good for your teeth, a Finnish healthcare innovation). I happened to buy some eucalyptus flavored ones to take with me to this conference, and had planned to give Mårten a pack, but forgot to do so.
With Peter's keynote being very short, and someone like Tomas Ulin completely absent, I came to look at Mark's keynote as most closely resembling a "state of the MySQL roadmap" kind of keynote you are used to listening to at a conference like this. It was of course completely from a Facebook point of view, but since we are a diverse community without a centralized Linus Torvalds figure, then that is the closest we can do.
Key figure from Mark's keynote: Thanks to the work done by his team, Facebook can fit 10x more data per MySQL server than without. Knowing that they have some thousands of MySQL servers, it is easy to make a calculation of that ROI.
You have to pardon another history lesson, but listening to Mark again remind me of the change in the past 3 years. In April 2009 Mark had not yet joined Facebook and Facebook was famous for running with just 2 DBAs. (I'm sure in 2009 they had more than 2 :-) Today the MySQL team at Facebook - when including all 3 MySQL teams - is a couple dozen strong and is the home of many rock stars of the community. In terms of developing MySQL their work is roughly equal to that of a Percona or a MariaDB, and this is counting purely in man-hours. In terms of symobolic value and maybe even relevance it is higher.
It is just yet another example of how our community has changed for the better: we have a diversity of strong contributors. (This year both Twitter and Google have also released in-house developed MySQL code.)
Quite many people said something like this with a surprised tone: "Isn't it great we can have such a great conference even if Oracle wasn't part of it..."
To me it wasn't a surprise at all. I mean, who could cover MySQL Cluster better than Johan Anderson? Or who could cover the NoSQL-to-InnoDB APIs better than the inventor of the concept, Yoshinori Matsunobu? Who could cover the optimizer better than Monty's team, or even the Percona team?
Personally, I always thought that if Oracle wanted to harm this conference, then they really fumbled by giving it away, they should have instead tried to control it as tightly as possible. On the other hand, if Oracle wanted to give the MySQL community a great conference, then this is exactly what we have got!
Even so, I wouldn't mind having Oracle back as a participant. I hope they consider it next year, lest we'll see even more resignations in 2013.
There is a saying that whenever a large number of MySQL'rs meet, disruptions happen. By this we eventually came to refer to quite literal things, like volcano's erupting and such. I also remember there been many MySQL company meetings were more than half of the company reported in sick after coming home!
As far as I know, we were this year saved even from this kind of drama, and everyone got home safely and peacefully. Maybe Percona just is better at avoiding drama?
To offset that ease, I had this year taken the whole family with me to see the California I visit every year. We spent 4 Easter days around San Francisco before the conference began. Coming home we then traveled a group with 3 adults and 3 small kids, all of whom had either flu or stomach flu or both. So, even on this account, at least for me personally, everything was just like before.
Speaking of lack of drama, there also was nobody that acquired Percona or anyone else at this year's conference. Instead, the week after SkySQL announced closing a 4M investment round and Mårten's company Eucalytpus announced a 30M investment round. (It's somehow qute to see them announced on the same day. Pretty much everyone at SkySQL used to work for Mårten at MySQL.)
So I just wanted to congratulate both startups and wish them good luck in spending the money!