November 25, 2014

Report from OSCON 2008

So I’m back from OSCON 2008. As usually it is fun conference to attend. Though I could only come for couple of days (Wed/Thu) as things are too busy here.

Same as previous year I would note the conference is not as much backend and high performance focused as MySQL Conference or even Velocity, though this year there are number of interesting talks for example about Hypertable and lucidDB.

Honestly I could visit less talks than I wanted because I had too many things to do – talk to bunch of people, staff Sphinx .ORG booth (Andrew could not come so that was on me).

Sphinx both turned out particularly interesting – unlike couple of years ago on MySQL Conference there were a lot of people coming by not to ask what Sphinx is but happy users saying thanks for the product. I also had to explain many times why is it so fast – the two orders of magnitude speedups people often experience compared to MySQL build in full text search look too good to be true. I’ve clarified this comes from the design consideration which was built well exactly for the task while full text search in database systems if often afterthought.
At the same time I go a lot of questions about real time index updates – this is the number one feature people are asking about and indeed there is an agreement this is important feature to get. Though it requires a lot of effort to implement right – without sacrificing speed or flexibility.

The MySQL presence on this conference was also very interesting. There was a SUN booth with quite limited MySQL focus (though Monty,Brian,Jay,Colin gathered quite a good crowds). It was also interesting I have seen no MySQL Sales and Marketing guys on this conference just community guys and engineers – Monty, Brian, Steward, Jay, Giuseppe, Colin which is probably good choice. OSCON exhibition had enough Commercial OpenSource and OpenSource crippleware on the show :)

There also were no big announcements from MySQL on OSCON – the biggest one actually came somewhat outside – the Drizzle – MySQL “community fork” was surely the biggest splash on the conference around MySQL.

Monty’s Maria storage engine was only one which had dedicated session on OSCON (I did session which covered various storage engines) which was fun and very well received.

It was also good the session on MySQL Proxy was presented by Giuseppe and Ronald Bradford – it is good to let bias free independent consultants to speak.

My talk on MySQL Storage Engines was pretty well received and I had interesting crowd attending.

Speaking about MySQL vs PostgreSQL it is worth to say PostgreSQL “owns” OSCON – The PostgreSQL booth had more engineers and large crowd about it, plus EnterpriseDB had very large presence. Interesting enough there is not much of heated discussions going any more – it looks like PostgreeSQL and MySQL are largerly ignoring each other at the conference – PostgreSQL mentions Oracle, DB2 etc as the competitor they are after and MySQL does not put anyone as direct competitor in the talks I’ve heard.

About Peter Zaitsev

Peter managed the High Performance Group within MySQL until 2006, when he founded Percona. Peter has a Master's Degree in Computer Science and is an expert in database kernels, computer hardware, and application scaling.

Comments

  1. Yeah LucidDB is a pretty interesting project. If you are looking for a column oriented open source dbms, take a look at MonetDB too – they have some very promising benchmarks.

  2. LenZ says:

    EDB may have not mentioned or compared itself against MySQL at OSCON, but biased and partially incorrect product comparisons like this one clearly indicate who they are targetting:

    http://www.enterprisedb.com/downloads/whitepapers/PGvsMySQL_LowRes.pdf

  3. LenZ,

    No issues with calling that whitepaper biased etc etc. But are you sure MySQL doesn’t have any similar whitepapers? I think they do. And the webinars are pretty silly sometimes too.

  4. “independent consultants to speed.” –> “to speak” I hope.

  5. peter says:

    Roland,

    Yeah monetDB is another one though it seems like it is more academic. In reality it would be very interesting to see Column Orientieered storage engine for MySQL too… the problem is however such databases often need very different optimizer to perform well and MySQL is not really flexible in this aspect.

  6. peter says:

    Lenz,

    You’re right. This will be shame for EnterpriseDB for years to come. Mark has a good post on it
    http://mysqlha.blogspot.com/2008/07/postgresql-vs-mysql-according-to.html

    Interesting enough PostgreSQL guys contacted me and few other MySQL community members to review their MySQL vs PostgreSQL comparison papers – this is much better approach.

  7. peter says:

    Baron,

    You’re right. MySQL marketing materials are often full of crap too.

Speak Your Mind

*