MySQL Conference with Severalnines, Xtrabackup, Nokia and much to learn

hingo's picture

It's now 2 hours until I depart for San Francisco and then Santa Clara for this year's MySQL Conference. Can't wait to meet everyone and to learn about cool new things (2 solutions for parallelizing replication, Drizzle, Xtrabackup...)

This year I am involved in two BoF sessions: Severalnines ClusterControl with Johan Anderson and Xtrabackup Manager with Lachlan Mulcahy. Both of these are about brand new topics, this being the first time they are presented, so I'm excited to be working with these guys.

Severalnines ClusterControl will have interesting news for both users of MySQL Cluster as well as MySQL replication. When I worked with MySQL Cluster, I would always use the configuration files and management scripts from severalnines.com. Early on this was a necessity, now it's more a convenience. Early versions of MySQL Cluster were really hard to configure right. Johan was the only one who really knew how, so using his configuration was the best way to make it work. Today MySQL Cluster has much improved defaults and easier to understand options, but I find that manually logging in to each server to run the same steps of installation, startup, whatever... quickly becomes boring, so controlling all this from one single command line is the right way to do it.

The new stuff is that the same ClusterControl software now supports also MySQL replication topologies. This is potentially interesting to a much broader audience than MySQL Cluster ever was. This is certainly on my list of things to check out when home from the conference. The other cool thing is added support for EC2 and Rackspace clouds. Together these two features now make it really simple to deploy a MySQL replication cluster with just making selections on a web page. Severalnines creates the needed amount of replication slaves for you, and the cloud vendor the needed amount of servers, while you mostly sit and watch. Things have changed since I first learned to do MySQL replication.

Xtrabackup Manager was launched just recently by Lachlan Mulcahy, also a former MySQL AB collague. We all know Xtrabackup from Percona, it is an open source alternative to Oracle's InnoDB HotBackup, allowing you to take online backups of MySQL. Xtrabackup Manager is the tool that allows you to schedule when it should automatically run Xtrabackup for you, where the MySQL servers are and where backups should be stored.

This is a very new project so it lacks a few things like, say, a nice user interface. It does already take backups, which is a plus for a backup management tool! But I was looking for something like that for my own needs, and have spent some time with it. Lachlan's code was relatively bug free and carefully written and the approach was very close to what I had in mind myself. Lachlan has also been very responsive to my feedback and it seems he is spending quite a lot of his time on this as there is progress every week. So one thing led to another and we now have a BoF together!

In addition to these two I'm behind-the-scenes involved in the MySQL Community awards. I'm glad to see they have been given their own spot right after the opening keynotes. Previously they've been an un-announced (though always expected) part of the opening keynote. I'm glad the conference organizers are highlighting the awards in this way. It's always nice to share some praise to people who really deserve it!

Finally, I should mention the session NoSQL with MySQL presented by my colleague at Nokia, Yekesa Kosuru and Peter Zaitsev (Percona). While I don't work in this project myself, I actually watched last year's presentation so I know it is an interesting system. Yekesa's team has been combining Voldemort together with a MySQL backend. These kind of "best of both worlds" solutions always inspire me - in this case Voldemort gives you all the CAP theorem replication stuff, while your data is still safely stored in MySQL, accessible by SQL should you still want that, and you can do monitoring and backups with familiar (and more importantly, already existing!) tools. You will naturally find me in the audience of this presentation too.

Other than this, I plan to learn about Drizzle, HandlerSocket, a few 3rd party replication solutions, Xtrabackup, attend the community dinner on Monday and Drizzle developer day on Friday. (On Saturday evening, today, I'm surprised to say, I still have no fixed plans - open for suggestions.) I'm also generally interested to get a feeling for where the momentum is - with Drizzle joining the MySQL forks game with a stable release, a lot has again changed from last year.

As for Mikael Ronström's teaser on what Tomas Ulin will announce in his keynote related to new MySQL Cluster API's, I'm betting on Memcached API.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Use [fn]...[/fn] (or <fn>...</fn>) to insert automatically numbered footnotes.
  • Allowed HTML tags: <h1> <h2> <h3> <h4> <p> <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <br> <sup> <div> <blockquote> <pre> <img>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically. (Better URL filter.)

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.
2 + 4 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.