Well, this week has been pretty hectic. We are really getting to the “move ALL the machines!” stage of the data center move, and the train we have going on April 9th includes many, many virtual machines. I am also preparing for the Percona Live: MySQL Conference and Expo in just over a week. Probably the most exciting news is that the Mozilla booth in the Dot Org Pavilion will have a Boot-to-Gecko device available. If you want to see Mozilla’s foray into the mobile device market, come see us at the conference!
This weekend I’ll be burning the midnight oil working on presentations, for the Percona Live: MySQL Conference and Expo (tutorial and lightning talk), SkySQL & MariaDB Solutions Day and for a presentation at Effective MySQL in New York City next week. But, that’s in the future, let’s delve into the past week, shall we?
- Ordered 10 copies of the MySQL Administrator’s Bible for the book signing on Wednesday, April 11th. I will be selling the book for US $40, since it’s easier to not have to make change, and there is an ATM on site. If you have your copy, I’m happy to sign it.
- Shipped a box of CD’s with sample OurSQL episodes. This is also for the book signing (I guess it’s a “media signing”?) on Wed night. But unlike the books, the CD’s are free, so make sure to stop by and get one! 25 episodes with links to the show notes website pages.
- Gathering folks to staff Mozilla’s Dot Org Pavilion booth, and swag to give away.
In non-logistical, actual database news:
- Released a new major version of Socorro, our crash-stats application, including database changes
- Moved a virtual machine for a PHPmyadmin server.
- Restored the backup to a development machine that had a bizarre turn of events reset the data back to December 2011.
- Created databases and users for the development and stage versions of wiki.mozilla.org and treestatus, a simple wsgi app to maintain the open/closed state of development trees.
- Dealt with our Addons primary db master crashing (again). This was after it had crashed 2 weeks ago and we were advised to update the firmware, which we had done. Needless to say, our secondary db is now our primary db, and we now have n+1 slaves (we normally operate with n+2 slaves). The crashed primary master is up and running, but we are not putting it back into production use until we get a definitive answer as to what’s wrong with it.
- Upgraded the staging master database for Bugzilla to Percona’s patched version of MySQL 5.1. This weekend I will be upgrading a production slave….I’m so excited! Note that the database is also being converted to innodb_file_per_table while we are at it.
- Updated permissions for our metrics team to be able to access a new custom Bugzilla field
- Debugged an application having “Aborted connections” on the MySQL side and CLOSE_WAIT sockets on the client side. (The application wasn’t closing the database handler properly)
- Worked on a migration plan for the Mozillians accounts, so that we could try to vouch them. Are you a Mozillian yet? Add yourself as a Mozillian!