I’ll be checking out MySQL’s “DBA Boot Camp” for Oracle DBAs on Thursday. Back next week to report on it.

Then, upcoming posts will discuss further topics from my talk, touching on scaling strategies, backups, and permissions. After that we’ll get to some more technically detailed issues that we encountered when we switched over to MySQL and got things up and running. Keep watching!…

Fedora receives $4.9m grant. MuleSource adopts OSI-approved CPAL license. FiveRuns and BitRock launch Ruby on Rails stack for the enterprise. (and more)

Gordon and Betty Moore Foundation Awards Fedora Commons $4.9M to start non-profit organization and develop open-source software for collaborative communities, Fedora Commons (Press Release)

MuleSource Announces Adoption of CPAL, MuleSource (Press Release)

FiveRuns Releases RM-Install, Powered by BitRock, FiveRuns / BitRock (Press Release)

Oracle CEO Ellison: Open source software is no threat, Haaretz, Guy Grimi …

I?m having a lot of difficulty understanding the debate surrounding MySQL?s decision to make their enterprise binaries and source less available than before. I was hoping that a few days into it I would get a better sense of the debate and why I was seeing the positions that are popping up. I realized, though, that the context of the discussion is deeply entrenched within the traditional software business model. Furthermore, it is complicated by MySQL?s IPO plans. I?ve seen people argue that MySQL is not as open, that MySQL is technically adhering to the GPL, that this is all... READ MORE

Catching up on some articles on the Planet MySQL feed, I just read the post by Dathan on how to promote a slave to be master by using MASTER_POS_WAIT(). The MASTER_POS_WAIT() is an excellent function that allows you to wait until the slave reaches a point at or after the given binlog position. Observe that after the statement issuing a MASTER_POS_WAIT() returns, the slave threads are still running, so this means that even if a STOP SLAVE is issued immediately after the statement with MASTER_POS_WAIT(), it is bound to move a little more before actually stopping. For Dathan's situation, this is not necessary, but wouldn't it be great if you could stop a slave at exactly the position that you want? Well, that …

What if last week’s announcement by MySQL had been the following?

Some major MySQL distributors have brought some issues to our attention. These distributors have been legally distributing the Enterprise version of MySQL to folks that do not have a MySQL support contract, without their knowledge or consent.

When a user installs an open source operating system and includes MySQL, should that user get the Community version or the Enterprise version? Since the source code was split back in October 2006, MySQL AB has felt that users should use the Enterprise version only if they intend to. We have made the source easily downloadable, so folks that want to use Enterprise, are able to use Enterprise.

However, given that distributors have been doling out Enterprise to unsuspecting users, we decided to change a few things. Firstly, we recognize that folks may be downloading Enterprise because they do not understand that Community …

This is fun. Probably submitting a bug report if I can get to the bottom of it.

Stupid pet trick with mysql 4.0:

mysqldump [equivalent of --opt without the --add-locks] --single-transaction --master-data | whatever

According to the manual, you'd expect this to work like it does on mysql 4.1+. A tiny lock to grab the master data info, then a consistent read is done via a transaction.

What actually happens? Full table lock, AND a transaction! Delicious.

Looking inside the 4.0 source code for mysqldump.c, it's obvious what the bug is. There's a check to exclude --single-transaction and --lock-tables, but --master-data enables --first-slave (later renamed to --lock-all-tables) silently, which does a lock on all tables.

Fun! So where's the bug, right? 4.0 is deprecated. Nah, the bug comes more recently.

See MySQL 5.0.24a. Attempt to use its …

Judges rules for Novell in SCO case. XenSource updates XenEnterprise. Red Hat releases Eclipse-based IDE for JBoss. (and more)

XenSource Raises the Bar on Enterprise-Class Virtualization with Industry?s Most Open, Scalable, High Performance Solution, XenSource (Press Release)

Red Hat Accelerates Application Development and Migration with Integrated Developer Tools, Red Hat (Press Release)

Actuate Harnesses Open Source to Advance Leadership in BI, Actuate (Press Release)

Fonality Launches Free Business Phone System with Free Worldwide …

It’s very nice that the LOAD XML patch finally got some attention. I wasn’t expecting it anymore. I worked on this with Erik a year ago. My C programming skill were to rusty at that time, that only my BNF code contribution was used and a lot of issues were never solved. With my C prog skills back up to a decent level, I;m planning to pick up the project. Now seems to be a better time than ever.

The current solution works well in certain situations, but definitely not in all. To insert referential data, you need to specify the id value at forehand, look at project_id in the example. That …

Dear Lazyweb,

I have all my old .netscape and .mozilla directories for all my past installations of Netscape, Mozilla, and Firebox, going back to 1996. I've got cert.db, cert5.db, and cert7.db files.

I want to extract out all of the certs and keys that I had added to those databases, and then import them into my current ~/.mozilla database.

Reading the docs and futzing around with NSS signtool, certutil, and pk12util isn't getting me very far.

Surely there is a tool or technique do doing this...

As we started to make the switch to MySQL, several of our managers and developers (and I) assumed we’d use clustering to improve availability and performance. (Performance, because they thought we’d be able to spread our load among clustered servers.)

I read up on storage engines and got to explain more than once that we couldn’t use clustering, much to their surprise and initial disbelief.

Clustering relies on a particular database engine (NDB) that is currently an in-memory-only engine. (Ie, all your data has to fit in memory.) Since our database is much bigger than our RAM, we couldn’t use it. (Future versions promise to allow data stored on disk.) We ended up using the InnoDB storage engine, and replication for availability and performance.

So what is a storage engine? If you don’t know, you need to find out before moving to MySQL.

MySQL implements a sort of data-handling architecture, into which …