Since the recent demise of my Thinkpad, and the musical hard drive game leading the Thinkpad's hard drive being installed in the Toshiba, the Toshiba has been operating with an insufficiently sized swap partition. This has led to severe performance problems, particularly when doing MySQL builds. So I enlarged the swap partition, deleting the operating system partition and in fact replacing it with two equally sized partitions, similar to the partition layout on the Toshiba's original hard drive.

So of course this required me to reinstall Slackware, MySQL, and PHP. I have found the Apache bundled with Slackware 10.2 to be ample for my needs so far. So first I ran my pulls script to pull and build all the different MySQL source trees, and installed the six instances I normally operate. So far so good. Then I turned my attention to installing PHP.

The first time I installed PHP (5.1.4) I must have had an extraordinary …

MySQL told me that it had a problem renaming a file. So I really thought I would go and help it by removing the file on the command line. But somehow I really felt that it didn’t really mean to rename a file. So I did a bit of googling and in the mysql bug tracker I found the problem.
The real problem occurs when you are using foreign key constraints with InnoDB. Even though I had only created an additional unique index on one of two columns where one of them was a foreign key. And when I tried to remove it mysql told me it had problems renaming a file :-(.
The solution was to remove the foreign key using ALTER TABLE tablename DROP FOREIGN KEY fkname. Now I was able to remove my index and then I added the foreign key constraint again.
Really ugly.

Stumbled into this today over at squarebits, using MySQL to generate bar graphs:

mysql> select Name, REPEAT("#", Value/20.0+1) Graph FROM Month;
+-----------+------------------------------------------------+
| Name      | Graph                                          |
+-----------+------------------------------------------------+
| January   | ####################                           |
| February  | ############################################## |
| March     | …

Spoke with Brian the other day on what was required to get NDB to be a pluggable engine - and started hacking.

The tricky bits invole dependencies of things like mysqldump and ndb_restore on some headers to determine what tables shouldn’t be dumped (hint: the cluster database used for replication).

Also, all those command line parameters and global variables - they’re fun too. It turns out InnoDB and PBXT are also waiting on this. In the meantime, I’ve done a hack that puts config options in a table.

Currently blocked on getting the embedded server (libmysqld) to build properly - but i now have a sql/mysqld binary with pluggable NDB. All libtool foo too.

Hopefully i’ll be able to post soon with a “it works” post

Inspired by PeterZ’s Opening Tables scalability post, I decided to try a little benchmark. This benchmark involved the following:

• Create 50,000 tables

• CREATE TABLE t{$i} (i int primary key)

I wanted to test file system impact on this benchmark. So, I created a new LVM volume, 10GB in size. I extracted a ‘make bin-dist’ of a recent MySQL 5.1 tree, did a “mysql-test-run.pl –start-and-exit” and ran my script, timing real time with time.

For a default ext3 file system creating MyISAM tables, the test took 15min 8sec.

For a default xfs file sytem creating MyISAM tables, the test took 7min 20sec.

For an XFS file system with a 100MB Version 2 log creating MyISAM tables, the test took 7min …

Have been meaning to post about the cluster setup I'm using for *functional* testing (not appropriate for performance testing or production environments).

The gist is that I'm using Parallels on OS X, running 4 nodes (1 management, 2 & 3 storage, 4 SQL) on 4 virtual machines. Each machine has between 128Mb and 256Mb of allocated RAM which isn't much for a cluster, but is as much as I can give from the 2Gb on the MacBook Pro. Each virtual machine has a version of MySQL 5.1.12 compiled from source. Actually, I built one VM with the compiled and installed source and then cloned that for the other 3 machines. The preferred flavor of Linux for these is Ubuntu server.

As I was setting up for a screenshot I walked through bringing …

I'll be at LCA 2007 in Sydney in January... Stewart and Colin are coming also, and quite possibly Brian. We're setting up a MySQL miniconf this time, and of course we would like to see some items from people who are not MySQL employees! See here:

This is a call for participation in the MySQL Miniconf which will be part of the linux.conf.au (LCA) 2007 conference in Sydney, Australia. The Miniconf will be run on Monday, January 15th, before the conference proper starts. This is the first conf of its type to be held in the Southern hemisphere and we are looking forward to participants from areas around Australia, New Zealand, South-East Asia, India and South America.

Any topic based around MySQL is allowed. Types of presentations that we're after, include:

MySQL MiniConf at linux.conf.au 2007
This implies I’m coming to linux.conf.au 2007 in Sydney next January. What’s more is that during the MiniConfs, we’ve got one for MySQL. Its on January 15th, and we’ve just put out the call for participation/papers. You have about eight (8) days left to submit a paper. So submit your tales of deployment, conference presentations, and I believe we’re even willing to accept “hand’s on” hacking sessions (ala what happened at MySQL Camp). Keep the wiki page handy, and submit goodies to mysql-miniconf[AT]mysql[dot]com.

Interview with Paddy Sreenivasan
Yes, Engineering Lead at Zmanda, they’re big on AMANDA and now …

Just received this note from from Yann Larrivée:

PHP Québec is pleased to announce the 2007 PHP Québec conference. The conference will take place in Montreal, Canada on March 14-15-16th 2007. We are looking for the best speakers willing to share their experience and skills with professional PHP developers from eastern Canada and the United States.

This year, the conference will feature 3 distinct tracks:

• Advanced Techniques: Providing in-depth details of PHP techniques
  
• Data Availability: Databases, XML, Web Services, VOIP, TOIP, WAP, etc
  
• PHP Beyond theory: Real solutions for real problems related to software development and project management

For more information, please visit the conference website..

The next version of CouchDbX comes with nice buttons and a cute interface. Enjoy the tease!