MySQL Performance Blog

Have you ever had a question about Innodb internal design or behavior which is not well covered in MySQL manual ? I surely had.
Now you have a great chance to have them answered !

Heikki Tuuri, Creator of Innodb will answer your Questions about Innodb at MySQL Performance Blog.

Please leave your questions as comments to this post by 5th of October and I will pass them to Heikki to reply merging with questions I have myself.

Note: due to Oracle policies Heikki will likely be unable to answer your questions about Innodb new features or timetables.

Many Open Source software solutions use database per user (or set of tables per user) which starts to cause problems if it is used on massive scale (blog hosting, forum hosting etc), resulting of hundreds of thousands if not millions of tables per server which can become really inefficient.

It is especially inefficient with Innodb tables both in terms of space (some tables would keep only couple of small rows, but require at least 16K page in Innodb), keeping all tables open in Innodb dictionary and number of other challenges in IO management and recovery. For MyISAM it works better but still overhead can get significant because table_cache can’t be made large enough and so a lot of table reopens needs to happen which requires table header modification, which is costly.

Of course if you can simply rewrite software to store multiple users per table it is best way to go, however quite typically this is way too much work and also requires constant patches as new software versions come out.
[read more...]

It always surprised me how little Innodb team seems to think about product usability/ease of use, when it comes to settings, performance management etc.

I could understand many things 5 years ago, like a lot of information being available only in hard to parse SHOW INNODB STATUS output or even uglier hacks with creating tables such as innodb_lock_monitor to get more detailed information free space specified in table comments (which need to be parsed) etc. 5 years ago Heikki was along and he had a lot to do to make things work well so a lot of these things were just done quick and dirty way.

It is however hard for me to understand why so many years later with significantly increased team not only many of these things remain unfixed but things are still done similar way ?
[read more...]