InnoDB Deadlock Count Patch

Deadlocks are a common occurrence in relational databases. They usually aren’t a problem until they start happening too frequently. Innodb can provide you with information about the latest deadlock in SHOW ENGINE INNODB STATUS. This can be useful for debugging but it’s almost impossible to get the rate at which deadlocks are occurring. This patch applies against MySQL 5.0.72sp1 and probably quite a few other versions. It adds a counter to show table status that tracks the number of deadlocks. In this example mysql-1> is connection 1 and mysql-2> is connection 2.

mysql-2> show global status like …

This is the Sharding for the masses: Introducing the SPIDER storage engine by Giuseppe Maxia, given at OpenSQLCamp, at FrOSCon, in August 2009. These are somewhat live notes, and the slides are available too.

Sharding for the masses View more documents from Giuseppe Maxia.

Why sharding? Scaling, of course. The MySQL way to solve this, is replication (even Yahoo! and Google use this).

When the master doesn’t have enough resources to cope with what you do (i.e. large data sets), replication chokes.

You can use proxies for sharding. There exists …

Disclaimer: This blog post is about things I did on my own free time, not endorsed by my employer.
A little over a month ago, Ronald posted a blog about the different query logging plug-ins that are available for Drizzle. This was pretty exciting news, especially when I saw the details that were included in the logs.

Meanwhile, a few weeks ago, I started looking at the REST API that comes with the MySQL Enterprise Monitor.

The result is that we can now see most of the information returned by the plug-in, on the Dashboard.

…

Every time I teach the MySQL Cluster architecture, someone inevitably asks "Isn't the management node (ndb_mgmd) a single point of failure?" The short answer: no. The management node is not a SPOF because the cluster can continue without it. However, it's inconvenient if your management node is down because the management node does several things such as:

• 
  
• Provide status information about the cluster and allow you to use the ndb_mgm for various maintenance tasks like taking a hot backup
  
• Own the cluster config file (therefore it must be running to start a node)
  
• Arbitration in case of a potential split-brain
  
• Logging
  

So while the management node can be down, it is nice to have a redundant one for failover. This is very easy to do:

1. 
   
2. Add 2 [NDB_MGMD] sections to …

Next Thursday (September 3, 13:00 UTC), we'll resume MySQL University sessions after the summer break with Reggie Burnett, head of Connector/NET development, giving a session on new features in Connector/NET 6.1.

For MySQL University sessions, point your browser to this page. You need a browser with a working Flash plugin. You may register for a Dimdim account, but you don't have to. (Dimdim is the conferencing system we're using for MySQL University sessions. It provides integrated voice streaming, chat, whiteboard, session recording, and more.) All MySQL University sessions are recorded, that is, slides and voice can be viewed as a Flash movie (.flv). You can find …

Next Thursday (September 3, 13:00 UTC), we'll resume MySQL University sessions after the summer break with Reggie Burnett, head of Connector/NET development, giving a session on new features in Connector/NET 6.1.

For MySQL University sessions, point your browser to this page. You need a browser with a working Flash plugin. You may register for a Dimdim account, but you don't have to. (Dimdim is the conferencing system we're using for MySQL University sessions. It provides integrated voice streaming, chat, whiteboard, session recording, and more.) All MySQL University sessions are recorded, that is, slides and voice can be viewed as a Flash movie (.flv). You can find …

In record time, less than a week after the conference (thanks to the free Pinnacle Video Spin and YouTube), all 11 videos that were taken at OpenSQLCamp Europe are online.

For those who missed the sessions, or just want to relive the fun!

Almost all the sessions were filmed; regrettably Darren Cassar’s Securich – MySQL user administration and security made easy! and Stephane Combaudon’s Minimizing data access with covering indexes were not.

The YouTube videos have the descriptions and resources from the official conference pages, and links to pages. If there is more information to add (for example, the slides from a talk are now online), or if …

Comparison of block touches in support projecting rows and columns:

These examples abstract out the cost to access a given set of rows to highlight the relative cost to return (project) rows and columns.  The assumption here is that the returned rows are generally contiguous. 

There are a number of operations whose cost will be different between a traditional row based dbms and a column based dbms.  For this discussion, an abstract table with 20 columns (4 1-byte columns, 8 4-byte columns, 8 8-byte columns) will be used with 10 million rows.  The row length is 100 bytes of data based on the data types listed. 

Additional assumptions/approximations include; data is stored in blocks/pages of 8k bytes, that any overhead for row identifiers is negligible, the primary measure of cost is block touches, distribution of rows within blocks is …

Recently I had a chance to take a look at Redis project, which is semi-persistent in memory database with idea somethat similar to memcache but richer feature set.

Redis has simple single process event driven design, which means it does not have to deal with any locks which is performance killer for a lot of applications. This however limits it scalability to single core. Still with 100K+ operations a second this single core performance will be good enough for many applications. Also nothing stops you from running many Redis instance on single server to get advantage of multiple cores.

I call Redis semi-persistent because it does not store the data on disk immediately but rather dumps its all database every so often - you have a choice of configuring time and number of updates between database dumps. Because dump is basically serial write Redis does not an expensive IO …

Code scanning and management vendor Black Duck reports the GNU General Public License v2 (GPLv2) now dipping below 50% share of open source software. While we already knew that GPLv2 was somewhat in decline from its far greater share of open source code over the last 5-10 years, it is useful to know what pool of code we’re talking about. We must also remember that while GPLv2 may not be as dominant as it once was and that other licenses, particularly GPLv3, are quickly gaining share, GPLv2 is still quite relevant to enterprise open source software, is used in a variety of newer and popular applications across the enterprise stack and is likely to remain in the top 10 licenses for a long time.

Regarding GPLv2 and Black Duck’s findings, some folks are rightly asking what code and how much of it are we considering where GPLv2 accounts for half or less of the …