As I run into the same mistake over and over again I am going to throw them out here in order to serve as a warning and a learning tool to others. Some of these may seem like I am beating a dead horse because you can find examples of these on other sites, in other blogs, or even in the manual. Despite the large amounts of data on the subjects these things still pop up. This is not a complete list, and this list is not in any order… so please don’t dispare if I miss your favorite common performance gaff. As I encounter more I will post them to the common mistakes category here on bigdbahead.com. Hopefully this list will grow to include a large subset of issues and maybe one or two people will take note and fix their applications and databases.

#1.) MORE IS NOT ALWAYS BETTER

We often see configuration examples where folks believe in the old adage more is better. Case in point if a 2MB sort buffer …

Well, in the “notes from the front line” part of this post…. I have been a MySQL DBA at The Pythian Group for three months (and 2 days) now. At most companies that is the probationary period, and I am still here, so that is a good sign….. So, after three months, how do I like [...]

We did a community townhall video session earlier this week with Jonathan Schwartz, Rich Green and myself.  You can view it online at http://ustream.tv/sun and at YouTube.  This is a pretty high-level discussion, but we responded to some questions that came in from the live chat about platforms, languages, patents, how Sun makes money from MySQL etc.  I was also hoping to resolve once and for all what remains one of the hot developer issues: vi vs emacs.  And in keeping with the MySQL tradition, we had a vodka toast at the end.

There were also more technical IRC sessions led by Kaj Arno and …

I'm talking at the MySQL User Conference once again this year presenting my Code Generators for MySQL Plugins and User Defined Functions in a formal session this time after my original "pizza & hacking" session on this topic on the 2006 UC (which ended up in hacking some other things on the way first ... photographs may follow ...).

The plugin and UDF code generators are based on the same foundation as my "pecl-gen" generator for PHP extensions so this might be of interest for hardcore PHP folks, too.

For more information see the Conference Schedule.

As Kaj announced, there was a IRC session with Ian Murdock on March 5.
Since I have been working in a virtual company for 18 months, I should not be surprised about this kind of events. Actually, IRC has become part of my life. I could not conceive working efficiently without it. When some new colleague asks me to have a simple phone call, without any IRC or other char facilities, I start shaking at the idea of such poor and antiquate way of communicating.
Even for one with experience, though, a public IRC meeting in a crowded #room is quite an experience. The "location" for the meeting is room #mysql on irc.freenode.net, a room that normally contains 400~500 people, where the level of noise is quite high.

Having a public …

As Kaj announced a few days ago, there is a IRC session with Simon Phipps scheduled for March 6, at 14:00 UTC, 15:00 CET.
Don’t miss it. It’s a great occasion to ask questions, and to chat to a bunch of gurus in the same room, all at once in true URC tradition.
The IRC meeting with Bob Brewin has been rescheduled to Wednesday, March 12, at 17:00 UTC, 18:00 CET.

I didn’t get to attend many sessions at the Tech Days, as I was mainly meeting with people, or at the PostgreSQL on Solaris booth (figures someone from MySQL should’ve been there too). I had plenty of interesting conversations with Tom Daly; we met by chance since I had a blue MySQL shirt and he figured he’d be cheeky and offer me a PostgreSQL one.

I was going to take a photo with him today, but he called me from the airport last night to say he had to go back. Oh well.

Laurie Wong, and me

Instead, you get a photo of Laurie and me (MySQL cap, PostgreSQL on Solaris t-shirt). I’ll blog later about the sessions that I did attend - I wish I could’ve done so earlier but staying at the Sheraton in Darling Harbour, meant that there was no in-room Internet access. …

Short version (aka, Executive Summary)

If you try to run a 32bit program under 64bit Ubuntu, there is a chance you might see this misleading error message:

-bash: ./executable_filename: No such file or directory

What the shell is actually trying to say is "You can't run 32 bit applications here without installing support for 32 applications!".

Longer version

We purchased the Webyog's MONyog MySQL monitoring tool, after some trial version testing in a Windows environment. When trying to run it on one of our Ubuntu servers, I got this:

myuser@myserver:~/MONyog/bin$ ./MONyog-bin
-bash: ./MONyog-bin: No such file or directory

I checked again: the file was there, permissions were okay. Hmm. Why was bash saying the file wasn't there? Here it is! It's not like the error came from the …

Another quiz based on comments from aka fenixshadow:

Two identical tables with FOREIGN KEY constraints within the tables are created:

CREATE TABLE t (
  id INT NOT NULL PRIMARY KEY,
  type INT,
  pid INT,
  FOREIGN KEY (pid) REFERENCES t (id)
) ENGINE=INNODB;

CREATE TABLE t1 (
  id INT NOT NULL PRIMARY KEY,
  type INT,
  pid INT,
  FOREIGN KEY (pid) REFERENCES t1 (id)
) ENGINE=INNODB;

The tables are created, and then the following updates are done in table t:

INSERT INTO t VALUES (1,1, NULL);
INSERT INTO t VALUES (2,1, NULL);
UPDATE t SET pid  = 2 WHERE id = 1;

Satisfied that all the updates have completed successfully, we attempt to add the data in t to t1:

INSERT INTO t1 SELECT * from t;

But are greeted with:

ERROR 1452 : Cannot add or update a child row: a foreign key
constraint fails

Question 1: Why did …

A ''tipping point'' is a concept, product or idea that becomes a hot commodity that attracts everyone's attention, interest or inspiration. There are always specific reasons and factors that are not easily identifiable why one product becomes a tipping point and others do not. Products not considered a tipping point usually never become a tipping point and ofter never understand why they didn't