From Webworkshops...
Scaling MySQL
But that doesn't mean I wear a skirt.
Tonight I took a trip to the Meetup Organizer Meeting. The group
is for Meetup Organizers to sit and talk about what they are
doing to promote their groups.
Since the PR VP from Meetup was attending, and offering free
pizza, I thought I would attend.
I went not expecting much. At best I was hoping to run into other
people in Seattle who run technology Meetups, and at worst I
thought I would be sitting around a table with five other group
organizers listening to someone complain about how no one ever
attended their meetings. I had already contacted someone at
Meetup.com to see if there would be anyone there who worked on
the site, and had found out that this would be non-geek sort of
event. (The site is powered by the MySQL database server, and I
always enjoy hearing how people make use of MySQL).
What I did not expect to find was a room filled with 50 people
who were passionate about their groups. The crowd was a …
I get stuck on the dumbest shit ever.
I have master-master MySQL 5.0 servers. I have a nightly snapshot
of the "slave" database, and I need to finish hacking in an
automagic method of storing binlogs along with the snapshot. I,
for the life of me, cannot decide how to do this. I've been
inching for hours.
Do I want to start up 10 tiny mysql instances with our entire
schema mirrored using the "blackhole" engine? How would I figure
out what position in the binlog correlates with the snapshot?
(not too hard I guess). How the hell would I keep that all in
sync as developers randomly add/remove tables?
From what *angle* do I grab the damn files? How does my normal
binlog rotator manage to communicate with the binlog slurper what
files it's already slurped? YAML? A special database on each
database? A file in the snapshot directory? Why do I care? :P Why
have I written so many of these scripts for so many …
Brian Aker starts work on a memcache engine for mysql. so your memcache cache acts just like a table.
the big thing here which I’ve seen asked for a couple of times on the memcached list is the ability to see a list of keys.
mysql > select * from foo1 WHERE k=”mine”;
freaking amazing.. I love these kind of mashups.
and the 2nd important event.
Django is starting a branch to integrate SQLAlchemy
There's a good blog by a former tech analyst, who also happens to be my twin brother, called OnDisruption. While I don't normally like to promote something as blatantly nepotistic (is that even a word?) as this, his blog doesn't cost anything to read so I don't think there's a conflict of interest here.
His most recent posting is on the "Eight Traps of Disruption" and does a good job analyzing some of the common mistakes businesses make when they think they are being disruptive. For those who lived through the dot com era in Silicon Valley, you may recognize a few of the examples. For others, yes, it sounds crazy, but people really did make these mistakes. And often to the tune …
[Read more]
“The key, the whole key, and nothing but the key, so help me
Codd!”
The database guys should like that one. Can’t remember where I saw it today to acknowledge reference.
Today MySQL published the press release with results of Database Contest (results on German available here http://www.mysql.de/ct-dbcontest).
Peter and me spent quite some time working on this project while being employed by MySQL and it is great to see results finally publicly available.
The story began about year ago when C'T magazine had called for Database competiton using Dell DVD Store benchmark (details available here: http://firebird.sourceforge.net/connect/ct-dbContest.html).
Most interesting results are (more orders per minute are better
):
MySQL5/PHP (our solution) : 3664 orders per minute
DB2/Java : 1537 opm
Oracle / Java: 1412 opm
…
Peter Zaitsev over at the excellent MySQL Performance Blog recently wrote an article on duplicated and redundant indexes -- any indexes which cover exactly the same columns as another index, or cover a leftmost prefix of another index. While there are subtleties, such as FULLTEXT indexes not being the same as non-FULLTEXT, for the most part this is sufficient criteria to raise possible duplicates to a DBA's attention. I opened my big mouth in the comments and said I could write a quick Perl script to discover possible offenders in just a few lines of code. Once I did that, I had to do it and give you the script. Here it is.
I'm giving a webinar on tag schema designs and folksonomy concepts tomorrow, from 1pm to 2pm EST. The webinar is chock full of SQL code and optimization tips, so it should be a fun time ...
We'll go over tagging database access patterns, many-to-many mapping techniques, using MySQL's storage engines and replication effectively to scale out the schema, and discuss the advantages of a normalized schema approach versus other popular approaches. Oh, and we'll make fun of rounded corners, too.
Beijing, August 29 2006 - MySQL AB, the developer of the world's most popular open source database, and Greatlinux Inc, a leading open source solutions and services provider in China, today jointly announced a strategic partnership to deliver commercial MySQL-based solutions to the burgeoning Chinese market.