Cannot give you all details yet and it's too early to say what the max performance we'll finally obtain, but currently we out-passed 100.000.000 (!) Read+Write transactions/sec on the latest MySQL 5.5 simply integrated on the Exadata X2-8 (full rack) instead of Oracle database!!! :-))

The result is so incredible.. - but we spent several days now to validate that workload is perfectly matching customer's requirement and really reproducing their production environment. The problem came from the performance issue they observed month to month and came in our Benchmark Center to test a new HW platform (and we're in competition here with HP and IBM, but they have already finished their testing and without a positive result (no comments.. ;-))).. Finally, no one of any tested platforms was able to keep the load expected for the end of the year.. And of course we think about Exadata :-)) but the problem that the customer's application is …

Longtime MySQL-er Lenz Grimmer

Lenz Grimmer

is leaving the MySQL side of the business to work in the Linux and Virtualization side of the Oracle business. He came from SuSE Linux and has retained interest in that area all the many years he helped make MySQL into the ubiquitous tool that it is. Lenz’s list of accomplishments is too long to note here and working with him on the MySQL Community Team has been a tremendous pleasure. He is not dropping off the face of the earth and will presenting at Collaborate 11 .

Please join me in wishing all the best for Lenz!

…

Building a Feed can be broken down into a few key questions.

Who can see what?
How many people can see it?

These key answers really dictates the design of data structure from a global read method to a global write method. A global read method can be summarized using some sql.

SELECT * FROM Activity WHERE userId IN (?,?,?,?,?) AND createDate > NOW() - INTERVAL 2 WEEKS;

Above is actually not an optimal SELECT, what is really done in my case is foreach '?' do a parallel query to each shard group of

SELECT * FROM Activity WHERE userId = ? AND createDate > NOW() - INTERVAL 2 WEEKS;

The reason why the 1st query is not optimal is due to the fact that their are two ranges in the 1st query. The IN clause is a range and createDate is a range thus you can't use a composite key (userId, createDate) the query is only using userId.


A …

This week we conclude the interview with Tokutek's co-founder and Chief Scientist, Martin Farach-Colton. Part 1 of the interview. TokuDB is a transactional, fully ACID-compliant storage engine that uses fractal trees for data and indexes, instead of MySQL's standard B-tree implementation. Combined with TokuDB's message-based architecture, TokuDB data and indexes do not fragment, have completely online column and index addition and removal and unlike InnoDB and XtraDB, do not fall apart when indexes no longer fit into memory.

read more

Welcome to this week's Last Week in Drizzle.  Today will be a relatively short edition due to the work everyone is doing preparing for the 2011 O'Reilly MySQL Conference and Expo and Google Summer of Code.
First Fremont Tarball
The first tarball of the Fremont development branch of Drizzle was created this week, following our tradition of releasing a tarball every two weeks.  It includes many experimental things such as the libdrizzle-2.0 separation and the multiple master to single slave replication.

For those wanting the stable release we suggest sticking to the Elliott branch which our GA was cut from.  New releases for this will be created much less frequently and will only include bug fixes.
Xtrabackup
Stewart …

Welcome to this week’s Last Week in Drizzle.  Today will be a relatively short edition due to the work everyone is doing preparing for the 2011 O’Reilly MySQL Conference and Expo and Google Summer of Code.

First Fremont Tarball

The first tarball of the Fremont development branch of Drizzle was created this week, following our tradition of releasing a tarball every two weeks.  It includes many experimental things such as the libdrizzle-2.0 separation and the multiple master to single slave replication.

For those wanting the stable release we suggest sticking to the Elliott branch which our GA was cut from.  New releases for this will be created much less frequently and will only include bug fixes.

Xtrabackup

…

Working at Percona has taught me to love being disruptive and creating innovative pricing models. In particular, their pay-as-you-go pricing in 15-minute increments, plus the ability to go from first contact to having them logged in and working in a few minutes with just a credit card, has been game-changing. I’ve always wanted to take that to the next level and create a new business, offering MySQL services with a radically simple pricing model.

Well, now I have. I’m offering the option to pay by the minute for MySQL consulting and support. And you don’t even need a credit card or approval from your boss. Best of all, I’ll talk geeky to you while I do it. Just call 1-900-SELECT and in no time at all, I’ll be fixing your servers while saying things like “I love the way you bash me with those shell scripts, you naughty boy! I’m gonna make you ‘friend class Baron’ so you’ll have access to my protected variables!”

There is 1 reserved connection for anyone having he SUPER privilege. And your replication account shouldn't need the SUPER privilege.
http://dev.mysql.com/doc/refman/5.5/en/too-many-connections.html

In Bug #24826 there is a patch to increase the number of reserved connections. This was done to make it possible to let the slave user use reserved connections.

You really should use the mysql_secure_installation utility when setting up a new server. This remove the privilege for root to login remotely.

With this revision, My xtrabackup branch has been merged into trunk.

What does this mean? It means that we now get a drizzlebackup.innobase binary which is the xtrabackup port for Drizzle. Exciting times.

Today we discovered a particularly subtle way of fucking up a server restart. After a routine configuration change and RPM upgrade, a colleague tried to restart an important master. That failed. The message: CODE: root@master ~]# /etc/init.d/mysql start
Starting MySQLCouldn't find MySQL manager (//bin/mysqlmanag[FAILED]erver (//bin/mysqld_safe)
The colleague tried multiple times, and finally resorted to manually typing a CODE: nohup mysqld_safe ... into a screen console, which he detached.

That took care of production for now and left us with an investigation. Why is the init script trying to start the MySQL manager?

It is not, and never tried to. What happen?


Continue reading "Fighting the mysqld init script" …