Alex Stetsenko

While all other TV channels were showing mystery or horror movies (no surprise – it’s Halloween night), TBS decided to broadcast an “old” good movie Music and Lyrics.

By the way the soundtrack album reached #5 on the Billboard Top Soundtracks Chart.

The 2010 edition of the MySQL Conference & Expo is already announced to take place in Santa Clara, CA on April 12-15.

Although schedule has not been published yet, obviously it is the most important event for MySQL professionals. The conference gathers over 2,000 open source and database enthusiasts from around the world.

At the same time Percona’s Director of Consulting has announced dates of Percona Performance Conference 2010. Of course it is April 12-13 🙂 . Neither place nor schedule have been unveiled yet. But last year it was the same building as MySQL Conference & Expo. In spite of it is not 100% MySQL related event, definitely some people will be pulled away from MySQL conference.

However Percona guys take their conferences very seriously. They have launched PerconaTV where video and slides of PPC 2009 are available.

By the way a lot of presentations and some videos are made available from MySQL Conference & Expo 2009 too.

These two events at one week could be a good reason to visit California. Looking forward to seeing additional details on both conferences.

Her other hits “Just Dance” and “Poker Face” have become international number one hits. Check Lady Gaga’s schedule if you want to see her live.

Developing a new section at this blog I would like to post information about one more database related event.

OpenSQL Camp 2009 will be an un-conference style get together for people interested in Open Source Databases. They are not focusing on any specific project, and hope to see representatives from a variety of open source database projects attend. It is taking place in Portland, OR on November 13-15.

Currently announced sessions are:

1 Virtual Pivot Tables in SQLite – Brian Duggan
2 BLOB handling and storage using MySQL/Drizzle and the PBMS engine – Barry Leslie
3 Checking MySQL’s replication sync with Ma’atkit tools – Sheeri K. Cabral
4 The wiki database – Domas Mituzas
5 The architecture & roadmap for mod_ndb – J.D. Duncan
6 Implementation of the new SQL parser in mod_ndb – Tulay Muezzinoglu
7 Developing a Universal Storage Engine API – Bradley C. Kuszmaul
8 xtrabackup – tricks beyond backup , Vadim Tkachenko
9 Sphinx – tricks beyond fulltext , Vadim Tkachenko
10 Performance is about Response time , Vadim Tkachenko
11 Redis – The missing piece of the puzzle – Morgan Tocker
12 What MySQL is bad at – Morgan Tocker
13 memcached and the database – Patrick Galbraith and Matt Ingenthron
14 Comparing non-relational databases: CouchDB, Tokyo Tyrant, MongoDB
15 A completely versioned schema – Ben Hengst
16 An Introduction to MongoDB – Mike Dirolf
17 Maatkit’s Swiss Army Knife: mk-query-digest – Baron Schwartz
18 Solving problems by writing Drizzle Plugins – Monty Taylor

The good news that it is free (unlike NoSQL East Conference). But the bad news are that space is limited to 120 attendees and only 7 places are still available at the moment.

Once again I hope to see some video or audio materials from this event.

Update 10/26: 10 more spaces are available now, i.e. total will be 130 instead of 120.

It seems that it is becoming a good tradition to post commercial soundtracks. I am not sure about Lincoln fans but not a few people are still visiting this blog not to read about MySQL but listen State Farm’s “Sixteen” song by Modern Music. Now it is Droid’s turn and more to come actually… So I am guessing marketing departments do not waste theirs budgets while looking for advertisement songs 🙂 .

It is not that I don’t like iPhone but Verizon Wireless has started a very aggressive campaign against iPhone while promoting a new Android powered phone Motorola Droid. Verizon seems to be the first who will unveil Android 2.0 although Donut (i.e. 1.6) has just been released for T-Mobile Customers.

It is always interesting to see how other universities teach their students.
Thanks to Massachusetts Institute of Technology (MIT) open courseware – lecture #1 of ‘Introduction to Computer Science and Programming’ (Fall 2008) is available online. This lecture (53.5 minutes) covers goals of the course, what is computation, introduction to data types, operators, and variables.

The NoSQL movement has been gaining some steam lately, with discussion forums and blogs popping up all around the web. Now NoSQL East conference is planned in Atlanta, GA on October 28-30. The focus is promised to be on real world use cases and high level overview of architecture.

Currently they have announced the following speakers and topics:

Arin Sarkissian (Digg): Cassandra
Kevin Smith (Hypothetical Labs): Redis
Kevin Weil (Twitter): Pig
Chris Curtin (Silverpop): Cascading
John Hornbeck (Engine Yard): MongoDB
Mike Miller (Cloudant): CouchDB
Cliff Moon (Microsoft / Powerset): Dynomite
Justin Sheehy (Basho): Riak
Mark Gunnels (Catamorphic Labs): HBase
Tim Anglade (af83): tin
Emil Eifrem (Neo Technology): Neo4j
Geir Magnusson (Gilt Groupe): Project Voldemort
Yuan Yu (Microsoft Research): Dryad/DryadLINQ
John Corwin (Yahoo!): Sherpa

The conference will take place in Georgia Tech Research Institute – Conference Center. Although this event has some good sponsors it is not that cheap as it could be in the open source world. So it should be possible to sign up with a promo code Stammy250 and save $50.

Unfortunately I am not sure if I am able to attend this event. Hopefully there will be some video or slide casts available later. For example Bryan Fink shared his presentation on the riak web-shaped data storage engine which was a part of New York NoSQL mini-conference on October 5th, 2009.

P.S.: I am not affiliated with anyone of the conference organizers. However I believe it is always good to see what’s going on in the world of databases as well as to learn RDBMS disadvantages and ways how to avoid or fix them.

In addition to SHOW TABLE STATUS there are at least two more ways to find out a table storage engine.

The first way is simply to query the INFORMATION_SCHEMA.TABLES.

SELECT `table_name`, `engine`
FROM INFORMATION_SCHEMA.TABLES
WHERE table_schema = DATABASE();

In case there is not permission to access INFORMATION_SCHEMA the second way requires only the SELECT privilege for the table.

SHOW CREATE TABLE TestTable;

A while ago I was involved in the technical part of an interview with a PHP developer. Obviously some questions were about MySQL. For instance – how to add a new column to existing table. The candidate tried to convince us that there was no way to do that except creating another table with new column and copy data from old table there. Sure thing that we advised him to read about ALTER command.

However it turned out that suggested ‘method’ could have sense. According to Chris at Everything MySQL blog a full dump and reload to a table with new column may be much more faster than alter table! For instance there is shown a test with 5Gb table and alter table is 47% slower. Of course such ‘alternative alter’ should be used only against slave as well as some other restrictions may apply.

Below is the suggested process.

1. If you have a slave, AND YOU SHOULD AT THIS POINT, run stop slave
2. Run the SELECT INTO OUTFILE local, if you have the space, or over nfs (not very fast)
3. DROP or RENAME THE TABLE you are trying to ALTER
— RENAME only if you have the space and you don’t trust THE PROCESS!
4. CREATE the same table with the ALTERATIONS, if you dropped, or a new table with the ALTERATIONS
5. IMPORT the file from step 2
6. Either RENAME the two tables, if you renamed, or start slave
7. Fail over the writes to the newly altered slave
8. Repeat steps 1 – 7 for the old master