MySQL@Facebook
I have written about the new support for group commit in the InnoDB plugin and the lack of support for it in the binlog. I will attempt to describe how commit is done for MySQL 5.1 and the InnoDB plugin. The current code can limit performance depending on your workload...
MySQL@Facebook
How does InnoDB do on a server that supports 10,000 IOPS? I am sure this wasn't the first thing on Heikki's mind as he was designing InnoDB 10 to 15 years ago...
MySQL@Facebook
MySQL@Facebook
@Adam - the equivalent of a fast disk array. You should be able to repeat some of the results with a RAM disk.
22. joulukuuta 2009 kello 6:19
MySQL@Facebook
InnoDB computes a checksum immediately prior to writing a page to disk and stores that on the page. It validates the checksum immediately after reading the page back from disk. I love this feature but I don't love the implementation of the checksum function and neither does Domas...
MySQL@Facebook

MySQL@Facebook There are several feature requests for INTERSECT and EXCEPT. I guess bugs implements bag semantics. See http://bugs.mysql.com/bug.php?id={1309,26477,31336,33200}

en.wikipedia.org
In SQL the UNION clause combines the results of two SQL queries into a single table of all matching rows. The two queries must have the same number of columns and compatible data types to unite. Any duplicate records are automatically removed unless UNION ALL is used.
MySQL@Facebook
MySQL@Facebook
I don't think the author of the 'elementary' comment works at MySQL/Sun.
21. joulukuuta 2009 kello 8:33
MySQL@Facebook

MySQL@Facebook Hooray, Sun/MySQL is a founding sponsor for the user conference. It would not have been the same without them.

datacharmer.blogspot.com
Data seem sometimes to have their own life and will, and they refuse to behave as we wish. Then, you need a firm hand to tame the wild data and turn them into quiet and obeying pets.
MySQL@Facebook

MySQL@Facebook Dear lazyweb, where is posix_fadvise on Mac OS X? The Facebook patch doesn't build on a Mac.

18. joulukuuta 2009 kello 13:33
MySQL@Facebook

MySQL@Facebook MySQL 5.5 is here with semi-sync replication. I like that feature. Wei did a great job on it.

datacharmer.blogspot.com
Data seem sometimes to have their own life and will, and they refuse to behave as we wish. Then, you need a firm hand to tame the wild data and turn them into quiet and obeying pets.
MySQL@Facebook

MySQL@Facebook There is a Facebook group for the 2010 User Conference.

The MySQL Conference & Expo is an interactive learning community--four days packed with connections to MySQL developers and open source experts who know their subject inside and out. You'll gain uniqu...
MySQL@Facebook
I need to hack MySQL to stop parsing LIMIT a, b. This is a high-risk feature for an OLTP RDBMS. I saw the following on a server that was already too busy. select ... from foo where ... LIMIT 2000000, 10; select ... from foo where ... LIMIT 2000010, 10; select ... from foo where .....
MySQL@Facebook
MySQL@Facebook
MediaWiki paging classes are index-based \o/ -- Domas
15. joulukuuta 2009 kello 14:35
MySQL@Facebook

MySQL@Facebook Yes, Virginia, there is a conference

en.oreilly.com
O'Reilly MySQL Conference & Expo
MySQL@Facebook

MySQL@Facebook Do you know about Galera for synchronous replication in MySQL? It is interesting, but I need to read more about it.

codership.blogspot.com
We just made a major software release, but I still don't see journalists queuing outside our office. Looks like I have to do the hard work and interview myself. In the following, I'll give rough reporter ...
MySQL@Facebook

MySQL@Facebook DRBD has been merged into the Linux kernel.

fghaas.wordpress.com
DRBD has entered a new phase. After being developed out of tree for 9 years, and after an extended review and streamlining phase since March, Phil submitted DRBD to be merged into 2.6.32 release of the Linux mainline kernel. ...
MySQL@Facebook
MySQL@Facebook
I (MarkC) think DRBD is great. Domas probably agrees. Can't speak for the others.
8. joulukuuta 2009 kello 17:56
MySQL@Facebook
How do you determine when InnoDB readahead is not working? Use MySQL 5.0, make your server IO bound and look for a server that sustains a high number of aio reads as reported by SHOW INNODB STATUS...
MySQL@Facebook
InnoDB supports group commit in the plugin. The binlog does not. There has been a feature request for this in InnoDB for a long time. There wasn't a request for the binlog until now -- see bug 49326. The previous note from Domas describes one workaround -- use XFS...
MySQL@Facebook
MySQL@Facebook
Don't mind me. I am still bitter about my_bool. Ask Stewart.
4. joulukuuta 2009 kello 13:05
MySQL@Facebook
We've noticed contention in binary log syncing to disk, so I got back to micro-benchmarking edge cases...
MySQL@Facebook
MySQL@Facebook
Stewart, I didn't see too many perf increases with p_fallocate() - XFS does reasonable preallocation itself anyway, but still has to sync metadata on each append. -- Domas
3. joulukuuta 2009 kello 8:53