In the new InnoDB lab release we support page sizes of 32k and 64k. This gives users even more choices on the page size, allowing you to further customize InnoDB for your particular workload.

There are some things worthy of note related to this new feature:

1. The extent size changes when the innodb_page_size is set 32k or 64k.
   The extent size is 2M for 32k page sizes, and 4M for 64k page sizes (the extent size is 1M for 4k, 8k, and 16k page sizes). If we do not enlarge the extent size then we will have too many extent headers on the allocation bitmap page, and the bitmap page will overflow.
2. The …

I still get asked about the lost+found directory enough, and so I wanted to provide a current update.

The lost+found directory is a filesystem directory created at the root level of a mapped drive. Thus this is common to see if you create your mysql datadir at the root level of a mapped drive.

In the past, you could ignore it, if it wasn’t too problematic for you, or you could move your datadir down a level, and then it wouldn’t be created in the datadir anymore.

However, there is now the –ignore-db-dir option. It is actually not too new (it’s been in MariaDB since 5.3.9 and 5.5.28, and in MySQL as of 5.6.3), but I don’t think many are too familiar with it.

But when you do run into this problem, some/many would prefer to add a single line to the config file rather than move the datadir.

To do this, just add the following option to your my.cnf file, under the [mysqld] section (it cannot be set …

I have recently posted up benchmarks of libAttachSQL vs. libmysqlclient using sysbench. Whilst these are great and shows the results I hoped for, this isn't what we designed libAttachSQL for. It was designed for non-blocking many connections per thread.

With this in mind I spent today knocking up a quick benchmark tool which replicates the Sysbench "Select" test but using libAttachSQL's connection groups on a single thread. The code for this can be seen in the new AttachBench GitHub tree. Of course the secondary reason for this is to try and hammer the connection groups feature, which of course did find a bug when I scaled to around 32 connections. This has been fixed in libAttachSQL's master ready for …

Padding Oracle On Downgraded Legacy Encryption

First off, the naming “convention” as of late for security issues has been terrible. The newest vulnerability (CVE­-2014-3566) is nicknamed POODLE, which at least is an acronym and as per the header above has some meaning.

The summary of this issue is that it is much the same as the earlier B.E.A.S.T (Browser Exploit Against SSL TLS), however there’s no known mitigation method in this case – other than entirely disabling SSLv3 support, in short, an attacker has a vector by which they can retrieve the plaintext form your encrypted streams.

So let’s talk mitigation, the …

Founded in 1998, Rackspace has evolved over the years to address the way customers are using data – and more specifically, databases. The San Antonio-based company is fueling the adoption of cloud computing among organizations large and small.

Today Rackspace is doubling down on open source database technologies. Why? Because that’s where the industry is heading, according to Sean Anderson, Manager of Data Services at Rackspace. The company, he said, created a separate business unit of 100+ employees focused solely on database workloads.

The key technologies under the hood include both relational databases (e.g., MySQL, Percona Server, and MariaDB) and NoSQL databases (e.g., MongoDB, Redis, and Apache Hadoop).

Last July Rackspace …

It has been just over 4 months since I started working on libAttachSQL for HP's Advanced Technology Group. Today marks the first (and hopefully only) RC release of the library.

Connection Groups

The only real new feature that has been added to 0.9.0 is the concept of connection groups which is something I'm pretty excited about. Internally libAttachSQL uses event loops to supply the non-blocking API. Connection Groups join a bunch of connections together into a group that uses a single event loop. This makes things much more efficient internally and makes applications easier to code too.

Here is a simplified example of how to use it (for a more detailed example see our example in the documentation).

First we need to create the group and add connections to …

A few months back, Michael wrote about reconnecting orphaned *.ibd files using MySQL 5.6. I will show you the same procedure, this time for partitioned tables. An InnoDB partition is also a self-contained tablespace in itself so you can use the same method described in the previous post.

To begin with, I have an example table with a few orphaned partitions and we will reconnect each partition one by one to the original table.

mysql [localhost] {msandbox} (recovery) > SHOW CREATE TABLE t1 G
*************************** 1. row ***************************
Table: t1
Create Table: CREATE TABLE `t1` (
[...]
KEY `h_date` (`h_date`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1
/*!50100 PARTITION BY RANGE (year(h_date))
(PARTITION p0 VALUES LESS THAN (2006) ENGINE = InnoDB,
PARTITION p1 VALUES LESS THAN (2010) ENGINE = InnoDB,
PARTITION …

The new InnoDB Labs release contains the ability to create and use independent multi-table general tablespaces.

This feature will provide a way to group tables together into tablespaces at a location and filename of your choosing.  Tables using row formats of Redundant, Compact, and Dynamic can be combined together into the same general tablespace. Compressed tables with the same key_block_size can also be combined together.

The SQL syntax for creating an empty general tablespaces is:

CREATE TABLESPACE `tblspace_name` ADD DATAFILE 'tablespace.ibd' [FILE_BLOCK_SIZE=n];

The filename can contain an absolute path or a path relative to the data …

Percona Toolkit’s  pt-table-checksum performs an online replication consistency check by executing checksum queries on the master, which produces different results on replicas that are inconsistent with the master – and the tool pt-table-sync synchronizes data efficiently between MySQL tables.

The tools by default use the CRC32. Other good choices include MD5 and SHA1. If you have installed the FNV_64 user-defined function, pt-table-sync will detect it and prefer to use it, because it is much faster than the built-ins. You can also use …

This year’s MySQL Central at Oracle Open World was an exhilarating experience. In contrast to the previous year’s MySQL Connect events, MySQL have now got their own Central at the main Oracle Open World. In the previous years, we were always short on time and trying to get a lot of sessions into just two days was just to much. This time I could both present sessions, attend sessions by other users, and also to talk to people in the MySQL community: something that I really enjoy and also find very valuable to see where we should be heading.

This year, the “MySQL Fabric Team” representation on MySQL Central was me and Narayanan Venkateswaran, which is heading the sharding solution in MySQL Fabric. Together with the conference, we also released MySQL Fabric 1.5.2 as the GA release of MySQL Fabric 1.5 containing a few new features:

• Server …