I recently wrote about non-deterministic queries in the replication stream. That’s resolved by using either MIXED or ROW based replication rather than STATEMENT based.

Another thing that’s not fully handled by STATEMENT based replication is temporary tables. Imagine the following:

1. Master: CREATE TEMPORARY TABLE rpltmpbreak (i INT);
2. Wait for slave to replicate this statement, then stop and start mysqld (not just STOP/START SLAVE)
3. Master: INSERT INTO rpltmpbreak VALUES (1);
4. Slave: SHOW SLAVE STATUS \G

If for any reason a slave server shuts down and restarts after the temp table creation, replication will break because the temporary table will no longer exist on the restarted slave server. It’s obvious when you think about it, but nevertheless it’s quite …

Scaling up a workload to many cores on a single host

Here are results for Shard-Query 2.0 Beta 1* on the Star Schema Benchmark at scale factor 10.  In the comparison below the “single threaded” response times for InnoDB are the response times reported in my previous test which did not use Shard-Query.

Shard-Query configuration

Shard-Query has been configured to use a single host.  The Shard-Query configuration repository is stored on the host.  Gearman is also running on the host, as are the Gearman workers.  In short, only one host is involved in the testing.

The …

As you probably know, Tungsten Replicator can replicate data from MySQL to MongoDB. The installation is relatively simple and, once done, replication works very well. There was a bug in the installation procedure recently, and as I was testing that the breakage has been fixed, I wanted to share the experience of getting started with this replication.

Step 1: install a MySQL server

For this exercise, we will use a MySQL sandbox running MySQL 5.5.31.

We download the binaries from dev.mysql.com and install a sandbox, making sure that it is configured as master, and that it is used row-based-replication.

$ mkdir -p $HOME/opt/mysql
$ cd ~/downloads
$ wget …

Shard-Query examines INFORMATION_SCHEMA.PARTITIONS to determine if a table is partitioned.  When a table is partitioned, Shard-Query creates multiple background queries, each limited to a single partition.  Unfortunately, it seems that examining INFORMATION_SCHEMA.PARTITIONS causes query plans to change after the view is accessed.

I have reported bug 69179 to MySQL AB  Oracle Corporation(old habits die hard).

Be careful: If you have automated tools (like schema management GUI tools) then make sure they don’t examine INFORMATION_SCHEMA.PARTITIONS or you may get bad plans until you analyze your tables or restart the database, even if using persistent stats.

I can only get the bug to happen when a WHERE clause is issued that limits access to a single partition.  It may be that the per partition statistics …

As you might know, I'm a big fan of JSON. One big reason is that I believe that JSON is closer to most developers view on data, whereas the Relational SQL based model is closer to what someone working with data itself or someone working with infrastructure. What I mean here is that neither view is wrong, but they are different.

So, given that, can we merge the Object JSON world with the relational model? Well, not JSON, but Hibernate does it quite well. This is one of my objects to the NoSQL world, that the datamodel is closely linked to the application at hand, and less so to data itself and to other applications. Stuff such as accounts, privileges, accounting data, orders and many other things are global, and are not specifically connected a specific application, but in many NoSQL applications, this is what it ends up being.

And there are not that many good solutions, how can I easily explore data in a NoSQL …

Backups is one of the most important part of any MySQL deployment, and nowadays, there’s a number of tools to choose from depending on how your organization implements them. The purpose of this post is to enumerate the main tools and some helpers that makes backing up and testing/restoring your backups more convenient. By all means this is not the complete list, I’m sure I am missing some, so feel free to add them through the comments.

The Core Tools

• mysqldump – is a logical backup tool for MySQL. It creates plain text files with SQL statements which you can directly import back to the server. Some would say mysqldump is not really a backup tool as you cannot get a consistent backup without disrupting operations while the server is running. I’d say this is just a limitation, if your dataset is small …

As far as I know there isn't a well accepted set of best practices for MySQL, but there are many best practices known and used by most MySQL DBA's. One of those best practices is that the datadir must not be equal to a mountpoint; it has to be a subdirectory of a mountpoint. I learned this the hard way a few years ago when I used a NetApp Filer via NFS as a data directory. The NetApp filer exposed the snapshots via a .snapshot directory. A database in MySQL is a directory, so MySQL thought that the .snapshot directory was a database. This resulted in some issues with our monitoring scripts, so we had to create a mysql_data directory and move all the databases to that directory.

For other setups directories like lost+found, .zfs, etc. gave similar issues.

In MySQL 5.6 a new feature was introduced to make it possible to make some databases hidden. To do this a ignore-db-dir option needs to be specified for each directory. …

Some of you know that I have published how to install Oracle VM Manager (OVMM) on a Dom0 host since Oracle released the Oracle VM 3. I have described why you possibly may want to do it in my very first post. Just want to mention here that it should be used for sandbox configurations only. You can find the previous post on how to install 3.1.1 OVMM version under Dom0 here. This time I talk about 3.2.2 version.

NOTE: At the time of writing ORACLE VM 3.2.3 SERVER (Patch 16410428) and ORACLE VM 3.2.3 MANAGER (Patch 16410417) became available. I didn’t have time to install the latest versions yet. However I do not expect that installation are significantly …

Hi guys, i was reading the new query cache plugin from Roland Bouman, now default in mariadb-5.5.31
This is a very old feature request at mysql (27 Oct 2006 12:31):
http://bugs.mysql.com/bug.php?id=23714

And recent mariadb (thanks Sergei reading my MDEV =) ) (2012-05-04 01:22)
https://mariadb.atlassian.net/browse/MDEV-249

Well this is a very nice piece of code...
Every time i google about "mysql performace", "mysql cache" etc ... i get something like this:
…

This week we talk about server and status variables relating to the performance schema and the ps_helper tool. Ear Candy is an sql_mode bug, and At the Movies is a performance_schema and ps_helper webinar.

Performance Schema Variables
MySQL 5.5 performance schema variables
MySQL 5.6 performance schema variables - many options are auto-sized


SHOW GLOBAL VARIABLES LIKE 'performance_schema%';

read more