This month’s Maatkit release, which just hit the download servers, contains a new feature for mk-query-digest: Apdex scores. The Apdex performance index is a sensible metric of how satisfactory the response time to a request is. Of all the scientific metrics I’ve seen, it’s the most practical; and of all the practical ones I’ve seen, it’s the most scientific.

To use this feature, you need to know what the Apdex threshold means, and you need to know how to configure this in mk-query-digest.

And then you can ask and answer practical questions such as “which classes of queries against this MySQL server are not satisfying users?” Or, because mk-query-digest supports the HTTP protocol, you can pose …

One of the major benefits from the CMake work available in MySQL 5.5 is that in the MySQL Release Engineering team we have been able to make it easy for users and developers to build MySQL exactly as we do for the official releases. For too long there has been a disconnect between the binaries produced as part of a regular ’./configure; make’ build and what we ship to users and customers.

We’re still not exactly where we want to be, there are still some parts which need to be integrated into the server tree, but for now it’s relatively straightforward to build exactly as we do.

Here are the instructions, using the mysql-5.5.6-rc.tar.gz source tarball as an example.

tar.gz

These are the generic instructions to build a tarball release. Note that we make use of CMake’s out-of-srcdir support to

• ensure the …

Based on discussions with several clients, we are strongly considering implementing a limited form of parallel replication. Single-threaded replication is one of the most severe limitations in the MySQL server.

We have a brief outline of the ideas at this wiki blueprint. So far, the "binlog order" idea is the only one that is workable. It has the added property that it is much more flexible in the future, so we could lift some of the restrictions. We will initially impose these limitations to keep things simple, and make sure we can actually get something working for our clients.

We're trying to help the users who are the most negatively affected by the limitations, and whose problems can be …

Python is a programming language that lets you work more quickly and integrate your systems more effectively. You can learn to use Python and see almost immediate gains in productivity and lower maintenance costs.

Python is a programming language that lets you work more quickly and integrate your systems more effectively. You can learn to use Python and see almost immediate gains in productivity and lower maintenance costs.

Sorry if you'll find this post stupid, but probably it'll safe a day for somebody else who may hit a wall like it was for me last Friday ;-))

So far, I've started my test scenario to analyze a workload with up to 1600 concurrent users on the 32 cores Linux server. Load is growing step by step by 200 users every 5 min, and once the load is reached 800-1000 users I've started to observe all my new clients ejected with the following error message:

"ERROR 1135 (HY000): Can't create a new thread (errno 11); if you are not out of available memory, you can consult the manual for a possible OS-dependent bug"

I may say, once you hit such an error on your MySQL server you're less than happy ;-)

What about the error?.. - this error is generic, and what is really important inside of the error message is its error code -- "errno 11" (well, of course you should check if you're really not short in memory, and in …

Python is a programming language that lets you work more quickly and integrate your systems more effectively. You can learn to use Python and see almost immediate gains in productivity and lower maintenance costs.

Hot Backup, Online Backup, Dynamic Backup - all synonyms for backups that run while at the same time database reads and writes continue on uninterrupted.

MySQL Enterprise Edition now includes a Hot Backup solution - MySQL Enterprise Backup, as the premier backup product for MySQL providing

• Fast and Low Impact Online/"Hot" backups for InnoDB
• Synchronized backups for Non-ACID MySQL engines and files
• Consistent point in time restore and recovery
• Full, Incremental, and Partial Backups
• Compression

We've built from a solid high quality InnoDB Hot Backup base and have and will continue to move it forward with new capabilities and features.

Learn more 

• Overview
  
• …

Register now for a free Percona webinar about migrating your MyISAM databases to InnoDB. Save the date: Dec 1, 2010 at 9:00 AM PST (California) time.

Update: the date was originally listed as December 2nd, but that was a mistake. It's December 1st.

We know that not everyone can attend expensive conferences that require travel. We're going to be doing highly technical webinars regularly, to help those people get some of the same kind of content they might get at the conferences where we present. This is our first such webinar.

With the change in default storage engine from MyISAM to InnoDB in the upcoming MySQL 5.5 release, we expect that a lot of users will be wanting to learn more about the benefits of switching to InnoDB, as well as topics such as:

• What kind of contraindications there are
• What the process might look …

I have a 5G mysqldump which takes 30 minutes to restore from backup.  That means that when the database reaches 50G, it should take 30x10=5 hours to restore.  Right?  Wrong.

Mysqldump recovery time is not linear.  Bigger tables, or tables with more indexes will always take more time to restore.

If I restore from a raw backup (LVM snapshot, xtrabackup, innodb hot backup), it is very easy to model how much longer recovery time will take:

Backup is 80G
Copy is at 70MB/s.
10G is already complete.
= ((80-10) * 1024)/70/60 = ~17 minutes


I can tell progress with mysqldump by monitoring the rate at which show global status like 'Handler_write'; increases and compare it to my knowledge of about how many rows are in each table.  But progress != a magic number like "17 minutes".  Not unless I do a lot of complex modeling.

I am …