Moving from MySQL to CouchDB: Part 1

I’ve started a little series on how to migrate your MySQL applications and databases over to CouchDB. Most of the process is about how you think about your data, not about the database itself, the application, or the interface to the database storage. There are some use cases for data storage that lend themselves to the CouchDB document model that provides some advantages over the table-based structure in MySQL. The first part of the series is Moving from MySQL to CouchDB: Part 1.

Left MySQL/Joined CouchOne

For many people this will be old news, but I guess It thought I should put up something official. At the end of September, I left MySQL/Sun/Oracle – that wasn’t an easy decision, mostly because I loved my job. It’s difficult to stop doing something that you enjoy so thoroughly and, over the years, have been so involved in. I did more than just get involved in the docs, I helped out with advice for different departments, worked on areas like DTrace, and of course helped write the documentation and enhanced many of the tools that enabled us to build such brilliant documentation. I managed to work with some amazing people, most of all the rest of my team who worked so hard to produce the manuals and content. The impetus to leave came from an opportunity to work with another excellent team on a different database, namely CouchDB. CouchDB reminds me of my early database work working on freeform text databases, with a nice open and easy structure, but with ways of getting standardized information out. I’ve joined CouchOne as Vice President of Documentation. The core of that is building an entirely new suite of documentation, starting from the ground up with everything from the build environment for the docs, to the content itself. Longer term there are lots of other things we are working on, but it will hang off that core reference documentation.

Reorganizing the documentation

Those of you that know the documentation well will be aware of the old page we used to have for the MySQL documentation. It was huge, and over the years we’d done a number of things to try and improve the layout and make it easier to find what you wanted. We had in-age links to jump to the different documentation types, and the old topic table that allowed you to jump to specific parts of the documentation. The problem was that the more documentation that we produced (and there are over a thousand docs in various formats now), the bigger the page got. When we added the individual topic guides, for example, we trebled the size of the page by adding the links for each individual topic guide. Ultimately that makes it increasingly difficult for you guys to find what you are looking for, despite the quick links and other elements. We’ve now changed all this and split the single, big, monolithic page of *every* piece of documentation that we create, and instead spread the documentation out over a number of pages. The actual documentation itself remains the same, and we still have the same range of documentation (in fact, it’s increased slightly as I’ve been able to squeeze in a few more formats and topic guide docs), but everything is still there. The key is the new sub-navigation bar that the Web team have provided us with:


The pages have been split out as follows:

  • MySQL Manual — the full, complete reference manuals
  • Workbench — the Workbench manuals
  • Expert Guides — the standalone guides for some of our more detailed products and system such as the Falcon storage engine and the MySQL Test Framework
  • Topic Guides — the topic reference, with the topic table at the top providing direct links into the 5.1 manual or standalone guides, and the full list of downloadable standalone guides.
  • MySQL Cluster — the full cluster manuals, including the guide to the MySQL Cluster API (NDBAPI)
  • Other docs — other documentation, not already mentioned, including the sample databases (Sakila, World, Employee), the help tables you can import into MySQL, and printed material and links elsewhere.
  • MySQL Uni — a page about the MySQL University, which is run by the documentation team, and which provides links to the MySQL Uni pages on Forge
  • About — information about the documentation team, who we are, and some statistics on the documentation we produce
  • Archives — archives of older manuals

We are aware of a few issues with some of the links to some documentation, and I’m working right now to address those problems, but all the documentation should be there and available. If it isn’t, please report a Bug.

Rebuilding the installation chapter

We have lots of things on the go right now (over and above the normal process of keeping things up to date), and one of the main projects for me is to do a complete rebuild of the installation chapter (Installing and Upgrading MySQL). I’ll be starting with the 5.1 manual, then the 5.4 manual. Any future manuals should be based on these so we should be up to date for future generations. What I’m doing:

  • Re-structuring the chapter to make it easier to follow on a platform basis. The old structure mixed content for different binary and source types, and different platforms, across a number of sections, making it very difficult to follow the instructions for your chosen platform.
  • Make some things generic. There are sections which are generic and apply to all (or at least many) different installation types.
  • Make some things more specific. Equally, there are some things that need to be spelled out more uniquely.
  • Remove some old, old, advice. We have notes in there going back 10 years or more. Among the favorite examples I’ve found is a piece of advice that says ‘If your machine has more than 16MB of RAM…’. These things are not helpful in the manual, and may just serve to confuse some people.
  • Remove some older platforms. Some of the platforms and advice go back and predate MySQL 5.1, and even MySQL 5.0 and 4.1. In many cases the OS information is for a system either no longer actively developed or supported (FreeBSD 3.x, or Solaris 2.5, for example). Again, we want to remove some ambiguous and potentially confusing information and advice here for platforms which we simply can no longer monitor.
  • Make it easier to keep up to date. The problem with the old organic structure is that knowing where to add new content, improvements, extensions, etc. becomes harder and harder. by merging and unifying the structure we will improve this, and in turn, improve the ability to find information.

In practice that means for at least the next month or so you will see a number of improvements and restructuring in the installation chapter for 5.1 and later manuals. I already have a list of about 35 items that need to be addressed, over and above the list above, but feel free to provide any additional suggestions and I’ll see what I can do to fit them in.

Dojo examples from UC2009

I know, I know, loads of people have been waiting for these…So here we go, I’ve finally sorted a downloaded version of the Dojo examples from the presentation I provided at the MySQL Users Conference 2009. There are three examples:

  • The auto-paging table example, which uses the functionality of the Dojo Toolkit and the QueryReadStore to automatically load content from a table.
  • The basic graphing example, which loads data dynamically and plots a graph.
  • And the zooming version of the same basic graph interface
  • There’s a README in the download that contains instructions on getting everything up to speed, although it should be fairly obvious. It’s attached to the bottom of this post too.Any questions for getting this to work, please go ahead and ask!Download the package hereUC2009 Working with MySQL and Dojo==================================MC Brown, 2009, and http://coalface.mcslp.comComponents:There are three examples here:- Dojo table with auto-paging (table_autopage.html and table_autopage.cgi)- Dojo Basic Graph (graph_basic.html and graph_ajax.cgi)- Dojo Zooming Graph (graph.html and graph_ajaz_zoom.cgi)You will also need the Dojo/Dijit and DojoX toolkit package from here: the combined package, and then extract the contents and place in the same directory as the CGI scriptsIntructions for use:For the Table example, you can create a table (or use an existing DB) from WordPress using the wp_posts tabls.In practice you need only create and populate a table with the following columns:id (int)user_nicename (char)post_date (datetime)post_title (char)To be compatible with the example without any more modifications.For the Graphing examples, you need a table as follows:CREATE TABLE `currencies` ( `currencyid` bigint(20) unsigned NOT NULL auto_increment, `currency` char(20) default NULL, `value` float default NULL, `datetime` int(11) default NULL, `new` int(11) default NULL, PRIMARY KEY (`currencyid`), UNIQUE KEY `currencyid` (`currencyid`), KEY `currencies_currency` (`currency`,`datetime`), KEY `currencies_datetime` (`datetime`)) ENGINE=MyISAM AUTO_INCREMENT=170048 DEFAULT CHARSET=latin1And then populate with date/value data according to the information you want to store.Once done, make sure you change the database information in the CGI scripts to populate the information correctly during operation.

MySQL 5.1 in OpenSolaris

If you’ve attended just one of my recent talks, either at the UC, LOSUG or MySQL University, you should know that MySQL 5.1.30 will be in the next official drop of OpenSolaris. In fact, you can find MySQL 5.1 in the current pre-release builds – I just download build 111 of the future 2009.06 release. Key things about the new MySQL 5.1 in OpenSolaris:

  1. Contains the set of DTRACE probes that also exists in MySQL 5.4 (see DTrace Documentation)
  2. Like the 5.0, we have SMF integration, so you can start, stop, monitor and change some of the core configuration through SMF
  3. Directory layout is similar to 5.0, with a version specific directory (/usr/mysql/5.1), and the two can coexist if you want to handle a migration from 5.0 to 5.1

To install MySQL 5.1, use the pkg tool to perform the installation. We’ve split the components into three packages:

  1. SUNWmysql51 contains the server and client binaries, and associated scripts.
  2. SUNWmysql51libs contains the client libraries, which you’ll need for all external (i.e. MySQL) tools, like DBD::mysql for Perl connectivity)
  3. SUNWmysql51test contains the MySQL test suite

To install:

$ pfexec pkg install SUNWmysql51

Once installed, you can start MySQL server through SMF:

$ pfexec svcadm enable mysql:version_51

You can set properties, like the data directory, by using svccfg:

$ svccfg svc:> select mysql:version_51 svc:/application/database/mysql:version_51> setprop mysql/data=/data0/mysql svc:/application/database/mysql:version_51> setprop mysql/enable_64bit=1

Any questions, please ask!

Speaking at CommunityOne West

Sorry for the (relatively) short notice, but I will be talking at Sun’s CommunityOne conference in San Francisco on June 1st.I’ll be talking about, and demonstrating, the DTrace probes we have put into MySQL in a joint presentation with Robert Lor who will be doing the same for Postgres.


CommunityOne West Badge
Our presentation is on the Monday afternoon. Check out the CommunityOne West Conference Site for more details and registration.

Presentations now on MySQL Conf Site Too

So I talked to the wonderful JB from O’Reilly who is responsible for all sorts of parts of the conference, and she’s now enabled my presentations on the MySQL Conference Presentations page. While you’re there downloading mine, make sure you go and download some of the others. In particular, you might want to try:

UC2009: Scale Up, Scale Out and High Availability: Solutions and Combinations

My presentation slides for the Scale Up, Scale Out and High Availability tutorial here at the MySQL Users Conference are now available for download: Scale Up, Scale Out and High Availability: Solutions and CombinationsThese have been on the MySQL Conference website for days now, but for some reason they don’t seem to have been freed yet. To those waiting I’m sorry for the delay in getting these uploaded.