The 2006 MySQL User Conference was this week. As part of his research, Tim O’Reilly asked several Web 2.0-related companies how they use databases. If you haven’t seen the posts on O’Reilly Radar, they’re Web 2.0 and Databases Part 1: Second Life, Database War Stories #2: bloglines and memeorandum, Database War Stories #3: Flickr, Database War Stories #4: NASA World Wind, and Database War Stories #5: craigslist.
There’s a lot for purists to hate, but given the choice between making it work and making it perfect, I’ll choose the former every time, then do my best to refactor toward the latter. (More on database agility next week.)
It’s somewhat draining hearing so many web 2.0 pitches which amount to ‘We’re going to work exactly like the tool you’re used to, but without synchronous server round-trips. Isn’t that great?!?’ No — that’s not great. It reminds of all the boom-startups who thought moving mail-order to the web was revolutionary … it’s just too obvious too really be intriguing.
So I didn’t have high hopes when I met a couple of folks working at one of my office-condo neighbors, and found out that their company (Kyube) is doing a web 2.0 CMS called SimplifyCMS. But they demo’d it for me, and I liked it a lot.
One of the recent MySQL enhancements was the INFORMATION_SCHEMA data dictionary, which gives a more coherent view on the DBMS meta-data.
Useful as it is, the information schema is becoming widely used, and many DBAs are learning how to use it as an administration and debugging tool. So useful, in fact, that it would be convenient, from time to time, to export its contents and pass it to a colleague to have a second opinion when a problem arises.
Unfortunately, there is no built-in tool to achieve this result. The standard mysqldump utility will skip the information schema on purpose. Since they are tables that are generated on-the-fly at server start-up, dumping them would not make sense to the normal user. However, the DBA is not a normal user, and such a dump would be highly desirable.
Here is a quick workaround that will produce a dump suitable for transferring the data to a new database.