Christiaan Kortekaas wrote:
> Hi All
> Just wondering if anyone has written a plugin to put all the eventum issue
> data into an Apache Solr index and keep it in sync?
> This would help with scaling (we have over 400k eventum issues and >3
> million full text emails) and allow for faceting (see a site like
> http://espace.library.uq.edu.au or http://www.jseeeker.com.au for how great
> search faceting can be) and much more powerful command line style searches.
Impressive, this is by far the largest installation I have heard of.
> No offence meant to MySQL on this, as they are not the same types of systems
> but Solr is great for search performance and features with a MySQL Innodb
> backend as the authoritative source. This is how we run the Fez software
> (http://sf.net/projects/fez) which itself its loosely based around Eventum
> code (though we are using Zend DB rather than Pear DB these days PDO is
Yes, using the InnoDB engine will provide much better performance. Internally at MySQL we
switched Eventum from MyISAM to InnoDB along time ago. I implemented a search plugin
system that would make it fairly trivial to add any indexing system. I used the very
excellent Sphinx search but it should be easy to work with Solr as well. Unfortunately I
haven't had time to merge this into the main tree, but there is some demand for it so I
will try to get it in within the next month. Feel free to harass me along the way to
remind me to get it done. Also if anyone else is interested in helping, please let me
Bryan Alsdorf, Lead Web Developer
Monty Program, AB. http://askmonty.org