Interesting that the article references a post I made here. In the end I did not end up needing that lucene code change for two reasons.
1) I ended up just clustering the RAMDirectory itself so the subclass of the collection ended up not being shared 2) In the coming release (early december) we'll be supporting subclasses of collections. Pretty neat stuff. Doing the exercise ended up leading to some performance tuning that made it work pertty well. That is also coming in the early december release. Cheers, Steve On 11/9/06, karl wettin <[EMAIL PROTECTED]> wrote:
Some people might find this interesting. I have personally not looked at it in depth: Engineers at TerraCotta have detailed a new way to cluster lucene, the popular text search library from Apache. Their method involves implementing the lucene RAMDirectory interface and using TerraCotta DSO to share the RAMDirectory across JVMs. http://www.infoq.com/news/2006/11/clustering-lucene-terracotta
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
