It uses twisted for xmlrpc.  Default threading method is async.  I am
just wondering, how come you guys didn't use twisted event loop?

Also, where (what file and format) do I config the thread pool size,
index store (FSDirectory) and such?

My understanding is Indexer works on a single object at a time.  There
is no batch indexing.  To do batch, you set RAMDirectory and flush the
buffer.  This is fine with my current requirement.

how do you assign different weights to title, description (I know this
is during the indexing)?

Thank you.
_______________________________________________
cps-users mailing list
[email protected]
http://lists.nuxeo.com/mailman/listinfo/cps-users

Reply via email to